Professional Documents
Culture Documents
PART 2 - FUTURES
PART 3 - OPTIONS
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 2/358
PART 2 - FUTURES
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 3/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 4/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 5/358
PART 3 - OPTIONS
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 6/358
11 Valuation of Options
Risk-Neutral Valuation
Riskless Hedge Portfolio Using a Simple Binomial Model
Risk-Neutral Valuation Using the Binomial Model
Risk-Averse Valuation Using the Binomial Model
Commodity Price and Return Distributions
Risk-Neutral Valuation of European Call Option
Evaluation of X Pr ob(ST X) ; of E ( ST | ST X ) and of E (cT )
Current Value of Call
Risk-Neutral Valuation of European Put Option
Properties of the European Call and Put Option Pricing Formulas
Change in Commodity Price
Percentage Change in the Commodity Price
Change in Delta
Change in the Exercise Price
Change in the Cost-of-Carry Rate
Change in the Interest Rate
Change in the Volatility
Change in the Time to Expiration
European Exchange Option
Valuation of American Options
Estimation of the Option Pricing Parameters
Historical Volatility and Implied Volatility Estimation
Summary
Appendix 11.1
Appendix 11.2: Approximations for the Cumulative Normal Density Function
Appendix 11.3: Cumulative Normal Probability Tables
Appendix 11.4: Partial Derivatives of European Commodity Option Valuation
Equations
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 7/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 8/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 9/358
1.4 SUMMARY
A look at the business pages of the newspaper reveals a bewildering array of price quotations for futures and
options. While futures contracts on agricultural commodities have been with us since the mid-1800s, futures
trading in financial assets-such as bonds, currencies, and stock indexes were introduced as recently as 1975
and have grown at an explosive rate since that time. Likewise, trading of options on financial and agricultural
commodities is a relatively recent event, dating to the founding of the Chicago Board Options Exchange in
April 1973. Today, call options trade on five U.S. exchanges in nearly 400 common stocks. Moreover,
options also trade on bonds, foreign currencies, stock indexes, and traditional agricultural commodities.
While futures and options contracts on a variety of underlying commodities1 have been developed,
certain principles of valuation and price behavior are common across all commodities. For example, the
essence of the price relation between the futures contract and its underlying commodity is captured by a
simple arbitrage argument, even though the types of commodities range from agricultural to purely financial.
In this book, we emphasize the principles that determine the value of a futures contract, an option contract,
and a futures option contract relative to the value of its underlying commodity. For example, consider the
S&P 500 stock index to be the underlying commodity. The index is a value-weighted stock portfolio
consisting of 500 large common stocks that trade predominantly on the New York Stock Exchange. The
Chicago Mercantile Exchange lists a futures contract on the S&P 500, as well as an option contract on the
futures, while the Chicago Board Options Exchange lists an option contract on the index itself. In other
words, for this particular commodity –the S&P 500 stock index portfolio– the four markets depicted in
Figure 1.1 trade simultaneously. Inextricable linkages exist among prices in these four markets, and, in this
book, we identify the nature of these price linkages and the implications they have for expected return/risk
management. In this chapter, we begin by defining futures, options, and futures options.
1
In this book, the term “commodity” is defined as being something of value. The commodity may be a foodstuff such as wheat, a
currency such as the Japanese yen, or a stock index such as the S&P 500.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 10/358
The profit from a long futures position initiated at price F0 is plotted in Figure 1.2a. For every dollar price
rise above F0 the investor makes one dollar. For every dollar decline below F0 , the investor loses one dollar.
The profit from a short position initiated at the futures price, F0 , is shown in Figure 1.2b.
As an example, suppose someone buys the July 1992 wheat futures contract listed in Table 1.1. At
the close of trading on November 13, 1991, the futures price ( F0 ) is reported to be $3.21 per bushel. The
denomination of this wheat futures contract is 5,000 bushels. Suppose further that the individual reverses his
position on the following February 6 by entering into a contract to sell July wheat. If the price of the July
wheat futures ( FT ) is $4.00 on February 6, the individual earns a profit of ($4.00 - $3.21) x 5,000 =
$3,950.00. Alternatively, if the futures price on February 6 is $3.00, the individual earns a profit of ($3.00 -
$3.21) x 5,000 = $1,050.00. Most futures contracts are, in fact, reversed in this manner prior to expiration.
Futures contracts are a means for reducing risk or assuming risk in the hope of profit, not a means of taking
possession of the underlying commodity. Users of the underlying commodity generally prefer a grade and
delivery location of the underlying commodity that are different from the grades and locations allowed under
the terms of the futures contract.
As a second example, consider the March 1992 S&P 500 index futures reported in Table 1.2.
Suppose someone buys this contract at the close of trading on November 13, 1991, at the reported price ( F0 )
of $400.35. The contract size for the S&P 500 futures is 500 times the price or $400.35 x 500 = $200,175.
Suppose the position is reversed on February 6 when the futures price ( FT ) is, say, $410.00. In this case, the
individual earns a profit of ($410.00 - $400.35) x 500 = $4,825.00. If, instead, the index falls to, say,
$390.00, the individual has a loss of ($400.35 - $390.00) X 500 = $5,175.00.
Every futures contract entered into has two sides: a willing buyer and a willing seller. If one side of
the contract makes a profit, the other side must make a loss. All futures markets participants taken together
can neither lose nor gain-the futures market is a zero-sum game.
FIGURE 1.1 Interrelations Between Commodity Market and Markets for the Commodity's Derivative
Instruments
Option
Commodity
Market
Market
Futures Futures
Market Options
Market
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 11/358
Profit
(a) Long Futures Position
Terminal
Futures
Initial Futures Price, F0 Price, FT
F0
F0
Terminal
Futures
Initial Futures Price, F0 Price, FT
TABLE 1.1 Prices of wheat futures contracts at the close of trading on Wednesday, November 13, 1991.
Lifetime
Open High Low Settle Change High Low Interest Open
---GRAINS AND OILSEEDS---
WHEAT (CST) 5,000 bu.; cents per bu.
Dec 347 ½ 352 ¾ 347 ½ 352 ½ +5¼ 369 ¼ 272 ½ 19,480
Mr92 349 ½ 353 ¾ 348 353 ½ +5¾ 367 279 23,394
May 334 338 ½ 334 338 +5¼ 352 ½ 280 ½ 5,174
July 317 ½ 321 316 ¾ 321 +4¾ 337 ½ 279 8,067
Sept 326 326 326 326 +5 341 292 690
Dec 333 ½ 335 332 335 +4¾ 351 329 ½ 703
Est vol 15,500; vol Tues 11,502; open lnt 57,508, -149.
Source: Reprinted by permission of Wall Street Journal, (November 14, 1991) Dow Jones & Company, Inc.
All. Rights Reserved Worldwide.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 12/358
FUTURES
TABLE 1.2 Prices of S&P 500 stock index futures contracts at the close of trading on
Wednesday, November 13, 1991.
S&P 500 INDEX (CME) 500 times index`
Open
Open High Low Settle Chg High Low
Interest
Dec 395.00 398.50 394-30 398.30 +1.00 401.50 316.50 139,341
Mr92 396.80 400.50 396.50 400.35 +1.00 404.00 374.70 7,544
June 398-30 402.35 398.30 402.20 +1.10 407.00 379.00 1,102
Est vol 42,125; vol Tues 41,413; open lnt 148,048, +916.
Indx prelim High 397.42; Low 394.01; Close 397.42 +.68
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 13/358
price of the underlying commodity increases, because, in that case, the call will be exercised, and the call
seller will be requested to purchase the underlying commodity at ST and deliver it to the call buyer at X .
The maximum gain to the call seller is C0 . It is evident from Figures 1.3a and 1.3b that the sum of the profits
of the call buyer and call seller at any terminal price, ST is zero. As in the case of futures markets, the option
market is also a zero-sum game.
The position of a put buyer –a long put position– depicted in Figure 1.4a is profitable if the price of
the underlying commodity falls below the exercise price by more than P0 , the initial price of the put. If the
price exceeds the exercise price at expiration ( ST X ), the put is not exercised. The maximum loss to the
put buyer is P0 and the maximum profit is X P0 . The position of the put seller-the short put position-
depicted in Figure 1.4b is the reverse image of the put buyer's position. The put seller has a maximum gain of
P0 and a maximum loss of X P0 .
Option buyers may choose to realize profits by exercising their options as we have just discussed.
More frequently, however, option positions are closed out by selling the option. At expiration, an option may
be sold for its exercise value. Before expiration, options usually sell for more than their exercise value. As a
result, it is usually, but not always, preferable to close out an option position prior to expiration by selling the
option rather than by exercising it. The gain or loss on the option is then just the change in the price.
An example using S&P 100 index options will, perhaps, make this discussion more concrete. Table 1.3
contains the closing prices of S&P 100 index options on November 13, 1991. The S&P 100 option contract
size is 100 times the index value. The December call with an exercise price of $370 has a reported price of
$7.375. This means that a call option buyer would pay a premium of $7.375 X 100 = $737.50 for the right to
“buy” the S&P 100 stock index at $370 x 100 = $37,000 any time before the expiration date2. If the index
level rises from $371.21 on November 13 to, say, $400.00 on November 30, the call option buyer can
exercise her option to earn a profit of ($400.00 - $370.00 - $7.375) x 100 = $2,262.50. Alternatively, the
option buyer can reverse her position by selling an option contract with the same terms. Suppose that on
November 30 the price of the call is $32.00. If the buyer of the option on January 16 chooses to sell the same
option, she realizes a profit of ($32.00 - $7.375) x 100 = $2,462.50. Note that closing the position by selling
the option as opposed to exercising it yields an additional profit of $2,262.50 - $2,462.50 = $200.00.
The fact that the option can be sold for more in the marketplace than the intrinsic value reflects the
time value of the option. On November 13, the 370 call can be sold for $7.375 and yet, if that same option is
exercised on November 13, its value is $371.21 - $370.00 = $1.21. The difference between the two values is
called the time value of the option and reflects the probability that the stock index will rise significantly from
its current level by the third Friday in February. The factors affecting the level of option premiums for
different striking prices and different maturities are described later in the book. The price of the 370
December put option in Table 1.3 is reported to be $5.25. This option is out-of-the-money since the current
value of the stock index exceeds 370. Someone who buys this put option on November 13 earns a positive
profit if the index level falls below $370.00 - $5.25 = $364.75 before the third Friday in December 1992.
It is important to recognize that the option writer (seller) faces payoffs exactly opposite those of the
buyer. If, for example, a call option is in-the-money at expiration, the option writer must deliver a
commodity worth more than the exercise price received by the writer. In terms of the call option example
above, when the buyer of the option chooses to exercise, the option seller in effect has to purchase the index
at $400.00 and deliver it to the option buyer at $370.00. This loss is offset in part by the premium, $7.375,
that the writer collected at the outset. The net loss to the writer equals the option buyer's net profit. On the
other hand, if the index level stays below the exercise price of the call until expiration, the option will not be
exercised, and the option writer keeps the premium collected from the buyer when the option contract was
written. This time the writer makes a positive profit, but, again, it is equal to the buyer's loss.
2
S&P 100 index options are American-style and-expire on the Saturday following the third Friday of the contract month.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 14/358
FIGURE 1.3 Profit Diagrams for Call Option Positions Held to Expiration
Profit
(a) Long Call
Exercise Price, X
Terminal Price of
Underlying
-C0 X+C0 Commodity, ST
C0 X+C0
Terminal Price of
Underlying
Exercise Price, X Commodity, ST
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 15/358
FIGURE 1.4 Profit Diagrams for Put Option Positions Held to Expiration
Profit
(a) Long Put
Exercise Price, X
Terminal Price of
X-P0 Underlying
-P0 Commodity, ST
P0
X-P0
Terminal Price of
Underlying
Commodity, ST
Exercise Price, X
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 16/358
TABLE 1.3 Prices of S&P 100 stock index option contracts at the close of
trading on Wednesday, November 13, 1991.
OPTIONS CHICAGO BOARD
S&P 100 INDEX- $100 times Index
Strike
Calls-Last Puts - Last
Price
Nov Dec Jan Nov Dec Jan
335 … … … 1/16 … …
340 30 31 ½ 31 ¼ 1/16 5/8 1½
345 23 27 28 1/16 7/8 2 1/8
350 21 ½ 21 1/8 … 1/16 13/16 3
355 16 3/8 19 195/8 1/16 1¾ 37/8
360 11 ¼ 14 ¾ 17 1/8 21/2 4½
365 6 3/8 10 5/8 13 ½ 5/16 3 5/8 6
370 2 5/16 7 3/8 10 ½ 1 1/16 5¼ 7¾
375 3/8 4 3/8 7 3/8 4 1/8 7 5/8 10 7/8
380 1/16 2 7/16 5¼ 9¼ 10 7/8 13 3/4
385 1/16 1¼ 3 1/8 … 151/4 …
390 1/16 5/8 1 7/8 18 5/8 20 20 ¾
Total call volume 142,982 Total call open Int. 387,114
Total put volume 128,390 Total put open Int. 421,693
The Index: Vol 371.22; Low 367.54; Close 371.21, +0.92
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
Allocation of Risk
Futures and options provide an efficient mechanism for allocating risk from those who wish to avoid risk to
those who are interested in bearing the risk. Futures contracts tend to arise when the underlying commodity
is costly or cumbersome to trade. For example, futures on agricultural commodities allow an investor to bear
the risk associated with holding an agricultural commodity without the troublesome details of trading in the
3
For an extended discussion of the economic purposes of futures and options, see Carlton (1984), Jaffe (1984), Peck (1985),
Silber (1985), and Stoll and Whaley (1988).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 17/358
actual commodity. In this manner, processors of the underlying commodity can pass on price risk to
investors and retain the economic function of processing the underlying commodity.
Options provide an additional benefit in allocating risk because the profit function for options is
different from the profit function for the underlying commodity or for a futures contract. As is evident from
Figures 1.3 and 1.4, the profit from options positions are asymmetric. Such an asymmetric payoff pattern is
useful for example, in dealing with situations that involve both quantity and price risk. Consider a farmer
who is interested in avoiding the risk associated with a drop in the price of the commodity that he grows.
Before the harvest, the farmer does not know the size of the crop or the price. Selling futures against the crop
would hedge the farmer against a price decline if the harvest were known, but a futures hedge would expose
him to risk if the harvest failed and prices increased because then the farmer, would not have the wheat that
he had committed to sell in the futures market. The farmer would take a loss in acquiring the wheat to
delivery against the futures contract. Buying a put option on the underlying commodity provides a more
effective hedge against price and quantity risk than selling futures. If prices fall, the put is exercised (or
liquidated at a profit). If prices rise, the put option expires, worthless, and the farmer realizes the revenues
from his crop, regardless of the size of the harvest. The cost of this one-sided protection for the farmer is the
put option premium. Similar examples exist for other underlying commodities. In addition to this hedging
use, options are also a useful portfolio management tool. For example, index put options can be used to limit
the downside risk of stock portfolios while retaining part of the upside potential.
Price Information
Some trading in futures and options markets, as in other financial markets, arises not because individuals
have a desire to shift risk, but because they have different information and disagree about the correct price of
the underlying commodity. This kind of informational trading is termed speculative trading. Society benefits
from speculative trading because the analysis and search for information on which it is based cause the prices
of futures and options and underlying commodities to correspond more closely to their correct values. Even
if an underlying commodity is traded, futures and options on that commodity are likely to increase the
interest and the number of judgments bearing on the underlying commodity's price. Because futures and
options prices are related to the price of the underlying commodity by an arbitrage relation, factors affecting
futures and option prices tend to be conveyed to the price of the underlying commodity; conversely, factors
affecting the price of the underlying commodity tend to be conveyed to option and futures prices. Thus, to
the extent that futures and option trading increase the total interest in an underlying commodity, the
commodity's price will be more broadly based and less likely to be influenced by only a few judgments.
Futures and option markets encourage increased research and analysis. If financial markets are like
other markets, this leads to greater efficiency in the production of information. Insofar as increased analysis
and increased interest improves the quality of prices, resources will be allocated more efficiently. The price
signaling benefit of futures and options is most evident in the case of agricultural futures. For example, the
daily newspaper prints futures prices for wheat deliverable many months in the future. That price is the price
someone has paid to receive delivery of a particular grade of wheat in the future. Producers or storers of
wheat can use this price as a signal for production and storage decisions. If the price is high relative to their
costs, more wheat will be produced and stored. If the price is low relative to their costs, less wheat will
produced and stored. In this way, the proper amount of wheat will be allocated for future consumption. Of
course, if the futures price is wrong-because it is manipulated or for other reasons resources will be
misallocated. Futures markets depend on the presence of many knowledgeable participants to avoid this. One
of the benefits of the introduction of futures is the fact that futures trading increases competition. Producers
interested in entering into contracts for delayed delivery are no longer compelled to deal with relatively few
users of a commodity.
As we shall see later, option prices depend primarily on the projected volatility of the commodity
underlying the option. As a result, option prices quoted in the newspaper provide information on the future
price uncertainty of an underlying commodity. Processors and users of the underlying commodity therefore
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 18/358
have information not only on the expected price of the underlying commodity but also on the price
uncertainty in the future.
Transaction Costs
It is sometimes argued that futures and options are redundant securities because any futures or option
position is achievable by trading the underlying commodity. For example, a long position in wheat futures
has the same profit or loss potential as a long position in the wheat itself. A long position in stock index
futures can also be achieved by buying a diversified stock portfolio. The payoff to a Treasury bond futures
contract can be replicated by a position in the underlying T-bonds. Similarly, option positions can be
replicated by appropriate trading strategies in the underlying commodity.
An important benefit of futures and options, however, is that they reduce transaction costs of
achieving certain risk return positions compared with the cost of trading the underlying commodity. It is
certainly much less costly to trade wheat futures than to trade wheat itself. But the lower trading cost in
futures markets also exists with respect to financial instruments. For example, the transaction costs
associated with trading index futures contracts are estimated to be 1/15 of the costs associated with trading
the corresponding underlying stocks. Index futures are therefore a less expensive means of trading claims on
a portfolio of stocks.
1.4 SUMMARY
In this chapter, payoffs to long and short positions in futures and options are illustrated. The economic
purposes of futures and options-risk transfer, price discovery, and reduced transaction costs-are discussed.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 19/358
2.5 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 20/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 21/358
Maturity Months
Too many maturity months reduce the depth and liquidity in any one month. Too few maturity months
reduce the usefulness of a contract. How these conflicting objectives are balanced depends on the underlying
commodity. For example, in wheat, five maturity months (March, May, July, September, December) are
traded at any time, reflecting the harvesting and marketing cycles for wheat. In silver, sufficient volume
exists to warrant the trading of ten different maturity months extending one and a half years into the future.
In Eurodollar futures, forty maturity months extending ten years into the future are traded.
It is worth noting that the number of days until maturity of a particular futures contract is changing as
the maturity date is approached. This is in contrast to various forward contracts, the prices of which are
quoted in the newspaper. For example, the newspaper may quote the three-month forward price of silver or
the three-month forward price of a currency. These quotes are for new contracts originated on that day. The
secondary market for these contracts after they are originated is not very active, so price quotes of existing
contracts do not appear in the newspaper.
Contract Size
Contract sizes vary considerably and are chosen to meet the needs of the users of the contract. In many of the
grains, the standard contract size is 5,000 bushels, or approximately $15,000 at the current price of the
commodities. Contract sizes are considerably larger in some of the financial futures. For example, futures on
Treasury bills and Eurodollars have contract sizes of $1 million.
4
For more analysis of contract design and the success and failure of futures contracts, see Black (1986), Carlton (1984), Johnston
and McConnell (1989), and Stoll and Whaley (1985).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 22/358
exchange regulation. Delivery of tangible commodities may be made at any time during an extended period,
such as two weeks, and usually takes the form of warehouse receipts giving claim to the commodity, which
is stored at an approved location. Delivery of financial commodities is usually more narrowly defined, with
delivery taking place through an approved bank. The buyer of futures (the long) is obligated to take delivery
if called upon to do so. The assignment of delivery notices by the exchange takes various forms. In some
markets, the oldest long is assigned the delivery notice. In other markets, delivery notices are assigned
randomly. Futures contracts differ as to the flexibility remaining to the long after the receipt of a delivery
notice. In some cases, usually in the tangible commodities, the long has the opportunity to pass the notice on
to someone else and to liquidate the futures contract. The shorts usually have the greatest flexibility because
they can choose the particular grade of underlying commodity that will be delivered as well as the exact
timing of delivery.
With American style call options, the buyer may request delivery –exercise the options– at any time
during the option’s life. The seller of the call is then obligated to make delivery. With European-style call
options, delivery may be requested only on the expiration date. Options tend not to be written on underlying
tangible commodities that may be difficult and cumbersome to deliver. Instead, options on Commodities are
written on the futures contract on those commodities. Exercising futures-option on corn, for example, is a
request for the option seller to deliver the long futures position in corn. When the long futures position is
received, the position may be held to maturity if delivery of the corn is desired. The exercise of a put option
on corn is a request for the option seller to deliver a short futures position in corn. Options on many financial
instruments, such as options on individual common stocks, call for delivery of the underlying financial
instrument.
Certain futures and option contracts call for cash settlement rather than delivery at maturity. The
buyer of a cash settled futures contract, holding the position until expiration, receives the difference between
the final settlement price of the futures contract and the price at which the contract was purchased.5 The final
settlement price of the futures contract is the cash price of the underlying commodity. The seller of the
futures contract receives a profit exactly opposite that received by the buyer. In the case of cash settlement
call options, the exercise of a call option results not in the delivery of an underlying commodity, but rather in
a profit equal to the difference between the price of the underlying commodity and the exercise price of the
option.
Cash settlement is particularly useful when the underlying commodity is difficult to deliver. In U.S.
markets, stock index futures and options are cash settled because it is difficult to deliver a large portfolio of
many different common stocks. Municipal bond futures and futures on the consumer price index are also
cash settled, because, in these cases, the underlying commodity is impossible to deliver. In the case of
municipal bond futures, the price of the underlying commodity is actually an average of dealer quotations in
municipal bonds.
5
Technically, this statement applies to forward contracts only. The futures contract holder has accumulated over the life of the
contract an amount equal to the difference between the final settlement price and the price at which the futures was purchased. In
the interest of clarity, we defer detailed discussion of the distinction between forward and futures contracts to Chapter 3.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 23/358
soft red wheat and hard red winter wheat are deliverable against the wheat futures contract, but hard red
winter wheat is selling in the marketplace at a lower price than soft red wheat, the person with a short
position will choose to deliver hard red winter wheat because it is the cheapest to deliver. Indeed, the futures
price at maturity reflects the price of the cheapest to deliver grade of wheat, not necessarily the grade
specified as standard in the futures contract. In T-bond futures, it is sometimes desirable to deliver a low
coupon, long-maturity bond and at other times desirable to deliver a high-coupon, short-maturity bond. The
eight-percent coupon bond specified as standard in the contract is usually not the bond being priced by the
futures contract.
The cheapest to deliver commodity may change during the futures contract life. The exchange
specifies the price relation between the deliverable grades prior to the start of the contract. As market
conditions change, some grades go to a premium or discount relative to the standard grade.
Point of Delivery
An important feature of futures contracts on tangible commodities is the number and the location of delivery
points. Transportation of tangible commodities to the delivery location may be costly. As a result, an
increase in the number of delivery locations benefits the shorts who are obligated to deliver. To see this,
suppose a wheat futures contract calls for delivery only in approved warehouses in Chicago. If wheat is in
relatively short supply in Chicago, it is possible for someone to buy up most of the remaining supply while at
the same time buying wheat futures contracts. Such an individual would have engineered a corner if the
market did not have sufficient time to ship wheat to the Chicago delivery location. As a result, it is
sometimes desirable to specify several delivery locations in a contract, thereby making it difficult to corner
the available supplies at all the delivery points.6
Exercise Price
In the case of options, a feature is required that is not required in futures contracts namely, the number of
exercise prices which should be available. It would be possible, for example, to have only one exercise price
for each option maturity. This is not done because the usefulness of an option is greatest when the exercise
price is close to the price of the underlying commodity. As a result, additional options with new exercise
prices are created whenever the underlying commodity’s price moves by a prespecified amount. For
example, a stock selling at $40 might have an option with an exercise price of $40. If the stock price
increases to $45, a new option with an exercise price of $45 would be initiated. The $5 increment at which a
new option with a new exercise price is initiated is determined by the exchange.7 All the options on a
particular underlying commodity with the same maturity are called an option series. The number of options
in an option series is determined by the price volatility of the underlying commodity and the price increment
at which additional exercise prices will be set. If the price increment is small, many options, each with
relatively little liquidity, will be created. If the price increment is large, few options will be created.
6
A recent book by Pirrong, Haddock, and Kormendi (1991) provides a detailed analysis of delivery terms for agricultural
commodities.
7
Stock option exercise prices at $25 or above have $5 increments. Below $25, the increments are $2.50.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 24/358
Types of Orders
In trading futures and options, as with other securities, investors can place a variety of orders. A market order
instructs the broker to trade at the best price currently available. A limit order instructs the broker to buy or
sell at a specific price. Naturally, the price given for a limit order to buy is below the current market price,
and the price given for a limit order to sell is above the current market price. A stoploss order is an order to
sell below the market or to buy above the market. A spread order instructs the broker to buy one contract and
sell a related contract. In a maturity or calendar spread, for example, the trader buys a contract in one
maturity month and sells a contract in the same commodity for a different maturity month. In an
intercommodity spread, the trader purchases a contract in one commodity and sells a contract in a different
commodity. An almost unlimited array of spread transactions is available in futures and option markets, and
these will be discussed in later chapters of this book. An important point to remember is that the trader in a
spread transaction is interested in favorable changes in the price differential between two contracts.
Types of Markets
Orders placed by customers with their brokers are transmitted through the brokerage firms’ back offices to
the floor of the appropriate exchange for execution. The mechanics by which such orders are executed differ
between options on securities and futures. Futures and futures option contracts are traded in a pit in an “open
outcry” format. Generally, one pit or ring is assigned to each commodity traded on an exchange. Traders
stand on the steps around the pit and trade pairwise with each other. Certain actively traded futures contracts,
such as T-bond futures or S&P 500 index futures, attract in excess of 400 traders in the pit. Orders are
received on the trading floor by telephone and transmitted to the appropriate trading pit by messengers.
Unlike the NYSE stock market where trading in a particular stock occurs sequentially in time at a particular
location on the floor, many transactions can occur simultaneously in an active futures contract. Futures
markets do not therefore guarantee the same degree of price and time priority that NYSE stock markets
guarantee because of the possibility that two simultaneous transactions might occur in different parts of the
trading pit at different prices or because limit orders held by a particular broker may for some reason not be
exposed to all other brokers in the crowd. Such price differences within the ring are, however, infrequent and
small because many traders on each side of the market are each searching for the best price. Competition
among floor traders thereby reduces any price deviations within the pit and also provides tremendous
liquidity for orders flowing in from the public.
Options on financial instruments are traded according to two different procedures. Options on
securities exchanges such as the American Stock Exchange and the Philadelphia Stock Exchange are traded
using the specialist system. In a specialist system, market orders are usually traded at the bid or ask price
quoted by the specialist on his own behalf or on the behalf of limit orders previously left with the specialist,
although there is an opportunity for other traders in the crowd to better the specialist’s price. Limit orders are
left with the specialist to be executed when the market price reaches the limit price. (In futures markets, each
floor broker has his or her own “deck” of customer limit orders.) The specialist system has been criticized
because only a single specialist makes a market in each option. As a result, investors do not have an
opportunity to shop for better prices from other marketmakers.
The Chicago Board Options Exchange (CBOE) system combines elements of futures markets and the
specialist system of stock markets. An Order Book Official (OBO) maintains the book of limit orders but
8
The Chicago Board of Trade (1989) provides useful information on trading procedures and other aspects of futures markets.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 25/358
does not trade for his own account. At the same time, many professional floor traders are prepared to trade
for their own accounts to absorb temporary imbalances and maintain market liquidity. More so than either
the futures market system or the specialist system, the CBOE floor trading system represents what many
have called for in the stock market: a system that combines competition among marketmakers with full
exposure of all limit orders through the open book of the OBO.
Types of Traders
Floor traders in futures and option markets can be divided into two general classifications: floor brokers and
professional traders. Floor brokers are agents who execute transactions for public customers such as
processors of tangible commodities or portfolio managers of financial instruments. Professional traders buy
and sell for their own accounts. Professional traders are sometimes called speculators because they take on
varying amounts of risk. Professional traders in futures markets are often classified into position traders, day
traders, and scalpers. Position traders take on risks and positions that are held for longer periods of time –
days or weeks. Day traders have a short horizon and take on positions that are usually liquidated at the end of
the day. Scalpers have a very short horizon and make their income primarily from short-term, minute-by-
minute transactions. Scalpers provide liquidity to other investors by buying at the bid price when public
customers desire to sell and by selling at the ask price when public customers desire to buy.
In stock exchanges, the major type of professional trader is the specialist. The specialist’s role in the
stock market is similar to the scalper’s role in the futures market. Both are responsible for maintaining
market liquidity, and profit from the spread between bid and ask prices. In a specialist system, however,
there is only one specialist for each security. In futures markets, many scalpers compete in each contract. In
securities markets, there are relatively few professional traders other than the specialist. Most position traders
and day traders in options and stocks submit orders from off the exchange through brokerage firms. By
contrast, in futures markets, a great deal of volume is the result of trading by professional traders on the
floor.
Trading Costs
The costs of trading options and futures consist of two components: the commission charges of the broker
and the price concession that may be necessary to execute the transaction. The price concession reflects the
fact that sales are made at the bid price of professional traders on the floor (scalpers or the specialist) and
purchases are made at the higher ask price of professional traders on the floor. In addition, the broker
carrying out a customer transaction is compensated by a commission. Commissions on futures and option
exchanges are competitively determined and vary from broker to broker. Commissions cover the back-office
services of the broker as well as the charges for floor brokerage and the clearing of transactions.
Price Reporting
On stock exchanges, the price and size of each transaction is reported on the ticker tape immediately after the
transaction occurs. In futures markets, not every transaction is reported because many transactions occur
simultaneously. Instead, price reporters in the trading pit report each different transaction price and each
different bid or ask price in the pit. Systems for price reporting differ among the futures exchanges. Some
record prices manually on a price board above the exchange floor. The prices are then entered into computer
terminals for transmission worldwide. Others enter price information directly into computer terminals, and
then the information is automatically displayed on the exchange floor and is transmitted worldwide. In
futures markets, statistics on the volume of trading are not available on a real time basis. Instead, such
statistics are compiled at the end of the day on the basis of transactions clearing data.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 26/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 27/358
losses, if any, or are permitted to withdraw profits, if any, each day. These payments from losers to gainers
are called variation margin and must be in cash.
Options have some features of common stocks and some features of futures contracts. As in the case
of common stocks, a payment is made when the option contract is entered into. The buyer of an option pays
a premium for an insurance service rendered by the seller of an option. For example, the buyer of a call
option has the right to purchase the underlying asset at a known exercise price and is insured against any
losses should the underlying asset price fall below the exercise price. Although no asset changes hands when
an option contract is entered into, payment for the “insurance service” is made, and money changes hands.
Under current margin procedures for futures options as well as securities options, the buyer of the call or put
pays 100 percent of the premium. The seller of the option is required to post margin as a performance
guarantee, and the margin must be at least as great as the current market value of the seller’s obligation. In
this way, the clearinghouse is assured that the seller’s obligations will be carried out.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados Diciembre 2010 Página: 28/358
9
Margin deposits are required from both the buyer and the seller. We assume a constant margin deposit of $750 (5%).
10
Margin is required of the seller only. Margin requirements are complex and vary across contracts and exchanges. We assume a margin deposit of $1000 plus the current
value of the option
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 29/358
Futures trading is a zero-sum activity, as is reflected in the example by the fact that the sum of profits of the
four traders is zero each day. On balance, A and B gain; C and D lose. To the extent that margin is pledged
in cash, there is a net loss to traders in the form of foregone interest on the funds pledged as margins.
However, if the margin is pledged in the form of interest earning assets such as U.S. Government Securities,
this loss is avoided.
Table 2.2 presents a corresponding example of trading and settlement in call options on a common
stock. The underlying contract is for 100 shares of the stock which on day 1 is selling for $30. The option
premium is assumed to be 10 percent of the contract value. The structure of the example is similar to the
futures market example but there are some important differences. First, an option transaction requires the
payment of a premium by the option buyer to the option seller. Thus, on day 1, A pays $300 to B. Second,
only sellers of options are required to post margin. Buyers of options pay the full premium and meet all their
obligations at the time the premium is paid. Third, profits and losses arising from changes in the option price
are not settled daily. The buyer of an option can realize gains only by entering into a transaction. Such gains
are not transferred by a daily settlement procedure. Sellers of options do, in effect, realize gains and losses
daily because they are required to adjust their margin positions by the amount of the gain or loss in the
option. That is, their margin is marked-to-market daily. Thus, individual B must post an additional $10 of
margin on day 2 to guarantee the ability to purchase the underlying stock and deliver it.
2.5 SUMMARY
In this chapter, we first discussed the difference between over-the-counter (OTC) option and futures markets
and organized exchange markets. In OTC markets contracts can be tailored to the particular needs of
customers, but secondary market trading is difficult. In organized markets, contracts are standardized so that
secondary market trading is facilitated. Another key distinction between organized futures and option
exchanges and OTC markets is the existence of a clearinghouse. The clearinghouse is central to the operation
of organized futures markets. It interposes itself between buyer and seller and guarantees contracts. It sets
margins for clearing members and settles profits and losses daily. In contrast, in OTC markets, no
clearinghouse exists. The parties make their own arrangements for guaranteeing the contract’s financial
integrity.
Next, the factors that must be considered in designing futures and option contracts are discussed. The
design of contracts balances the benefits of a narrowly and clearly defined contract against the benefits of a
broadly defined contract that is less susceptible to manipulation and is more liquid.
Trading procedures are examined next. The basic types of orders that may be used are similar across
markets, but trading procedures differ quite markedly and range from the open outcry procedures of futures
markets to the specialist system used on certain option exchanges. Types of traders, trading costs, and price
reporting procedures are also discussed.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 30/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 31/358
4.7 RESUMEN
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 32/358
500
400
Ft
300
Prices
200
St
100
0 Sit
-100
-200
Time
Un protector con una posición en una mercadería quiere tomar una posición en el mercado de futuros que lo
proteja contra una disminución en el precio del tipo que se muestra en la Figura 4.1. La cobertura es efectiva
si el precio al contado de la mercadería base se mueve de la misma manera que los contratos de futuros de la
mercadería. Otra manera de decir lo mismo es que no hay cambios inesperados en la base, Ft St . La Tabla
4.1 contiene una ilustración numérica simple de una cobertura corta para alguien en el negocio de almacenar
trigo. En la Tabla 4.1, nosotros asumimos que el que almacena el trigo compra una tonelada de trigo el 1º de
setiembre y se protege vendiéndolo mediante un contrato de futuros para el 1º de diciembre. El 1º de
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 33/358
Tabla 4.1 Resultados de la utilidad de una cobertura corta, asumiendo una base constante por mes
Mercado en Efectivo (Spot) Futuros a Diciembre
Fecha Transacción Precio Transacción Precio
1º de setiembre Comprar 1 tonelada a 3.00 Vender futuros a 3.09
1º de noviembre Vender 1 tonelada a 2.70 Vender futuros a 2.73
Ganancia -0.30 0.36
Ganancia neta 0.06
Ganancia neta menos costos de almacenaje de $0.03 por mes 0.00
Después de dos meses, el 1º de noviembre, el precio del trigo cayó drásticamente y el trigo se vendió
a $2.70, causando una pérdida de $0.30 en el mercado financiero. Los precios de los futuros, sin embargo,
también cayeron a $2.73, resultando en una ganancia de $0.36. La diferencia entre los precios de los futuros
y el precio en efectivo se ha acortado de $0.09 a $0.03, pero la base por mes sigue en $0.03 ya que sólo falta
un mes para el vencimiento del contrato de futuros. Dado que el protector estaba corto en el mercado de
futuros, la ganancia allí de $0.36 compensa de más la pérdida de $0.30 en el mercado en efectivo o al
contado. Esta ganancia refleja el hecho que la base total, Ft St , se acorta conforme se va acercando el
vencimiento para reflejar el tiempo reducido a lo largo del cual se debe mantener la mercadería. Sin
embargo, el protector mantuvo la mercadería básica durante dos meses e incurrió en gastos de almacenaje,
los cuales hemos asumido son $0.06 por tonelada. Como resultado la ganancia neta de $0.06 por los cambios
en los precios de los contratos futuros y los precios al contado se compensan con el costo de $0.06 de
mantener la mercadería. La Tabla 4.1 ilustra el caso en el cual la cobertura es completamente efectiva en el
sentido que el protector no sufre pérdidas ni obtiene ganancias.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 34/358
contrato de futuros. En el caso de mercaderías disponibles, la convergencia de los precios de los futuros y al
contado garantizan una ganancia de $0.09, lo cual es igual a la base cuando se estableció la cobertura.
Si los costos de almacenamiento para el mes restante se mantienen en $0.03 por tonelada, se obtendrá
una utilidad en exceso de $0.02 en el último mes, la cual es suficiente para compensar la pérdida de $0.02 en
los primeros dos meses. Por otro lado, si los costos de almacenamiento se elevan a $0.05, como se implica
por la base el 1º de noviembre, la pérdida el 1º de noviembre es inevitable. La pérdida no se origina por la
falta de efectividad de la cobertura, sino por la falla de la cobertura de no asegurar los costos de
almacenamiento durante todo el período de tres meses.
Tabla 4.2 Resultado de las utilidades de una cobertura corta con riesgo de base
Mercado al Contado Futuros a Diciembre
Fecha Transacción Precio Transacción Precios Alternativos
1º setiembre Comprar 1 tonelada a 3.00 Vender futuros a 3.09 3.09
1º noviembre Vender 1 tonelada a 2.70 Comprar futuros a 2.75 2.71
Ganancia 0.34 0.38
Ganancia neta 0.04 0.08
Ganancia neta menos costos de almacenaje de $0.03 por mes -0.02 0.02
Esta es sólo otra manera de decir que el tenedor del inventario enfrenta dos fuentes de riesgo: (a)
fluctuaciones en el precio de la mercadería y (b) fluctuaciones en el costo de mantener la mercadería. Los
mercados de futuros pueden ser perfectamente efectivos en la cobertura del riesgo de precio, pero no son
efectivos para proteger el segundo riesgo.
El segundo caso ilustrado en la Tabla 4.2 es una disminución en el precio de los futuros a $2.71, lo
cual implica un estrechamiento o reforzamiento de la base. En este caso, existe una ganancia neta de $0.08,
que cubre de más los costos de almacenaje de $0.06 en los primeros dos meses. En este ejemplo la base ha
cambiado en una dirección favorable. Siempre que el protector no haya contratado un almacén en el último
mes, esta ganancia es real. Sin embargo, si el protector ha pedido prestados fondos y ha contratado un silo de
granos para todo el período de tres meses, se incurre en costo de almacenaje en el tercer mes, ya sea que se
almacene el trigo o no. En ese caso la ganancia no es real. Si se va a incurrir de todas maneras en el costo de
almacenaje de $0.03, es mejor mantener el trigo un mes más para obtener la base de $0.01, la cual compensa
en parte el costo de almacenamiento de $0.03.
El ejemplo anterior y las implicancias enumeradas a continuación se basan en que el protector tiene
una posición en la mercadería base del contrato de futuros. La cobertura de mercaderías no-entregables se
discute en la siguiente sección. A continuación veremos un resumen de las implicancias de los ejemplos en
esta sección (esto es, ejemplos donde el protector mantiene una mercadería disponible y protege utilizando
los futuros) bajo dos encabezados:
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 35/358
$0.00. No existe riesgo en este ejemplo (irreal) porque los precios son pre-establecidos el 1º de
setiembre.
c. Puede no ser posible fijar los costos de acarreo de una mercadería por adelantado. Como resultado, los
tenedores de una mercadería están sujetos a cierta cantidad de riesgo base que se origina por costos de
acarreo inciertos.
La única fuente de riesgo de base cuando una mercadería se puede entregar contra un contrato de futuros es
el riesgo del costo de acarreo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 36/358
Notación
La notación utilizada en el desarrollo del ratio de cobertura óptimo es la siguiente:
S0 precio al contado inicial. (Esta variable corresponde al primer ingreso en la columna de mercado al
contado de la Tabla 4.1; esto es, el valor $3.00.) Estamos omitiendo el subíndice i indicando el precio al
contado.
F0 precio inicial de futuros. (Este valor corresponde al primer ingreso en la columna del mercado de
futuros de la Tabla 4.1, eso es, el valor $3.09.)
ST precio al contado incierto al momento futuro T . (Esta variable corresponde al valor del 1º de
noviembre de la columna al contado de la Tabla 4.1, el cual es uno de los muchos resultados posibles en ese
momento.)
FT precios de futuros en T . (Esta variable corresponde al valor el 1º de noviembre en la columna de
futuros de la Tabla 4.1, el cual es uno de los muchos resultados posibles en ese momento.)
nS número de unidades de mercadería al contado en cartera. ( nS es positivo para las posiciones largas y es
negativo para las posiciones cortas.)
nF número de contratos de futuros en cartera. ( nF es positivo para las posiciones largas y es negativo para
las posiciones cortas.)
~ ~
~h ( ST S 0 )nS ( FT F0 )nF (4.1)
~
El término ( ST S 0 ) es el cambio de precio (aleatorio) por unidad de la mercadería a lo largo de la vida de la
~
cobertura, y el término ( FT F0 ) es el cambio de precio (aleatorio) del contrato de futuros a lo largo de la
vida de la cobertura. La ganancia esperada para el protector a lo largo de la vida de la cobertura es
~ ~
E (~h ) [ E ( ST ) S 0 ]nS [ E ( FT ) F0 ]nF (4.2)
La expresión (4.2) proporciona una manera conveniente para ilustrar dos ideas. Primero, si asumimos que el
precio actual de los futuros es un mecanismo de predicción del precio de los futuros en el momento T como
se implica en la Figura 4.1, esto es, si
~
F0 E ( FT ) (4.3)
el último término de la ecuación (4.2) se elimina. La utilidad esperada sobre la cartera de cobertura depende
solamente del cambio esperado en el precio al contado. En un mercado de costos de acarreo, la diferencia
entre el precio al contado actual y el precio al contado futuro esperado representa los costos de almacenaje.
Segundo, si la mercadería está disponible y se mantiene hasta el vencimiento, la convergencia asegura que
~ ~
E ( ST ) E ( FT ) (4.4)
Combinando (4.2), (4.3) y (4.4) obtenemos
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 37/358
En otras palabras, la utilidad esperada sobre la cobertura es directamente proporcional a la base inicial como
se indicó en la Tabla 4.1. El resultado general dependerá de los costos de acarreo incurridos por el protector
que no son incorporados en la ecuación (4.1). Incluso si la mercadería es no-entregable, la ganancia esperada
es igual a la base, siempre que la diferencia entre el precio de entrega inmediata y el precio al contado en la
Figura 3.1 permanezca constante.
Los precios futuros realizados y los precios al contado se desvían de los precios esperados; por lo
tanto, las ganancias realizadas son mayores o menores que las ganancias esperadas. Esto es cierto, en
particular, para la cobertura cruzada en cuyo caso no hay oportunidad para entregar la mercadería. La
cobertura cruzada óptima es la cobertura que minimiza la desviación a partir de la ganancia esperada. Ahora
veremos una derivación de la cantidad protegida óptima basada en este criterio.
~h ~ ~ n
( ST S 0 ) ( FT F0 ) F . (4.6)
nS nS
~ ~ ~ ~
Ahora definamos los términos del cambio de precio, S ( ST S 0 ) y F ( FT F0 ) , y sustituimos estos
términos en (4.6),
~h ~ ~
S h F . (4.7)
nS
Como el protector está preocupado por minimizar el riesgo11, utilicen (4.6) para escribir la variación de la
utilidad de la cartera de cobertura.
h2 S2 h 2 F2 2h SF , (4.8)
SF
h* . (4.9)
F2
11
El marco referencial del ratio de cobertura media-variación desarrollado aquí es similar al desarrollado en Ederington (1979). En
este marco referencial, se asume que el protector sólo está interesado en minimizar su exposición al riesgo. En un marco
referencial más general, se permitiría al protector considerar no sólo la variación en el precio sino también el cambio de precio
esperado para la determinación del ratio de cobertura óptimo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 38/358
Por lo tanto, el ratio de cobertura óptimo, h* , depende de la covariación entre el cambio en el precio al
contado y el precio de los futuros con relación a la variación del cambio en el precio de los futuros. Es
interesante notar que la expresión para el ratio de cobertura óptimo, SF / F2 , es la pendiente de la regresión
~
normal de mínimos cuadrados (OLS) del cambio del precio al contado, S , sobre el cambio en el precio de
~
los futuros, F . El enfoque de la regresión de los mínimos cuadrados para calcular el ratio de cobertura
óptimo se describe a continuación.
donde 0 y 1 son parámetros de la regresión, ~ es un término aleatorio con E (~ ) 0 . La ecuación (4.10),
la cual es trazada en la Figura 4.2, muestra que el cambio en el precio al contado tiene tres componentes: (a)
un componente constante no-aleatorio de cambio en el precio 0 , la intersección en la figura; (b) un
~
componente aleatorio que es relacionado sistemáticamente con el cambio en el precio de los futuros, 1 F , y
(c) un componente aleatorio singular, ~ , que no está correlacionado con el cambio del precio de los futuros
y es representado por las distancias verticales entre la línea y los puntos fuera de la línea en la Figura 4.2.
Retorno Esperado
El término de intersección, 0 , captura cualquier cambio esperado en el precio al contado acompañado por
~
un cambio esperado en el precio de los futuros. De la Figura 3.1, sabemos que, E ( F ) 0 , el cambio del
~ ~
precio al contado esperado es igual a la base, esto es, E ( S ) E ( ST S 0 ) F0 S 0 . El modelo de regresión
~
establece que el cambio en el precio al contado esperado es igual a 0 bajo la premisa que E ( F ) 0 (y que
E (~ ) 0 ), por lo tanto el término de intersección en (4.10) representa la base. La base, a su vez, refleja los
costos de almacenaje que el abastecedor de los activos debe recuperar mediante el aumento del precio. El
~
término, 1 F , refleja el hecho que los cambios aleatorios en el precio de los futuros se reflejará en el precio
al contado de acuerdo con el coeficiente de inclinación, 1 . El término ~ refleja el riesgo de base, el cual se
~
origina del hecho que ciertos cambios aleatorios en S , , son únicos para la mercadería al contado y no están
correlacionados con el cambio del precio de los futuros.
12
En Pyndyck y Rubinfeld (1981, Capítulos 1-6) se proporciona una excelente explicación de la regresión de mínimos cuadrados.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 39/358
Figura 4.2 Cambio del Precio al Contado Versus Cambio del Precio de los Futuros
Cambio Precio
Spot, S
1
1
Cambio Precio de
Futuros, F
~h ~ ~ ~
0 1 F h F 0 (1 h) F (4.11)
nS
~
La Ecuación (4.11) muestra claramente que la utilidad de la cartera de cobertura, F , puede ser
independiente de los movimientos de los precios al contado y futuros al establecer h 1 . Si 1 1 , un
cambio de un dólar en el precio al contado es equiparado con un cambio de un dólar en los precios de los
futuros. Esto implica que los precios al contado y de los futuros se mueven a la par. En este caso, la
cobertura óptima es h 1 , o una cobertura de 100 porciento. Si 1 0.0 , los precios de los futuros y el
precio al contado no están relacionados y no tiene sentido proteger y el ratio de cobertura óptimo es cero
(h 0) .
Efectividad de la Cobertura
Una cobertura es completamente efectiva sólo si los cambios en el precio de los futuros y el precio al
contado están perfectamente correlacionados. Esto significa que el término de error aleatorio, ~ , en (4.19)
siempre es cero. Aunque podría ser óptimo proteger 100 porciento, la cobertura puede no ser totalmente
efectiva porque pueden originarse desviaciones de la relación promedio, representada por los puntos fuera de
la línea en la Figura 4.1.
Para poder medir la efectividad de una cobertura, primero se mide la falta de efectividad de una
cobertura. La falta de efectividad de una cobertura es medida por el ratio de variación del cambio en el
precio de la cartera protegida ( h2 ) sobre la variación del cambio del precio de la cartera sin cobertura
( h2 cuando h 0 ). De la ecuación (4.8), el cambio del precio de la cartera sin cobertura sobre la variación
del cambio del precio sin cobertura es S2 . De (4.11), la variación del cambio del precio de la cartera
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 40/358
protegida óptimamente ( h 1 ) es 2 , esto es, la variación del término residual, ~ , es el modelo de
regresión. El ratio que mide la falta de efectividad de la cobertura es por lo tanto 2 / S2 ; este ratio puede
variar entre 0 y 1. Dado que la efectividad de la cobertura es simplemente el complemento de la falta de
efectividad, la efectividad de la cobertura se puede medir como
2
1 . (4.12)
S2
Para protegerse contra múltiples fuentes de riesgo de precio, se realiza una regresión del cambio en el
precio al contado sobre los cambios de precios de varios contratos futuros en la forma,
~ ~ ~ ~
S 0 1 F ,1 1 F , 2 n F ,n ~, (4.13)
Los coeficientes de regresión estimados 1 a n , son los respectivos ratios de cobertura para cada uno de los
n contratos de futuros. Para protegerse contra todas las fuentes de riesgo, establezcan hi ̂ i para todos
los contratos futuros. Para proteger solamente exposiciones de riesgo seleccionadas, establezcan hi ̂ i
para los contratos de futuros correspondientes a los riesgos seleccionados.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 41/358
da cierta guía sobre este tema. El término de error, ~t , en (4.14) es gobernado por las siguientes premisas:
E ( ) 0, E ( ), E ( 2 ) 2 y E ( , ) 0 . El intervalo de cambio en el precio debe ser elegido de
t t F ,t t t t 1
13
Para 1989,se utilizan los precios de los contratos de futuros del índice S&P 500 de marzo de 1989, junio de 1989, setiembre de
1989, diciembre de 1989 y marzo de 1990 para generar la serie de cambios de precios de los futuros. El contrato de marzo de 1989
es utilizado hasta el último día de negociación el 16 de marzo de 1989. En ese día, se registran tanto los precios los contratos de
marzo de 1989 como de junio de 1989. El precio de los contratos de marzo se utiliza en combinación con su precio el día anterior
para calcular el cambio de precio de los futuros del 16 de marzo. El precio de los contratos de junio es utilizado en combinación
con su precio el día siguiente para calcular el cambio del precio de los futuros para el 17 de marzo. Este procedimiento de
“empalme” permite que la serie de cambios del precio de los futuros sea continua a lo largo del año.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 42/358
Tabla 4.3. Resumen de los cálculos del coeficiente de ratio de cobertura utilizando los cambios14
diarios, semanales y quincenales del precio del índice S&P 500 y los futuros del índice S&P 500
durante el año calendario 1989
S ,t 0 1 F ,t t . Intervalo de Confianza de 95 Porciento.
Intervalo n ̂1 S (ˆ1 ) 15 Inferior Superior Rango R2
Diario 251 0.8034 0.0226 0.7589 0.8478 0.0889 0.8352
Semanal 51 0.9914 0.0163 0.9586 1.0241 0.0673 0.9867
Quincenal 25 1.0013 0.0170 0.9662 1.0365 0.0703 0.9928
Un número de resultados interesantes surgen de la Tabla 4.3. El primero es que pendiente calculada
utilizando cambios de precios diarios, 0.8034, es drásticamente diferente de aquellos utilizados para los
cambios de precios semanales y quincenales, 0.9914 y 1.0013, respectivamente. Esta diferencia resulta de los
problemas arriba mencionados. Por ejemplo, el efecto de precio oferta/demanda introduce el problema de
errores en las variables en la regresión de precios diarios, la cual tiende a sesgar los coeficientes hacia abajo.
Otra comparación interesante en la Tabla 4.3 es la de los intervalos de confianza. Manteniendo otros
factores constantes, el rango del intervalo de confianza (el error estándar) en 1 , debería aumentar conforme
el intervalo de cambio de precio se incrementa de diario a semanal y quincenal porque más y más
información se está perdiendo. Sin embargo, ese no es el patrón que aparece en la Tabla 4.3. El error
estándar y el rango del intervalo de confianza son menores para las regresiones de cambio de precios
semanales y quincenales que en la regresión de cambio de precio diaria. Una vez más, este resultado refleja
los problemas asociados con el uso de cambios de precios diarios.
La comparación de los resultados semanales y quincenales también favorece el uso de los datos
semanales en el procedimiento de cálculo. Las magnitudes del coeficiente están muy cercanas, y aún así el
error estándar y el rango del intervalo de confianza son menores para la regresión semanal. El error estándar
inferior refleja el hecho que en la regresión semanal y la quincenal se introduce casi el doble de información
sobre cambios de precios.
Relacionado a la selección del intervalo apropiado de cambio de precio se encuentra el horizonte de
tiempo de la cobertura. El marco referencial de la cobertura desarrollado en este capítulo es para un solo
período. En principio, cuando el ratio de cobertura es calculado mediante el análisis de regresión, el cálculo
del ratio de cobertura es independiente de la distancia en las observaciones del precio en la regresión. Sin
embrago, el término de intersección aumentará conforme se incremente el tiempo entre las observaciones de
los precios porque es un estimado de la base entre los precios de futuros y al contado durante la duración del
intervalo de observación.
Finalmente, para terminar la discusión, también es importante notar que el parámetro estimado 1 , se
produce de una regresión de datos históricos y que la cobertura que estamos elaborando es para un período
futuro (es decir, los estimados de a son ex-post, pero las decisiones de cobertura ex ante). Al aplicar este
procedimiento, nosotros estamos invocando implícitamente una premisa de inmovilidad en la relación entre
los cambios de precios al contado y de los futuros. A priori, debemos está cómodos con que esa premisa sea
razonable.
14
Los cambios de precios del índice S&P 500 y los contratos de futuros del índice S&P 500 se calculan como S ,t ( St St 1 ) 6+, y
F ,t ( Ft Ft 1 ) , respectivamente.
15
s () es el error estándar del coeficiente de regresión.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 43/358
0.9586 1 1.0241 .
Este rango implica que tenemos 95 porciento de confianza que el número óptimo de contratos de futuros a
venderse esté entre 271.26 y 289.79.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 44/358
Tesoro tiene una denominación de $100,000, de manera que el número de unidades de la mercadería al
contado en términos de futuros de bonos del Tesoro es 100. El nivel del índice S&P 500 era 353.40 a fines de
1989 y la denominación del contrato de futuros S&P 500 es 500 veces el índice, de manera que el número de
unidades de la mercadería al contado en términos de los futuros S&P 500 es 56.59. Utilizando los estimados
de coeficiente reportados en la Tabla 4.4, el número óptimo de futuros de bonos del Tesoro a venderse es 100
x 0.4473 = 44.73, y el número óptimo de futuros S&P 500 a venderse es 56.59 x 0.0159 = 0.90.
Tabla 4.4. Resumen de los estimados del coeficiente del ratio de cobertura utilizando cambios16 de
precios semanales para bonos de Mobil Oil al 8½ porciento con vencimiento en el año 2001 (MO),
contrato de futuros del bono del Tesoro CBT (TBF), y contrato de futuros S&P 500 de CME (SPF)
~ ~ ~
durante el año calendario 1989: MO ,t 0 1 T B ,F ,t 2 SP ,F ,t ~t ,
n ̂1 s (ˆ1 ) 17 t (ˆ1 ) 18 ̂ 2 s (ˆ 2 ) t (ˆ 2 ) R2
Futuros de Bono del Tesoro y S&P 500:
51 0.4473 0.0407 10.98 0.0159 0.0091 1.75 0.7651
Solo Futuros de Bono del Tesoro:
51 0.4756 0.0382 12.46 0.7553
Solo Futuros de S&P 500:
51 0.0554 0.0154 3.59 0.1920
4.7 RESUMEN
este capítulo comienza con una explicación de la cobertura corta tradicional en la cual el propietario de una
mercadería asume una posición corta en futuros para asegurarse contra una disminución en el precio de la
mercadería. Un protector en corto asegura la base, Ft S t . Los protectores en corto enfrentan el riesgo que la
base pueda cambiar a lo largo del tiempo. En el caso de los protectores de mercaderías entregables o
disponibles, el riesgo de base representa el riesgo de un incremento en los costos de mantener la mercadería.
En el caso de los protectores de mercaderías no disponibles contra un contrato de futuros (cobertura
cruzada), el riesgo de base también representa el riesgo de cambio de los precios relativos de la mercadería y
el contrato de futuros.
En este capítulo es desarrollado un marco referencial para la cobertura según el análisis de cartera
media/variación y también dentro de un marco referencial de regresión de mínimos cuadrados. Se define el
ratio de cobertura óptimo y se discute la medición de la efectividad de la cobertura. El marco referencial de
cobertura óptima es entonces aplicado en el contexto del manejo del riesgo de una cartera de acciones y de
una cartera de bonos, dando cuidadosa consideración a la naturaleza de los datos del precio del título valor y
el cálculo de la regresión.
16
Los cambios en los precios son calculados como i ,t I t I t 1 donde I representa el bono de Mobil Oil, el contrato de futuros del
bono del Tesoro o el contrato de futuros S&P 500.
17
s() es el error estándar del coeficiente de regresión.
18
t () es el valor- t del coeficiente de regresión bajo la hipótesis nula que 0 .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 45/358
5.2 IS THE FUTURES PRICE AN UNBIASED ESTIMATE OF THE EXPECTED SPOT PRICE?
5.5 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 46/358
Hedgers enter the futures market to reduce or eliminate the risk of a commodity position. Conceivably, the
futures market could consist solely of hedgers. For example, farmers may want to sell wheat futures, while
processors of wheat may want to buy wheat futures. Both these parties are hedgers and could have futures
transactions that exactly offset each other. But usually the transactions of the long and short hedgers are not
exactly offsetting. Liquid and active futures markets typically require the participation of speculators.
Speculators analyze information concerning futures contracts and their underlying commodities in the hope
of identifying, and profiting from, futures contract mispricings.
5.2 IS THE FUTURES PRICE AN UNBIASED ESTIMATE OF THE EXPECTED SPOT PRICE?
~
The expected spot price, Et ( ST ) , is the market’s expectation at time t of the spot price when the futures
contract expires (time T ). At any given time t prior to contract maturity, the expectation is likely to be
wrong. Unexpected events after time t cause the actual spot price to be different from what was expected.
Over a long period of time and many futures contract cycles, however, expectations should, on average, be
~
realized, that is, Et ( ST ) avg ST , in other words, the average spot price at maturity ought to reflect the
expectation at some prior time.
In order for speculators to make money, futures prices must trend upward toward the expected spot
price when speculators are long futures, and futures prices must trend downward when speculators are short
~
futures. A market in which futures prices trend upward [i.e., where Ft Et ( ST ) ] is said to exhibit normal
backwardation. This situation is illustrated as the upward sloping line in Figure 5.1. A market in which
~
futures prices trend downward [i.e., where Ft Et ( ST ) ] is said to exhibit contango. This situation is depicted
in Figure 5.2 as the downward sloping line. If there is no trend in the futures price, the futures price is an
~
unbiased estimate of the expected spot price, that is, Ft Et ( ST ) .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 47/358
Price ~
Et (ST )
Convergence
Ft
St
Time
Price
Ft
Convergence
~
Et (S T )
St
Time
Students of futures markets have long discussed whether or not speculators as a group make money. Some,
like Keynes (1930), Hicks (1939), and Cootner (1960a, 1960b) argue that speculators make money because
they bear risk and must be compensated for their risk-bearing services. They usually argue that speculators I
0 tend to be long futures because hedgers tend to be short.19 Sales by hedgers force the futures price below
the expected spot price and lead to the situation of normal backwardation shown in Figure 5.1. Speculators
make money on the upward trend in futures prices.
Others, particularly Telser (1958, 1960), argue that speculators as a group are not risk-averse and do
not require compensation for risk. This is possible if there are different categories of speculators.
Professional speculators have to make money. Otherwise, they would be unable to support themselves. But
19
That is, on balance, there are more short hedgers than long hedgers.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 48/358
amateur speculators could lose money to professional speculators, so speculators as a group just break even.
Amateur speculators consist of two categories-gamblers and fools.
Gamblers enjoy the risks of small futures positions. They know the risks and the fact that there is a
house-take, but they enjoy the game. Fools believe they have a successful strategy. They think they know
how to make money in futures, but do not. Fools tend to lose money and then withdraw from the market. The
supply of fools is replenished by Barum’s Law. (There’s a sucker born every minute!) Telser thus argues that
speculators as a group do not make money even though they bear risk. If Telser’s argument is true, hedgers
are better off because they are provided insurance at no cost.
Finally, some argue that the amount of risk actually borne by speculators is small, if risk is properly
measured. Dusak (1973) takes this position. In modem finance theory, the appropriate measure of risk is the
amount of risk that cannot be diversified away. In other words, risk is measured in a portfolio context. Dusak
argues that commodity risk can be diversified away so that the systematic risk of a commodity is zero. That
means that speculators do not require a risk premium. Competition among speculators for futures contracts
will then drive the futures price to a point such that the futures price equals the expected spot price. To the
extent that the systematic risk of futures contracts is negative, speculators might be willing to accept losses.
For example, suppose futures were a good inflation hedge. In such a situation, speculators would be willing
to lose money in futures as a way to reduce the risk in other parts of their portfolio.
~
where r is the riskless rate of interest, RM is the rate of return on the market portfolio, and is the asset’s
beta or relative systematic risk coefficient. Equation (5.1) says that the expected rate of return on a risky
~
asset, E ( R ) , equals the riskless rate of return, r , plus a market risk premium equal to the market price of
~
risk, [ E ( RM ) r ] , times the asset’s relative systematic risk level, . Assume that the entire asset return is
~ ~
derived from price appreciation (depreciation), that is, R (VT V0 ) / V0 , where V is the asset price. Then
equation (5.1) may be rearranged to derive an expression for the current asset price:
To see how the capital asset pricing model applies to futures contracts, recognize that the terminal value of a
long futures position is
~ ~
VT FT F0 , (5.3)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 49/358
~
where F0 is the futures price when the contract is negotiated and FT , is the uncertain futures price when the
contract expires.20 Substituting (5.3) into (5.2) yields
However, the value of a futures position when the contract is first entered into is zero (i.e., the futures
position involves no initial investment outlay). Setting V0 0 in (5.4) and rearranging yields
~ ~ ~
where RF , is the futures return, that is, RF ( FT F0 ) / F0 . Simplifying, equation (5.6) can be written as
~ ~
E ( RF ) [ E ( RM ) r ] F , (5.7)
Figure 5.3 contrasts the expected return-risk relation for assets with the expected return/risk relation for
futures contracts. The two lines depicted in Figure 5.3 have the same slope, however the line corresponding
to assets has an intercept equal to the riskless rate of interest while the line corresponding to futures contracts
goes through the origin. Since asset positions require an investment outlay, an expected rate of return of at
least the riskless rate is earned on the asset position. Since futures positions require no investment outlay, the
minimum expected return for a futures contract is zero. In equilibrium, the expected rate of return on futures
equals only a market risk premium equal to the market price of risk times the futures contract beta.
In the preceding section, we discussed three alternative views of the expected return to holding futures
contracts. Under the Keynes/Hicks/Cootner view, speculators earn a positive risk premium. That view is
reflected in Figure 5.3 as a position along the futures pricing line with a positive level of systematic risk. The
positive return is earned from the upward drift in futures prices when speculators are long. An alternative
view is that of Dusak, who argues that futures contracts have no systematic risk. In other words, the beta of a
futures contract is zero. In that case, futures contracts are also on the futures contract pricing line but at the
point where the line crosses the origin and where beta equals zero. Under the Dusak view, there is no upward
drift in futures prices because no compensation for risk is necessary. Under the third view, that of Telser,
speculators in the aggregate earn no risk premium, even though risk exists. The Telser position in Figure 5.3
is an average of the position of professionals and amateurs in the figure. Professionals earn a normal risk
20
In (5.3), the daily settlement feature of futures is ignored. The interest earned or paid on daily settlements ought to be a part of
the profit or loss on the futures position. Usually the interest amount is very small. In addition, we have shown in Chapter 3 that, if
interest rates are known and transaction costs are zero, futures market positions can be adjusted so that expression (5.3) is an
appropriate measure of profit on a futures contract.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 50/358
premium and are on the futures pricing line. However, gamblers and fools earn negative returns as shown in
the figure. The average of these two points is the Telser position-positive risk but no return.
FIGURE 5.2 The Capital Asset Pricing Model and Alternative Views of the
Returns to Speculators
~
Et ( RF )
Assets
Futures
r Keynes/Hicks/Cootner
Telser Professional
Beta
Dusak Telser
Telser Amateurs
The empirical evidence on risk and return in futures markets is ambiguous and makes it difficult to
distinguish among these three alternative views. Telser and Cootner debated vehemently in the 1960’s as to
the meaning of the data for corn, wheat, and cotton. Cootner maintained that an upward drift in futures prices
was observable, while Telser argued it was not.
In a comprehensive investigation using semi-monthly price data for corn, wheat, and soybean futures
during the period May 1952 through November 1967, Dusak (1973) concludes that the expected futures
returns equal zero. To support her conclusions, she estimates (a) the mean realized futures return and (b) the
systematic risk coefficient for each of the futures contract months of the three underlying commodities. The
systematic risk coefficients are estimated using the OLS regression,
~ ~
RF ,t F F ( RM ,t rt ) t , (5.8)
~
where the proxy for the market return, RM ,t , is the price appreciation on the S&P 500 stock portfolio, and the
proxy for the riskless rate, rt , is the return on a T-bill with fifteen days to maturity. The results are reported
in Table 5.1. In Table 5.1, note first that in only two of the sixteen cases reported is the mean realized return
significantly different from zero (i.e., twice its standard error), and in both of those cases the realized return
is negative. These results are further corroborated by the estimates of the systematic risk coefficients. In only
one of the sixteen cases is the beta of the futures contract significantly different from zero. The lack of
covariation of the futures returns with the market return is also reflected through the low R 2 values reported
in the table.
All in all, the Dusak results appear to support the position that futures prices are unbiased predictors
of expected spot prices, however, the generality of the results is not known. Other investigators have found
different results as far as agricultural and metal futures markets are concerned.21 A broader range of
21
Bodie and Rosansky (1980), for example, analyze futures prices for the period 1949 and 1976 and conclude that futures
contracts on average have a positive return. Unfortunately, the Bodie/Rosansky results are suspect given that they use annual data
(and, hence, very few time-series observations) in their statistical analyses.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 51/358
underlying commodities and longer time series of daily or weekly prices are among the experimental
improvements necessary to determine which of the competing theories is best supported.
correlation between the futures price and the price of the underlying commodity but also the relation between
the futures and the hedger’s entire portfolio of all assets.23
Figure 5.4 shows how hedgers and speculators interact to determine a futures price in relation to the
expected spot price and the current spot price. The figure assumes homogeneous expectations on the part of
hedgers and speculators. Hedgers are distinguished from speculators because they have a position in the
underlying commodity. The HH schedule in Figure 5.4 depicts the futures market position that hedgers as a
group would like to hold for alternative futures prices. Note that the HH schedule crosses the vertical axis
below the expected spot price. That means that hedgers would sell futures at prices below the expected spot
price because of the attractiveness of risk transfer. The position of the HH schedule depends on the nature
and size of the underlying commitment. The slope of the schedule depends on the amount of price risk of the
underlying commodity and the degree of risk aversion of hedgers.
22
s () is the estimated standard error of the regression coefficient.
23
See Stoll (1979) for a more detailed discussion of this point.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 52/358
~
H E0 ( S T )
S
F0*
Short H
Futures
Position
Long
Speculators Short Speculators Long
Position Position
The SS schedule depicts the futures market positions that speculators would accept for alternative futures
prices. The SS schedule is downward sloping and intersects the vertical axis at the expected spot price. When
the futures price equals the expected spot price, the speculator has no incentive to take a futures position
either long or short. When the futures price falls below the expected spot price, speculators earn a risk
premium by taking a long position; and when the futures price is above the expected spot price, speculators
earn a risk premium by taking a short position. The downward sloping SS schedule implies that a larger risk
premium is required to induce speculators to take a larger position.24
In Figure 5.4, the equilibrium futures price, F0* , is determined such that the short position taken by hedgers
~
equals the long position taken by speculators. Speculators expect to receive a risk premium of E0 ( ST ) F0* ,
and hedgers expect to pay that risk premium. Hedgers hold real assets (like wheat or common stocks) and
sell futures to avoid risk. Speculators accept the risk; and, in return, earn a risk premium. Figure 5.4 is
consistent with the Keynes/Hicks/Cootner view and a capital asset pricing model in which the underlying
commodity has systematic risk.
Under the Telser and Dusak views of speculators, the SS schedule would be perfectly horizontal and would
~
cross the vertical axis at E0 ( ST ) . In such a case, hedgers would receive insurance at no cost, and speculators
would, as a group, not earn a risk premium.
It is possible that the risk premium is time-varying, particularly in agricultural commodities, which have a
seasonal harvest cycle. In the case of a commodity like wheat, for example, hedgers might be long wheat and
short wheat futures in the fall after the harvest has come in, and they might be short wheat and long wheat
futures in the spring when handlers of wheat make commitments to deliver wheat that they do not yet have in
hand. In terms of Figure 5.4, such a seasonal pattern would imply an HH schedule below the SS schedule in
the fall and an HH schedule above the SS schedule in the spring. In the fall, speculators are long futures and
~ ~
F0 E0 ( ST ); and in the spring, speculators are short futures and F0 E0 ( ST ). Futures prices would display
normal backwardation in the fall when speculators indirectly bear the risk of the long positions in the
commodity that has been harvested. In the spring, futures prices would display contango when speculators
indirectly bear the risk of the short positions in the underlying commodity assumed by hedgers.
24
Such an increase in risk premium as a function of position size is not modeled in the standard one-period capital asset pricing
model.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 53/358
5.5 SUMMARY
In this chapter, the role of speculators and the returns speculators can expect in an efficient market are
discussed. Some researchers argue that futures prices are unbiased estimates of subsequent spot prices, which
means that futures prices do not trend up or down over time. In that case, speculators as a group do not make
profits. The absence of speculative profits is possible if amateur speculators lose to professional speculators
so that speculators as a group do not make profits. The absence of profits is also possible if the risk of
holding futures contracts is fully diversifiable. If the risk can be diversified away, no risk premium is
required to induce speculators to hold futures contracts.
On the other hand, some researchers argue that futures prices trend up (normal backwardation) or
trend down (contango). If futures prices have a trend, speculators as a group make profits commensurate
with the level of risk that they assume from hedgers.
This chapter also derives the expected return to speculators under the Sharpe-Lintner capital asset
pricing model. Since futures contracts require no investment, the expected return of a futures contract equals
only the futures contract’s market risk premium (no riskless return is earned). The last section of the chapter
contains a model in which hedgers and speculators interact and determine an equilibrium futures price.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 54/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 55/358
6.9 RESUMEN
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 56/358
25
La notación en esta columna corresponde con el mes del año calendario (por ejemplo: 1 es enero, 2 es febrero y así
sucesivamente).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 57/358
TABLA 6.1 Cont. Especificaciones de los Contratos de Futuros para Mercancías Físicas
Mercancía Meses de los Unidades/
Horas de Negociación 26 Último Día de Negociación
(Bolsa de Valores) Contratos Cambio Mínimo en Precio
Alimentos y Fibras
10 toneladas métricas
Cacao (CSCE) 9:30-2:15 (EST) 12,3,5,7,9
$1(10)
37,500 libras
Café (CSCE) 9:15-1:58 (EST) 3,5,7,9,12
$0.0001($11.20)
Azúcar 112,000 libras Última día útil del mes
10:00-1.43(EST) 1,3,5,7,10
(mundial)(CSCE) $0.0001($11.20) anterior al mes de la entrega
Actual + 17 50,000 libras
Algodón (CTN) 10:30-3:00 (EST)
sucesivos $0.0001($5)
Jugo de Naranja 15,000 libras
10:15-2:45 (EST) 1,3,5,7,9,11
(CTN) $0.0005($7.50)
Ganada vacuno vivo 40,000 libras 20avo día calendario del mes
9:05-1:00 (CST) 2,4,6,8,9,10,12
(CME) $0.00025($10) del contrato
Del tercer al último día útil
9:25-2:00 1,3,5,7,9,12,1,3,5,7 25,000 libras
Cobre (CMX) del mes en que vence la
(EST) ,9,12 + act +3 $0.0005($12.50)
entrega
8:20-2:30 2,4,6,8,10,12, act 100 onzas troy
Oro (CMX)
(EST) +2 10 centavos($10)
8:20-2:30 50 onzas troy
Platino (NYM) 1,4,7,10, incl. act 3
(EST) 10 centavos ($5)
8:25-2:25 1,3,5,7,9,12 5,000 onzas
Plata (NYM)
(EST) act + 2 1/10 centavos ($5)
Petróleo
18 meses
Petróleo Crudo 9:45-3:10 consecutivos 1,000 barriles
(NYM) (EST) comenzando el 1 centavo ($10)
mes actual
15 meses
Petróleo Refinado 9:50-3:10 consecutivos 42,000 US galones
(NYM) (EST) comenzando el $0.0001
mes actual
15 meses
Gasolina 9:50-3.10 consecutivos 42,000 US galones
(NYM) (EST) comenzando el $0.0001
mes actual
Los contratos de futuros sobre mercancías físicas a ser entregadas a la opción del corto en algún momento en
el mes de la entrega. El primer día en que se puede realizar la entrega es el primer día de notificación. En
muchas mercancías, la posición larga que recibe una notificación de entrega tiene la oportunidad de vender
sus contratos de futuros y re-entregar la notificación; pero una vez que se termina la negociación de los
futuros, todas las posiciones largas circulantes no tienen otra opción que aceptar entrega. En la práctica, sólo
una pequeña parte de los contratos futuros ingresados alguna vez resultan en entrega.
La actividad de negociación en varios contratos de futuros varía según cambian las necesidades del
cliente y conforme surgen contratos competitivos. A fines de 1989, la CFTC aprobó 121 contratos de futuros
de mercancías físicas27. Algunos de estos nunca tuvieron éxito. Algunos que antes fueron exitosos ahora son
dominantes, Otros se negocian pero son muy inactivos. Ahora veremos los factores que dan origen a los
mercados de futuros de mercancías físicas y que determinan su éxito o su fracaso.
26
La notación en esta columna corresponde con el mes del año calendario (por ejemplo: 1 es enero, 2 es febrero y así
sucesivamente).
27
Informe 1989 de la Comisión de Negociación de Futuros de Mercancías.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 58/358
Incertidumbre
Dado que el propósito principal de los mercados de futuros es cubrir los riesgos, los mercados de futuros no
surgen si el precio de la mercancía no es incierto. Si el sostén del precio agrícola determina el precio del
trigo, no surgirán mercados de futuros de trigo. Los futuros de monedas no existirían en un sistema de tasas
de cambio fijos. La negociación de los contratos de futuros de café muere cuando el cartel internacional del
café “estabiliza” el precio del café en un nivel fijo. La incertidumbre en los precios se origina por la
incertidumbre sobre la oferta de mercancías y la incertidumbre sobre la demanda de las mercancías. La
incertidumbre sobre la cantidad relativa en el lado de la oferta y el lado de la demanda varía según el tipo de
mercancía.
Aún cuando las mercancías agrícolas producidas más estacionariamente ha aumentado durante cierta
parte del año en todo el mundo, la oferta es incierta debido a que el tamaño de la cosecha se ve afectado por
condiciones climatológicas. Por otro lado, la demanda global de la mayoría de los alimentos agrícolas y
petróleos es razonablemente estable dado que los patrones de consumo no cambiaron drásticamente de un
período a otro. Sin embargo, la incertidumbre de la demanda surge incluso después que un cultivo ha sido
cosechado porque la oferta de una mercancía sustituta puede ser incierta. Una cosecha adicional de maíz, por
ejemplo, puede afectar adversamente el precio del trigo.
Las mercancías en producción continua –como el petróleo, gas, madera y ciertos metales– enfrentan
incertidumbre tanto en el lado de la oferta como de la demanda. Las huelgas y los incrementos inesperados
en los costos afectan la oferta. Al mismo tiempo, la incertidumbre en la demanda es mayor que en las
mercancías agrícolas porque las mercancías de producción continua tienden a ser materiales industriales que
están sujetos al ciclo comercial.
Algunas mercancías físicas, específicamente oro y plata, tienen una oferta casi fija en el sentido que
las existencias circulantes de estas mercancías son mayores con relación a la producción anual. Como
resultado, la incertidumbre en el precio surge principalmente en el lado de la demanda. El oro es como
muchos instrumentos financieros que también tienen una oferta casi fija. Los precios del oro y los
instrumentos financieros dependen de las tasas de interés, inflación y otros factores macroeconómicos.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 59/358
no-competitivos cuando la producción de la mercancía subyacente es monopolizada o cuando son pocos los
compradores de la mercancía. En dichos mercados, el peligro es tan grande que el precio cash puede ser
manipulado para producir ganancias artificiales sobre el contrato de futuros (como a un interesado recién
llegado).
Al mismo tiempo, los mercados de futuros pueden mejorar la competencia en un mercado que no es
completamente competitivo. Por ejemplo, la negociación de los futuros de petróleo crudo le da a los usuarios
la oportunidad de asegurar el precio del futuro de petróleo sin negociar contratos a largo plazo con los
productores. El mercado de futuros proporciona una alternativa para negociar directamente con el productor.
Por esta razón, los productores con frecuencia se oponen a la introducción de los mercados de futuros. El
ejemplo más famoso de oposición de los productores a la introducción de mercados futuros es la oposición
de los agricultores de cebollas a la introducción de los futuros de cebollas, lo cual resultó en una prohibición
del congreso a la negociación de los futuros de cebollas en 1958.
Almacenamiento y Entrega
Las mercancías físicas sobre las cuales se basan los contratos de futuros pueden ser almacenadas directa o
indirectamente. El almacén usualmente se considera una necesidad para un contrato de futuros exitoso al
basarse en las opciones de compra del contrato para la entrega de la mercancía en una fecha posterior, y la
entrega se puede realizar solamente si la mercancía puede ser mantenida hasta la fecha de la entrega. Sin
embargo, si se puede producir la mercancía para la entrega, el almacén no sería necesario. Por lo tanto, un
contrato de futuros de huevos frescos existe, sin embargo, claramente los huevos no se pueden almacenar en
el sentido usual. La entrega no es el problema porque la disponibilidad futura de los huevos se puede
asegurar al tener las gallinas. Los huevos son almacenados indirectamente, como si estuviesen en las
gallinas. Similarmente, el ganado vacuno vivo no se puede almacenar en el sentido usual; no obstante,
claramente se puede almacenar (y alimentar) para su entrega posterior. Por lo tanto, en este sentido más
amplio, el requerimiento de almacenamiento y entrega se satisfacen para todas las mercancías físicas.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 60/358
Inventario
Tiempo
En el momento antes de la cosecha, como t en la Figura 6.2, normalmente existe un precio de futuros
para un contrato que vence en el momento H 1 , justo antes de la cosecha y para un contrato que vence en
un momento H 1 , justo después de la cosecha. Estos precios de futuros son pronósticos de los precios al
contado en los respectivos vencimientos de los contratos de futuros. La base en el momento t para el
contrato de futuros que vence antes del final de la cosecha del año pasado, Ft ( H 1) S t , representa un
mercado de cargo de acarreo y normalmente es positivo para reflejar los costos de almacenamiento. La base
en el momento t para el contrato de futuros que vence inmediatamente después de la cosecha,
Ft ( H 1) S t , representa un mercado invertido porque se espera que la cosecha nueva disminuya los
precios por debajo del nivel en el momento t .
Las figuras son útiles para identificar las tres principales fuentes de riesgo en mercancías estacionales.
Para un agricultor individual, el riesgo más importante es el riesgo de cantidad, esto es, el riesgo que se
relaciona con el tamaño de la cosecha en la cosecha estacional. Dado el clima y otros factores, el agricultor
no sabe la cantidad de cultivo que será cosechado.
El segundo riesgo es el riesgo de precio, el cual no sólo está presenta en el punto de cosecha sino
también durante el resto del año. Alrededor de la cosecha, la variabilidad de los precios refleja incertidumbre
sobre la cosecha agregada. El riesgo de precio y el riesgo de cantidad están relacionados. Si la cosecha
agregada es deficiente, el precio será mayor de lo normal. Si la cosecha agregada es buena, el precio será
menor que el normal. El agricultor cuya cosecha es representativa del conjunto agregado encontrará que el
precio y la cantidad están negativamente correlacionados –una cosecha deficiente se asocia con precios más
altos de lo normal, y una cosecha buena se asocia con precios menores de lo normal. Esta asociación
negativa mitiga colectivamente los riesgos de precio y cantidad. La cosecha de cada agricultor, sin embargo,
no está perfectamente correlacionada con la cosecha agregada dado que las condiciones climatológicas en las
diferentes partes del país varían. El agricultor que tiene una cosecha deficiente cuando otros agricultores han
tenido una buena cosecha estará particularmente mal porque el precio de la cosecha será bajo y al mismo
tiempo tendrá poco para vender. Correspondientemente, un agricultor con una cosecha excelente obtiene
grandes utilidades cuando otros agricultores tienen una mala cosecha. Después de la cosecha, el riesgo de
precio aún se mantiene porque la demanda por la mercancía es incierta. La demanda por la mercancía es
incierta porque la oferta de sustitutos es incierta y porque la demanda del consumidor final puede fluctuar.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 61/358
Precio
Ft(H - 1)
St
Ft(H + 1)
H
H
Tiempo
t H-1 H+1
Una vez que ya se ha cosechado, se saca el cultivo del almacén. Como ya vimos, la mayor parte del
riesgo durante esta parte del ciclo viene del lado de la demanda. Sin embargo, también esta presenta otra
fuente de riesgo, el riesgo del almacenamiento. Los costos de almacenamiento incluyen alquiler de almacén,
el costo del interés de los fondos destinados a la mercancía, seguro, mano de obra y manipuleo, y
desperdicio. Las fluctuaciones en estos costos afectan la rentabilidad del almacenamiento y el precio de la
mercancía en el período entre cosechas.
Se pueden utilizar los mercados de futuros para cubrir el riesgo del precio de la mercancía, pero los
mercados de futuros son menos adecuados para cubrir el riesgo de cantidad o el riesgo del costo de
almacenamiento. La discusión de la cobertura a continuación, por lo tanto, enfatiza el riesgo de precio de la
mercancía.
Mercancías No-Estacionales
Las mercancías no-agrícolas como los metales y petróleo son producidas de manera continua. Para todas
excepto los metales preciosos, los inventarios son relativamente pequeños con relación al consumo y no
fluctúan estacionalmente como los inventarios de mercancías agrícolas. Sin embargo, los inventarios pueden
agotarse rápidamente si se interrumpe la producción. Por ejemplo, una huelga en una mina de cobre puede
detener la producción de cobre y causar que los inventarios de cobre se agoten. Similarmente, las
restricciones de la OPEP sobre la producción de petróleo crudo pueden agotar los inventarios de petróleo. En
esos casos, el precio spot puede verse drásticamente afectado. En el caso de los metales preciosos como el
oro y la plata, los inventarios son grandes con relación a la producción, y los precios son determinados
principalmente por la demanda. La interrupción de la producción de metales preciosos toma más tiempo en
afectar los inventarios.
El patrón de precios de los recursos naturales fue analizado primero por Hotelling (1931), y más
recientemente, por Miller y Upton (1985). En un mundo de certeza, el Principio Hotelling establece que el
margen de utilidad de la minería de un recurso natural, el precio del recurso extraído, ST , y los costos de
producción marginales netos por unidad, CT , –se incrementan a la tasa de interés, r * :
( ST CT ) ( S 0 C0 )(1 r * ) (6.1)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 62/358
donde T se refiere al período futuro y 0 se refiere al período actual.28 Intuitivamente, la idea es que el
margen de utilidad debe incrementarse para cubrir los costos de los fondos dedicados a las reservas del
recurso natural. Si el margen de utilidad en el período futuro, T , es menor que la cantidad especificada por
(6.1), los productores incrementarán la producción en el momento 0 en lugar de esperar a producir en T . El
margen de utilidad podría invertirse a r * para obtener más en T que si la producción hubiese sido demorada
hasta mañana. La acción disminuye S 0 y disminuye el margen de utilidad en el momento 0 hasta que se
satisfaga la condición (6.1). A la inversa, si el margen de utilidad en T es mayor que la cantidad especificada
en (6.1), el valor actual de lo que el productor ganaría mañana excedería el margen de utilidad de hoy. El
productor estaría mejor restringiendo la producción hoy y produciendo más mañana. Podría pedir prestado
contra la utilidad de mañana para dar una cantidad que exceda el margen de utilidad en el momento 0. El
proceso de producir más mañana disminuye ST hasta que se satisfaga la condición (6.1). En un mundo con
incertidumbre, se mantiene una versión modificada del Principio Hotelling en la que se utilizan los valores
~ ~
esperados de ST y CT y la tasa de retorno esperada ajustada al riesgo es r * .
~ ~
Si existe un mercado de futuros, (6.1) se mantiene con FT sustituyendo por ST . Con esa sustitución y
cierta manipulación, (6.1) se puede escribir de la siguiente manera
~ ~
FT S 0 (1 r * ) CT C0 (1 r * ). (6.2)
Si no existen costos de producción, ya sea en el presente o en el futuro, el precio de los futuros está por
encima del precio spot por el costo del interés de los fondos destinados a la mercancía. Esto es lo mismo que
el modelo del costo-de-acarreo discutido en el Capítulo 3. El modelo de costo-de-acarreo dice que el precio
de los futuros de un contrato que vence en T excede el precio spot por el costo de mantener la mercancía. En
~
la ausencia de costos de almacenamiento aparte del costo del interés, se tiene que FT S 0 (1 r * ) . Este
equilibrio simple es apropiado para una mercancía como el oro que existe en grandes cantidades en forma
refinada.
La Ecuación (6.2) también muestra por qué el precio de los futuros de recursos naturales puede estar
temporalmente por debajo del precio spot. Esto puede suceder si el costo de producción actual, C0 , es
anormalmente alto, como en el caso de una huelga o una interrupción del suministro. En la crisis de Kuwait
de 1990-91, el precio spot del petróleo se elevó drásticamente con relación al precio de los futuros,
~
reflejando una elevación en C0 con relación a CT . A lo largo del tiempo, conforme los costos de producción
regresaron a la normalidad, el precio de los futuros de petróleo regresó a su prima normal sobre el precio
cash del petróleo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 63/358
es que los costos de almacenamiento resulten mayores de lo esperado. Por ejemplo, si los costos de
almacenamiento resultaron ser 3.5 centavos al mes, el comerciante pierde 1.5 centavos al comprar la
mercancía a $3.00 el 1º de setiembre y entregándola contra el contrato de futuros a $3.09 el 1º de diciembre.
La única manera de cubrirse contra este riesgo es realizar un contrato forward por los costos de
almacenamiento, algo que no siempre es posible. Si todos los costos de almacenamiento son asegurados de
antemano, el comerciante está en posición de tener un retorno libre de riesgo (asumiendo que no existe el
riesgo de incumplimiento). En la práctica, es poco probable que los comerciantes aseguren todos los costos.
La Tabla 6.2 analiza el resultado de la cobertura del comerciante si la mercancía es vendida a un cliente el 1º
de noviembre, un mes antes del vencimiento del contrato de futuros. El precio cash de la mercancía el 1º de
noviembre se asume en $2.70, una disminución de 30 centavos desde el 1º de setiembre. Al vender futuros,
el comerciante ha eliminado el riesgo de tal movimiento adverso en el precio de la mercancía. Sin embargo,
se mantiene el riesgo base. Como vimos en el Capítulo 4, el riesgo base es el mismo que el riesgo de costo de
almacenamiento si la mercancía es entregable. El efecto sobre el riesgo base se muestra al asumir tres precios
alternativos de futuros el 1º de noviembre –$2.73, $2.75 y $2.71– cada uno implicando una base diferente. Se
da un incremento en la base (debilitamiento) de cinco centavos cuando el precio de los futuros cae a $2.75 en
lugar de $2.73. Esto produce una pérdida neta de dos centavos. El comerciante tiene la opción de volver a
almacenar el 1º de noviembre para obtener una base de cinco centavos por mes, pero sólo es rentable si los
costos de almacenamiento son asegurados en un nivel menor. Si los costos de almacenamiento no son
asegurados, el incremento en la base refleja un incremento en todo el mercado de los costos de
almacenamiento, lo cual probablemente afectará al comerciante analizado en la Tabla 6.2. Por otro lado, una
reducción (reforzamiento) de la base a un centavo da como resultado una ganancia neta de dos centavos para
el comerciante, asumiendo que no tiene la obligación de pagar costos de almacenamiento para el mes de
noviembre.
Asuman ahora que la mercancía no se puede entregar a la presentación de un contrato de futuros. En
este caso, el riesgo base refleja el riesgo del precio de la mercancía así como el riesgo del costo de
almacenamiento. En la ausencia de entrega, los precios cash y los precios de los futuros no necesitan
converger al vencimiento. Por ejemplo, el comerciante de trigo en Oklahoma no puede entregar su trigo a un
costo razonable a la presentación del contrato de futuros en Chicago. Como resultado el precio del trigo en
Oklahoma cae con relación al precio del contrato de futuros, aunque la caída diferencial no puede ser mayor
que los costos de transporte. Por lo tanto, una cobertura en corto no es completamente efectiva. Mientras más
distante esté la mercancía cash en grado y espacio, mayor será la posibilidad que el precio cash y el precio de
los futuros se muevan en diferentes direcciones entre el tiempo de la cobertura y el vencimiento del contrato
de futuros. Como vimos en el Capítulo 4, la efectividad de la cobertura dependerá del grado de correlación
entre el precio cash y el precio de los futuros.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 64/358
incrementos en el precio del maíz mientras que realiza estos arreglos, el exportador compra 100 contratos de
futuros para diciembre. Si el precio del maíz sube, la ganancia del contrato de futuros compensa la pérdida
en el contrato de exportación. Conforme se adquiere el grado de maíz especificado y es embarcado a Nueva
Orleans, el exportador cierra su posición de futuros.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 65/358
generalmente van acompañados por incrementos en el precio de la harina. Si ese fuera el caso, se mantiene el
margen de utilidad y la cobertura no reduce el riesgo. Por ejemplo, si el molinero asegura el precio del trigo
con un contrato de futuros y los precios del trigo y la harina caen, él sufre una pérdida debido a la
disminución en el precio del output mientras que el precio del input está fijo en el nivel original más alto
dado por la cobertura. Si él no hubiese protegido el costo del input, el trigo, estaría mejor porque la
disminución en los precios de la harina sería compensada por la disminución en el precio del input. Por otro
lado, si se elevan los precios del trigo y la harina, la cobertura del precio del trigo produce una ganancia
global porque el precio del output se eleva mientras que el precio del input está fijo por la cobertura. Este
ejemplo, el cual asume que los precios del input y el output están correlacionados, se resume en la Tabla 6.3.
Es evidente que en este caso la “cobertura” –por la cual queremos decir: asegurar el precio del input– en
realidad incrementa la variabilidad de las utilidades. Sin embargo, si los precios del input y el output no están
correlacionados, la cobertura tendería a reducir el riesgo.
Muchos productores enfrentan este tipo de problema. Los fabricantes de dulces deben determinar si
van a proteger el precio del azúcar, cacao y otras materias primas. Los productores de cereales podrían
desear dar cobertura a los costos de los cereales. Los productores de cables eléctricos podrían desear dar
cobertura al costo del cobre. Los criadores de ganado podrían desear dar cobertura al precio de su alimento.
En cada caso, se debe determinar el deseo de asegurar el precio del input. Es más, si es deseable asegurar el
precio de los inputs, el fabricante podría elegir contratos forward con los proveedores o la cobertura del
mercado de futuros.
Ahora presentamos una formulación más precisa de la cobertura óptima cuando los precios del output
y el input son inciertos. La formulación es una modificación de la discusión de cobertura óptima presentada
en el Capítulo 4 para permitir la incertidumbre relacionada con el precio del output del productor así como el
precio de los inputs de las mercancías.
La notación utilizada en esta sección es la siguiente:
~
PT precio incierto de una unidad de output en el momento futuro T .
QP Número de unidades del producto a venderse
~
ST precio cash incierto del input en el momento futuro T . Nosotros asumimos sólo un input, aunque
fácilmente se podrían incluir inputs adicionales.
QS Número de unidades del input requeridas para producir QP unidades del producto.
F0 precio de futuros en el momento 0.
~
FT precio incierto de los futuros en el momento futuro T .
nF número de contratos futuros en cartera. ( nF es positivo para una posición larga y negativo para una
posición corta).
K costos fijos incurridos en el proceso de producción.
La utilidad incierta, ~h , de un productor que vende QP unidades, utiliza QS unidades como inputs y protege
nF unidades en el mercado de futuros es
~ ~ ~
~ P Q S Q ( F F )n K
h T P T S T 0 F (6.3)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 66/358
al dividirla entre QP , la ecuación puede re-enunciarse en términos de la utilidad por unidad de output:
~h ~ ~ QS ~ n K
PT ST ( FT F0 ) F (6.4)
QP QP Q P QP
Ahora definimos una unidad del input como el tamaño del contrato de futuros. En otras palabras, si el input
es azúcar, ahora definimos una unidad de azúcar como 112,000 libras, que es un contrato de futuros.
También definimos una unidad del output como la cantidad producida. Por el número de unidades contenidas
en el contrato de futuros. Por lo tanto, si se utilizan 112,000 libras de azúcar para producir 1,000,000
chocolates, una unidad de output es 1,000,000 chocolates. Esto significa que QS / QP 1.0 .29 Con estas
convenciones la ecuación se puede escribir como
~h ~ ~ ~ K
PT ST ( FT F0 )h
QP QP
donde h nF / QP es el ratio de cobertura. Para ser consistente con el Capitulo 4, escribimos la ecuación con
cambios en precios agregando y restando P0 S 0 al término de la mano derecha:
~h ~ ~ ~
c P S h F (6.5)
QP
donde c P0 S 0 K / QP
Siguiendo el procedimiento del Capítulo 4, la variación de la utilidad por unidad puede ser calculada como
h2 P2 S2 h 2 F2 2 PS 2h PF 2h SF . (6.6)
h bS bP . (6.9)
*
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 67/358
Si tanto el input como el output reaccionan de la misma manera al cambio en el precio de los futuros, esto es,
si los coeficientes de regresión son iguales, h* 0.0 , es el caso de una cobertura natural perfecta.
Para algunos productores, los precios del output pueden ser muy estables o pueden fijarse mediante contratos
a largo plazo de manera que bP 0.0 . En este caso, la cobertura óptima se determina por bS , la sensibilidad
del precio del input ante el precio de los futuros. El productor comprará futuros para asegurar el costo de los
inputs. Para algunos productores, el precio del input podría ser estable o puede fijarse mediante la compra de
futuros de manera que bS 0.0 . En este caso, el productor venderá futuros para cubrirse contra una
disminución en el precio del output.
Como vimos en el Capítulo 4, la cobertura óptima también puede desarrollarse comenzando con
ecuaciones de regresión que relacionen los precios del output y el input con el precio de los futuros:
~ ~
P a P bP F e~P , (6.10)
~ ~
a b ~
S S S e (6.11)
F S
29
Se debe notar que la empresa tiene un ratio fijo input/output y que planea producir un número fijo de unidades, todas las cuales
serán vendidas. En realidad, la empresa tiene cierta flexibilidad en la manera que combina los factores de producción, y puede no
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 68/358
En el análisis de regresión, las premisas que dictan el término de error son E (e ) 0, E (e F ) 0 . También
asumimos que los términos de error sucesivos en el tiempo son independientes. Los resultados de la
regresión se proporcionan en la Tabla 6.4. El calculo del ratio de cobertura óptima es
hˆ* bˆS bˆP = 0.9738 - 0.6487 = 0.3251 .
En otras palabras, por cada treinta galones de gasolina que la refinería produce, debe comprar
contratos de futuros por 0.3251 barriles de petróleo aún cuando se necesite un barril para producir treinta
galones de gasolina. La razón por la que sólo se compran 0.3251 barriles de futuros petróleo es que la
refinería tiene una cobertura natural parcial que se da por hecho que el precio recibido por la gasolina
compensa en parte cualquier cambio en el precio del petróleo utilizado para producir la gasolina.
Los valores pequeños de R 2 quieren decir simplemente que la efectividad de la cobertura no es
mucha. Los cambios en los precios de los futuros de petróleo sólo explican el 19.56 porciento de la variación
en los precios de la gasolina y sólo el 59.44 porciento de la variación de los precios del petróleo crudo de
Texas. Por lo tanto, el ratio de cobertura de 0.3251 puede resultar incorrecto después de todo. Por ejemplo,
supongamos que el estimado b̂S , está un error estándar por debajo de su verdadero valor y que el estimado
b̂P está un error estándar por encima de su verdadero valor. Entonces
h* = (0.9738 + 0.1098) - (0.6487 - 0.1741) 0.6090,
lo cual es casi el doble del ratio de cobertura estimado.
TABLA 6.4 Resumen de la regresión cambios de los futuros de petróleo crudo “amargo”, y
futuros de petróleo “ligeramente dulce” y cambios de precio de futuros de gasolina sin plomo de
noviembre de 1988 a noviembre de 1989.
Número de observaciones = 5330
P = -0.0312 dólares ˆ P = 0.9000 dólares
S = 0.1161 dólares ˆ S = 0.7994 dólares
F = 0.2580 dólares ˆ F = 0.6370 dólares
â P = -0.1985 s (aˆ P ) = 0.1187
b̂P = 0.6487 s (bˆ ) = 0.1741
P
2
R = 0.1956 ˆ e = 0.6516
P
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 69/358
Otra manera de expresar esta relación es decir que la base no puede exceder el costo del almacenamiento:
Ft S t Bt . Los comerciantes que almacenan mantendrán la mercancía sólo si la base, que puede ser vista
como el precio de almacenamiento, cubre los costos de almacenamiento. Se debe notar que el costo de
almacenamiento, Bt , es el costo marginal del almacenamiento por el período de tiempo en cuestión. Otra
tonelada de trigo será almacenada si el costo marginal, Bt , de almacenar esa tonelada es menor que el precio
recibido, Ft S t , por almacenar esa tonelada.
El comerciante que almacena toma el precio del almacenamiento como dado, pero en conjunto el
precio del almacenamiento depende de la interacción de la oferta y demanda del almacenamiento. La línea
sólida en la Figura 6.3 trata la curva de oferta agregada del almacenamiento para mercancías agrícolas como
trigo, maíz y soya. La curva se basa en datos reales que muestran cuánto está almacenado a cada valor de la
base, F S . A la izquierda del punto A, el costo marginal out-of-pocket del almacenamiento, B ,
representado por la línea punteada, excede la base.31 A la derecha del punto A, la curva de oferta del
almacenamiento coincide con el costo marginal del almacenamiento, B . El segmento horizontal extendido
de la curva de la oferta significa que los costos de almacenamiento son constantes por encima de este rango.
A la izquierda del punto A, las cantidades positivas son almacenadas inclusive cuando la base es menor que
el costo marginal del almacenamiento, B , dado que los productores derivan una utilidad de conveniencia por
tener la mercancía. Mientras menor sea la cantidad de mercancía en existencia, mayor deberá ser la utilidad
de conveniencia para compensar el hecho que el precio recibido por el almacenamiento es menor que el
costo marginal del almacenamiento. Presumiblemente el productor traspasa el costo del almacenamiento en
el precio final del producto.
Utilidad de Conveniencia
Q
QL A QM QH
La demanda del almacenamiento varía a lo largo del año de cosecha. Es mayor inmediatamente después de la
cosecha y disminuye conforme se va utilizando el cultivo. La línea vertical QH representa la demanda de
almacenamiento después de una cosecha excelente. Se interseca con la curva de oferta donde los costos
marginales de almacenamiento están subiendo, lo cual refleja el hecho que se deben estar utilizando medios
de almacenamiento deficientes y de costo alto para manejar la cosecha. Más adelante en la estación de
cosecha, la curva de demanda se traslada hacia la izquierda a un punto como QM , y la base cae.
31
La línea punteada se inclina hacia arriba sobre la premisa que los costos marginales del almacenamiento para los productores
que almacenan son mayor cuando se almacenan pequeñas cantidades; en otras palabras, existen economías de escala hasta el punto
A.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 70/358
TABLA 6.5 Precios de futuros de soya y la base mensual implícita en octubre de 1989
Vencimiento del Contrato Precio de Soya en centavos Base mensual32
Noviembre ‘89 552.5
Enero ‘90 564.25 5.88
Marzo ‘90 576.00 5.88
Mayo ‘90 586.00 5.00
Julio ‘90 592.25 3.18
Agosto ‘90 590.00 -2.25
Setiembre ‘90 573.00 -17.00
Noviembre ‘90 572.00 -0.50
La oferta del almacenamiento de mercancías no-estacionales como el petróleo, cobre y otros tendrá una
forma similar. Sin embargo, la demanda de almacenamiento no tendrá el patrón estacional predecible que es
típico en las mercancías agrícolas.
La sobre producción de petróleo colocaría una presión de demanda en los medios de almacenamiento
(un movimiento hacia la derecha sobre la función de la oferta de almacenamiento) y causaría un incremento
sobre la base.
Una huelga laboral en la producción de cobre reduciría la demanda de almacenamiento y causaría un
movimiento hacia la izquierda sobre la función de oferta de almacenamiento, tal vez inclusive hasta un punto
donde la base sea negativa. Así como en una mercancía agrícola, la base negativa significa que se anticipa un
incremento en la producción, en este caso, cuando acabe la huelga.
32
La base se calcula como la diferencia entre los precios de contratos de futuros adyacentes dividida entre el número de meses que
separan el vencimiento de los contratos.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 71/358
A diferencia de los que tratan de evitar el riesgo del precio de la mercancía, los especuladores asumen el
riesgo con la esperanza de obtener utilidad. La posición más riesgosa que puede asumir un especulador es
simplemente estar largo o corto en una mercancía. Una posición menos riesgosa es realizar un spread –
comprar(vender) un contrato de futuros y vender(comprar) otro contrato relacionado. Por ejemplo, un
meteorólogo que anticipa una sequía en Kansas podría decidir obtener una buena utilidad comprando futuros
de maíz. Si se equivoca, puede perder bastante. Un spread es menos riesgoso porque los precios de los
futuros tienden a moverse juntos. Los márgenes típicos se conforman de un vencimiento equiparado con otro
vencimiento en la misma mercancía o una mercancía con una mercancía relacionada para la misma fecha.
Por ejemplo, la sequía podría afectar los precios del maíz de manera diferente que los precios del trigo. En
este caso, podría ser deseable comprar futuros de maíz y vender futuros de trigo. Si los precios de los futuros
de maíz y trigo bajan juntos, no hay pérdida. Sólo si el maíz disminuye más que el trigo hay pérdida para el
especulador.
Los especuladores en futuros de mercancías, como en otros vehículos de inversión, deciden qué
comprar o vender sobre la base de un análisis fundamental o técnico. Los analistas fundamentales examinan
la oferta y la demanda de una mercancía y tratan de predecir la oferta y la demanda futura y así los cambios
futuros en el precio. En mercancías agrícolas, el clima es un factor importante en la oferta. En otras
mercancías, pueden ser importantes los factores políticos o la probabilidad de huelgas laborales. El análisis
técnico se enfoca en el patrón de los precios anteriores con la esperanza de predecir los cambios futuros en el
precio. Los técnicos preparan un diagrama de la conducta de los precios y el volumen de negociación y
buscan patrones que predigan los cambios futuros en los precios.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 72/358
pruebas empíricas concluyen que estas reglas no son rentables. Al llevar a cabo estas pruebas, se debe tener
cuidado en especificar la regla por adelantado antes de ver los datos, ya que uno siempre puede encontrar
alguna regla que hará dinero si se aplica a una secuencia particular de precios históricos.
Otra implicancia de los mercados eficientes es que el análisis fundamental tampoco puede proporcionar
retornos anormales si el análisis se basa solamente en la información pública disponible para todos los
analistas. En los mercados eficientes, toda la información pública se refleja en el precio actual. En otras
palabras, la información pública disponible en t no se puede utilizar para predecir el precio en t 1 .
Presumiblemente, los recursos son gastados en recolectar información con la esperanza de descubrir
información que no sea de conocimiento general, de manera que periódicamente se puedan obtener retornos
anormales. En los mercados eficientes, estos retornos anormales no deben, en promedio, exceder el costo de
adquirir la información especial que proporcionan estos retornos anormales.
Un tercer enfoque para probar la eficiencia de los mercados de futuros de mercancías es examinar
subgrupos de inversionistas, como comerciantes profesionales y consultores de inversión, para ver si ellos
pueden obtener retornos anormales. Qué es un retorno anormal requiere de una mayor discusión. Es poco
probable que el retorno anormal de los comerciantes y consultores profesionales sea cero; por que si lo fuese,
¿cómo alimentarían a sus familias? Uno esperaría que los inversionistas profesionales, aquellos que invierten
tiempo y dinero en el análisis, generen utilidades de negociación positivas o cobren comisiones. En la bolsa
de valores, un subgrupo muy popular para examinar son los fondos mutuos. Allí se ve que un fondo mutuo
típico no supere el desempeño del mercado, aunque cobran comisiones que permiten a los gerentes de cartera
alimentar a sus familias (algunas veces regiamente). Los descubrimientos en los mercados de futuros físicos
son consistentes con los mercados eficientes. Los tratados de Rockwell (1967) y Houthakker (1957)
examinan los retornos de los grandes protectores, los grandes especuladores y los pequeños comerciantes en
el mercado de futuros físicos. Estos estudios concluyen que los grandes especuladores sí obtienen utilidades,
lo cual es consistente con la idea que los especuladores profesionales deben tener una utilidad. Los estudios
no se ponen de acuerdo en si otros especuladores hacen o pierden dinero. Según la hipótesis nula de retornos
esperados cero, las ganancias de un grupo de especuladores deben compensar las pérdidas de los
especuladores restantes. Rockwell argumenta que este es el caso. Houthakker argumenta que los
especuladores pequeños también hacen dinero, lo cual significa que él rechaza la hipótesis nula de retorno
cero para los especuladores.
Elton, Gruber y Rentzler (1987, 1989) han examinado recientemente el desempeño de inversión de
los fondos de mercancías para el período 1979 a 1985. Los fondos de mercancías son análogos a los fondos
mutuos en la bolsa de valores en que son manejados profesionalmente. Durante los seis años analizados, el
retorno promedio anual del período de tenencia es –0.0007. Este retorno no refleja todos los costos de
transacción que se requiere paguen los inversionistas. Por lo tanto, los fondos de mercancías tienen un
desempeño inferior a otros instrumentos de inversión menos riesgosos como los títulos valores del gobierno.
El desempeño de los fondos de mercancías no apoyan la idea que los gerentes profesionales pueden obtener
utilidades positivas. De hecho, el desempeño es tan malo que Elton, Gruber y Rentzler (1989) cuestionan el
raciocinio de los inversionistas de fondos.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 73/358
La tasa de cambio se define como el precio en dólares de la moneda extranjera. La relación anterior,
conocida como la Ley de un Solo Precio (LOP), se mantiene debido al arbitraje de mercancías. Si el precio
en dólares de una mercancía excediese el costo de compra la mercancía en un país extranjero, alguien
dedicado al arbitraje compraría la mercancía en el país extranjero y la importaría a Estados Unidos,
depreciando así los precios en Estados Unidos y elevando los precios en el país extranjero. Supongamos, por
ejemplo, que el precio de una tonelada de trigo en Gran Bretaña cuesta 2.00 libras; la tasa de cambio es 1.60
dólares por libra; y el precio de una tonelada de trigo en Estados Unidos es 3.10 dólares. Alguien dedicado al
arbitraje podría comprar una tonelada por 3.10 en Estados Unidos y venderla en Gran Bretaña por 2.00 libras
y convertir las libras en (1.6)(2) = 3.20 dólares, lo cual proporciona una utilidad de 10 centavos por tonelada.
Este arbitraje eleva el precio del trigo en Estados Unidos y lo disminuye en Gran Bretaña hasta que el LOP
se re-establece. En la práctica, el LOP no se mantiene exactamente porque el arbitraje de mercancías es
costoso. Los costos de transporte y otros costos de transacción llevan a diferencias espaciales en el precio
entre los países. Así como el precio del trigo es diferente en Kansas y la ciudad de Nueva York, así el precio
del trigo establecido en dólares puede varias entre los diferentes países. Además, las comparaciones de
precios, inclusive de mercancías estrechamente definidas, usualmente no toman en cuenta totalmente las
diferencias de grado en las mercancías.
También se puede definir el LOP para los precios de los futuros:
Fd ,t (T ) Ft X (T ) F f ,t (T ) (6.16)
donde Fd ,t (T ) y Ff ,t (T ) son respectivamente los precios de futuros domésticos y extranjeros para los
contratos de mercancías que vencen en T , y Ft X (T ) es el precio de los futuros del contrato en moneda
extranjera con vencimiento en T . El LOP se debería mantener más estrechamente para los precios de los
futuros o los forward porque se pueden lograr acuerdos óptimos para el transporte a lo largo del tiempo hasta
el vencimiento.
TABLA 6.6 Desviaciones de la Ley de un Solo Precio para los precios de mercancías
estadounidenses y británicas33
Mercancía Media Desviación Estándar
Plata al contado 0.005 0.0200
Plata - forward de tres meses -0.0050 0.0146
Cobre al contado 0.0038 0.1959
Cobre - forward de tres meses 0.0116 0.0258
Futuros de café 0.1991 0.0941
Futuros de cacao -0.0386 0.0795
Futuros de trigo -0.2305 0.2023
La Tabla 6.6 presenta evidencia, tomada de Protopapadakis y Stoll (1983), sobre desviaciones del
LOP para ciertos precios inmediatos, forward y de futuros. Según el LOP, la media y la desviación estándar
del LOP serán cero porque el LOP es una relación de arbitraje no-aleatoria. La desviación promedio es
pequeña para la mayoría de las mercancías, particularmente para las mercancías como la plata que son
33
Las observaciones del precio de la mercancía y la tasa de cambio son semanales. El período observado es 1972-1980 para la
mayoría de las mercancías. Los datos del trigo son para el período 1976-1979.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 74/358
definidas precisamente. Las grandes desviaciones de la media en café y trigo se pueden atribuir a factores
especiales. En 1977-80, el precio de futuros de café estadounidense era relativamente alto con relación al
precio de futuros británicos debido a la manipulación de un contrato de futuros de café en Nueva York
realizado por un cartel de café sudamericano. El precio de los futuros de trigo estadounidense era bajo en
comparación con el precio de futuros británicos debido a subsidios en los precios agrícolas en Gran Bretaña
y la CEE mantuvo artificialmente un precio alto.
6.9 RESUMEN
En este capítulo, se identifican los principales contratos de futuros de mercancías físicas y se discuten los
factores que dan origen a los mercados de futuros de mercancías físicas. Se analizan los patrones de
inventario y precio para las mercancías con suministro estacional y las mercancías no-estacionales.
Se explica y modela el uso de contratos de futuros como una herramienta de cobertura para los
almacenan mercancías, los comerciantes de mercancías, y para los productores. Con frecuencia un productor
tiene precios inciertos en sus inputs y outputs, algo que hace que el problema de la cobertura sea más difícil
que si sólo existiese incertidumbre en el precio de una mercancía. En esta situación se presenta un marco
referencial de cobertura óptima y se ilustra el cálculo de la cobertura óptima.
34
Los detalles de este intento de acoso están en la Junta de Negociaciones de Chicago (1990).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 75/358
La base en mercancías físicas depende de la cantidad de la mercancía que debe almacenarse. Después
de la cosecha, la demanda de almacenamiento es alta, lo cual causa un movimiento a lo largo de la curva de
oferta de almacenamiento a un costo mayor de almacenamiento y a una base mayor. Cuando la demanda de
almacenamiento disminuye, la base también disminuye.
Los especuladores en futuros de mercancías físicas tienen dificultades para obtener utilidades, igual
que en otros mercados. La evidencia implica que los mercados de futuros son eficientes en el sentido que no
se logran utilidades anormales de manera consistente. Las reglas de negociación técnica no son vistas como
rentables (después de los costos de negociación). Los fondos de mercancías administrados profesionalmente
se desempeñan por debajo del mercado. El hecho que parezca que los especuladores no obtienen utilidades
implica que se provee a los que deseen una cobertura el servicio de asumir el riesgo a un costo muy bajo.
Los contratos de futuros de mercancías son utilizados por los que desean una cobertura y los
especuladores en muchos países. Este capítulo muestra cómo los precios de la misma mercancía están
vinculados en términos de monedas diferentes. El capítulo termina con una discusión sobre monopolios y
acoso corto.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 76/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 77/358
7.6 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 78/358
Arguably, the most exciting financial innovation of the 1980s has been the introduction of stock index
futures contracts. These contracts, written on the value of various stock index portfolios, provide important
benefits to stock portfolio managers. The uses and benefits of these contracts are described in this chapter.
We begin with a description of the history of stock index futures contracts in the U.S. and an explanation of
current contract designs. The second section details the composition of the stock indexes that underlie
currently traded index futures contracts.
Section 3 describes the index arbitrage that holds the cost of carry relation in alignment and explains
the concept of “program trading.” In section 4, the intraday price behavior of the index and its futures
contracts is investigated to see how well the price movements in the two markets are synchronized. The
chapter concludes with an illustration of hedging with stock index futures contracts.
35
The composition of the various stock indexes is discussed in the next section.
36
In June 1987, the Chicago Mercantile Exchange and the New York Futures Exchange changed the settlement of their S&P 500
and NYSE index futures contracts from the close of trading to the open in an attempt to mitigate concern about occasional
abnormal stock price movements in the “triple witching hour.” The futures contracts on the Major Market and Value Line indexes
continue to settle at the close. For an analysis of the effects of this change, see Stoll and Whaley (1991).
37
Margins are adjusted when the risk of the underlying index changes perceptibly. Prior to the October 19, 1987, stock market
crash, speculative margin on the S&P 500 futures contract was $6,000. Immediately following the crash, speculative margins were
set as high as $20,000.
38
The index composition is described later in this chapter.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 79/358
500 futures contracts were active on November 13 –the December 1991 and the March and June 1992
contracts. The estimated trading volume on that day was 42,125 contracts. The implied dollar stock
equivalent of this volume of trading is at least $398.30 x 500 x 42,125, or $8.39 billion. As is usually the
case, the nearby futures contract is the most active, as is reflected through the higher open interest figure for
the December contract. The underlying S&P 500 index level, 397.42, is also reported in the table, just below
the futures price summary.
TABLE 7.1 Contract specifications of stock index futures contracts trading in the U.S.
Index Trading Contract Units/Minimum
Last Day of Trading40
(Exchange) Hours Months39 Price Fluctuation
S&P 500 8:30-3:15 500 x index/0.05
3,6,9,12 Third Thursday
(CME) (CST) ($25)
NYSE Index 9:30-4:15 500 x index/0.05 Thursday Preceding third
376,9,12
(NYFE) (EST) ($25) Friday
Major Market 8:30-3:15 3 current months 250 x index/0.05 First business day prior to
Index (CBOT) (CST) plus 3,6,9,12 ($12.50) Saturday following third Friday
Value Line 8:30-3:15 500 x index/0.05
3,6,9,12 Third Friday
Index (KC) (CST) ($25)
Mini Value Line 8:30-3:15 100 x index/0.05
3,6,9,12 Third Friday
Index (KC) (CST) ($5)
TABLE 7.2 Stock index futures contract prices at the close of trading on Wednesday,
November 13, 1991.
FUTURES
S&P 500 INDEX (CME) 500 times index
Open High Low Settle Chg High Low Open Interest
Dec 395.00 398.50 394.30 398.30 +1.00 401.50 316.50 139,341
Mr92 396.80 400.50 396.50 400.35 +1.00 404.00 374.70 7,544
June 398.30 402.35 398.30 402.20 +1.10 407.00 379.00 1,102
Est vol 42,125; vol Tues 41,413; Open Int 148,048, + 916.
Index prelim High 397.42; Low 394.01; Close 397.42 +0.68
39
The notation used in this column corresponds to the month of the calendar year (e.g., 1 is January, 2 is February, and so on).
40
All stock index futures contracts are cash settled.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 80/358
CBT –Chicago Board of Trade. CME –Chicago Mercantile Exchange. KCBT –Kansas City Board of
Trade. NYFE –New York Futures Exchange, a unit of the New York Stock Exchange.
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones &
Company, Inc. All Rights Reserved Worldwide.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 81/358
where N is the number of stocks in the index. This market value is then scaled by a divisor so that the index
in period t is
N
n pi ,t
i 1 i ,t
St (7.2)
Divisort
The divisor represents what the stocks currently in the index would have been worth in a base period. In the
base period, the divisor is the market value of the stocks in the index,
Divisor0 i1 ni , 0 pi ,0
N
(7.3)
Over time, the numerator of (7.2) changes because stocks enter or leave the index or because shares are
issued or repurchased by companies. Because such changes do not reflect a change in the value of the stocks,
an adjustment to the divisor is made on the day that a change in the index composition occurs. The new
divisor on day t is just the old divisor on day t adjusted by the ratio of the market value of the new index
composition on day t divided by the market value of the old index composition on day t ,
Both the S&P 500 and NYSE Composite indexes are value-weighted. The S&P 500 consists of 500 common
stocks, the majority of which trade on the NYSE, although about fifty stocks trade on the American
Exchange and in the OTC market. The index was designed by Standard & Poors’ to contain stocks from a
broad variety of industry groupings. The market value for the base period of the S&P 500 is based on the
average market values of the component stocks during the years 40; 1941 through 1943. At that time, the
index was set equal to 10. The NYSE Composite contains all common stocks traded on the NYSE, slightly
more than 1,500 in number. The base period for the NYSE index is December 31, 1965, at which
time the index was set equal to 50. As Table 7.2 shows, the values of the S&P 500 and NYSE Composite
stocks indexes were 397.42 and 219.37, respectively, at the close of trading on November 13, 1991,
reflecting percentage gains of 3,874 percent and 339 percent, respectively, from their base periods.
N
pi ,t
St i 1
(7.5)
Divisort
In a price-weighted index, the divisor in the base period equals the sum of the prices of the stocks in the base
period, that is,
Divisor0 i 1 pi , 0
N
(7.6)
Like a value-weighted index, the divisor of a price-weighted index is adjusted to reflect stock splits and stock
dividends so that the index level remains unchanged during the stock split/stock dividend process [i.e., in the
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 82/358
manner of (7.4)]. Unlike the value-weighted index, however, the divisor of the price-weighted index is
unaffected by new stock issues or share repurchases.
The best known price-weighted arithmetic index is the Dow Jones Industrial Average (DJIA), which
consists of thirty “blue-chip” stocks. In an attempt to create an index that mimics the price movements of the
DJIA, the American Exchange created the Major Market Index (MMI). This price-weighted index contains
twenty stocks, seventeen of which are also members of the DJIA. Table 7.2 shows that the value of the MMI
at the close of trading on November 13, 1991, was 327.25.
RS ,t N i 1 (1 Ri ,t ) 1 .
N
(7.7)
This return is used to update the index from the previous period,
St St 1 (1 RS ,t ) . (7.8)
Currently, the only equal-weighted geometric index is the Value Line Index. It consists of approximately
1,700 stocks. Approximately ninety percent of the Value Line index capitalization is from shares traded on
the NYSE, one percent from AMEX, and nine percent OTC. The Value Line index and its futures contracts
are of limited interest for two reasons. First, the index weights all stocks equally so small stocks have as
much impact on the index movements as large stocks. For an index to track the behavior of the “market,”
much greater weight should be placed on large capitalization issues. Second, geometric averaging causes the
rate of return on the index to be less than the rate of return that would be earned by an equal-weighted
investment in each of the 1,700 stocks. As a result, price movements (returns) of the Value Line index are
not as strongly correlated with most stock portfolios as are other indexes, which makes the Value Line
futures contract less useful for hedging purposes. Table 7.2 shows that the open interest of the Value Line
futures is much lower than the futures contracts on the other indexes. The Value Line index closed at 326.47
on November 13, 1991.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 83/358
41
The value-weighted arithmetic index consists of 100 shares of Stock A and 100 shares of Stock B. At time 0, the market
capitalization is 4,000, which is adjusted to an index level of 100.
42
The price-weighted arithmetic index at time 0 equals the sum of the share prices of Stock A and Stock B divided by the divisor.
43
The equal-weighted geometric index equals 100 in the base period. The value at time 1 equals the time 0 index value times the
square root of the product of one plus the rate of return on Stock A and one plus the rate of return on Stock B.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 84/358
TABLE 7.4 Summary statistics of weekly percentage rates of price appreciation in five U.S. stock
indexes during the calendar year 1989.44
Means and Standard Deviations of Index Returns
Index Mean Return Standard Deviation
DJIA 0.4537 1.6640
MMI 0.5072 1.7453
S&P 500 0.4481 1.5825
VL 0.1809 1.3180
NYSE 0.4127 1.4916
Contemporaneous Correlations Between Pairs of Index Returns
Index MMI S&P 500 VL NYSE
DJIA 0.9779 0.9774 0.8880 0.9750
MMI 0.9497 0.8104 0.9403
S&P 500 0.9137 0.9972
VL 0.9337
Ft St e ( r d )(T t ) , (7.9)
where Ft and St , are the time t prices of the futures contract and the underlying stock index, respectively.
Note that the derivation of this relation in Chapter 3, as it applies to stock index arbitrage, implies that the
cash dividends, as they accrue through time, are being reinvested in the stock index portfolio.
Assuming that cash dividends are a constant, continuous proportion of the index level may be
inappropriate, particularly for a narrow-based index like the MMI, where the small number of stocks in the
index portfolio implies an obvious discreteness and seasonality of cash dividend payments.45 In such a case,
an assumption that the amount Di and the timing ti of the discrete cash dividends paid during the futures
contract life (i.e., between time t and time T ) are known is usually used. Furthermore, rather than assuming
that the dividends are being reinvested in the stock index portfolio, dividends are assumed to be reinvested at
the riskless rate of interest until the futures contract expires.
Under these assumptions, stock index arbitrage involves the transactions shown in Table 7.5a. The
~
long position in the index portfolio provides a terminal value equal to the uncertain index price ST plus a
n
known aggregate dividend income (plus accrued interest) i 1
Di e r (T ti ) . The stock portfolio position is
financed completely with riskless borrowings, which are repaid at time T at a cost St e r (T t ) ). The short futures
~
position has a terminal value ( ST Ft ) . Since the arbitrage portfolio involves a zero investment outlay and
has no risk, the net terminal value of the portfolio must equal zero for the market to be in equilibrium. Thus,
under the assumption of known discrete dividends, the cost-of-carry relation is
44
Rates of price appreciation axe computed on the basis of the closing index levels each Wednesday during 1989. Cash dividends
paid on index stocks are not considered.
45
Harvey and Whaley (1992) show pronounced seasonality in the cash dividends of the S&P 100 index, which contains
approximately forty percent of the market value of the S&P 500 index. In particular, during the period 1983 through 1989,
dividends tend to be highest in the months of February, May, August and November.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 85/358
Ft S t e r (T t ) i 1 Di e r (T ti )
n
(7.10)
TABLE 7.5a Index arbitrage transactions for establishing the relation between index futures and
underlying index prices, assuming known discrete cash dividends. Ft St e r (T t ) i 1 Di e r (T ti )
n
A simple version of the cost-of-carry relation arises if one assumes dividends and interest are paid at the end
of the period corresponding to the life of the futures contract. Table 7.5b presents arbitrage transactions for
this case and shows that this simple cost-of-carry relation is
Ft St (1 r * d * ) (7.11)
where r * is the rate of interest and d * is the dividend yield over the remaining life of the futures contract.
Violations of the cost-of-carry relation (7.9), (7.10), or (7.11) signal profitable index arbitrage opportunities.
If, for example, the observed futures price is above the theoretical futures price as implied by the right-hand
side of (7.9), (7.10), or (7.11), arbitrageurs sell futures and buy the underlying stocks, driving the price of the
futures down and the prices of stocks up. The arbitrage becomes unprofitable (3 when the futures price
reflects the cost of carrying the underlying stocks, that is, the interest cost less the cash dividends.
TABLE 7.5b Index arbitrage transactions for establishing the relation between index futures and
underlying index prices, assuming dividends and interest are paid at maturity. Ft St (1 r * d * )
Position Initial Value Terminal Value
~
Buy index portfolio St ST d * S t
Borrow St St St (1 r * )
~
Sell futures contract 0 ( ST Ft )
Net portfolio value 0 Ft St (1 r * d * )
Unlike typical basis arbitrage, the underlying commodity is a precisely weighted portfolio of
common stocks, rather than a single asset. For example, engaging in index arbitrage with the S&P 500 index
requires a mechanism for buying or selling quickly and simultaneously all 500 stocks in the S&P 500 index
portfolio. Since the simultaneous purchase or sale of the stocks in a precisely weighted and timely fashion is
beyond human capability, computers and computer programs are usually used to place transaction orders as
well as to assist in the execution of those orders. For this reason, trading of portfolios of stocks is called
program trading, although program trades can also be done by manually preparing order tickets for each
stock. NYSE statistics define a program trade as any order for a portfolio of 15 or more stocks.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 86/358
The interest rate from the Treasury bill substitute strategy, r * , is 2.5 percent. In other words, a pension fund
that might ordinarily invest $3,486,000 in T-bills to earn 1.6 percent over 73 days could invest the same
amount of money in a Treasury bill substitute to earn 2.5 percent over 73 days. To do so, the $3,486,000 is
invested in the index portfolio (i.e., 10,000 units of the index are purchased) and twenty index futures
contracts are sold (recall each index futures is 500 times the index value). Over the 73-day period, the index
portfolio will generate $27,900 in cash dividends and the index level will appreciate by 5.90 relative to the
futures (because the futures price and index level converge at the end of 73 days), for a total price
appreciation of $59,000. The overall rate of return on the Treasury bill substitute position is (27,900 +
59,000)/3,486,000 or 2.5 percent.
Stock Replacement
A second example of how index futures may be used to generate a higher return than an investment with
equivalent risk is a stock replacement strategy. When the actual futures price is below the theoretical futures
price, an arbitrageur enacts a short arbitrage –the short sale of stocks and the purchase of futures contracts.
But stock portfolio managers, too, can profit from such an opportunity by selling their stock portfolios and
using the proceeds to buy index futures and Treasury bills, that is, by engaging in stock replacement.
To illustrate a stock replacement strategy, consider the previous example in which the current S&P
500 index level is 348.60, the time to expiration of the nearby S&P 500 futures contract is 73 days, the rate
of return on a 73-day T-bill over the next 73 days is 1.60 percent, and the future value of the cash dividends
on the S&P 500 over the next 73 days is $2.79. However, this time, assume the nearby S&P 500 futures price
is $350.25. On the basis of these figures, the theoretical futures price is
~ 50,000,000 ~ ~
VS &P 500,T ( ST 2.79) 143,430.87 ST 400,172
348.60
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 87/358
in 73 days. On the other hand, if he liquidates his S&P 500 stock portfolio and buys T-bills and the nearby
S&P 500 futures contract, the portfolio value for the stock replacement strategy (SRS) will be
~ 50,000,000 ~
VSRS ,T ( ST - 350.25) 50,000,000(l.016)
348.60
~ ~
143,430.87 ST - 50,236,662 50,800,000 143,430.87 ST 563,338 .
Note that the stock replacement strategy is certain to have a terminal value $163,166 higher than the stock
portfolio strategy. The fact that this incremental value is certain reflects the fact that, while each strategy’s
terminal value is uncertain, both strategies have equal risk. If the observed futures price is below its
theoretical level, however, the stock replacement strategy will dominate.
46
See Stoll and Whaley (1987, p. 18).
47
In August 1990, the NYSE implemented a rule requiring a downtick on each stock in an index arbitrage program purchase if the
DJIA rose by 50 points or more and an uptick on each stock in an index arbitrage program sale (short or from a long position) if
the DJIA declined by 50 points. This rule is counter productive because it impedes index arbitrage.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 88/358
The efficacy of the index arbitrage process has been examined in a number of theoretical and
empirical papers.48 In general, these papers find that observed futures prices can deviate from the theoretical
futures price specified by arbitrage conditions by more than normal transaction costs. This is particularly the
case for deviations of the futures price below the theoretical price. Such deviations may be difficult to
arbitrage though because of the short sale restrictions and because of the lack of a sufficient number of
institutions willing to engage in stock replacement strategies.
Ft St e ( r d )(T t ) ,
should be satisfied at every instant t during the futures contract life. If such is the case, the instantaneous
rate of price appreciation in the stock index equals the net cost-of-carry of the stock portfolio plus the
instantaneous relative price change of the futures’ contract. To see this, take the natural logarithm of (7.9) at
time t and at time t 1 :
Several implications follow from (7.14) under the assumptions that the short-term interest rate and the
dividend yield rate of the stock index are constant and that the index futures and stock markets are efficient
and continuous:
~
a. The expected rate of price appreciation on the stock index portfolio E ( RS,t ) equals the net cost of
~
carry ( r d ) plus the expected rate of return on the futures contract E ( RF,t )
b. The standard deviation of the rate of return on the futures contract equals the standard deviation of
the rate of return of the underlying stock index.
c. The contemporaneous rates of return of the futures contract and the underlying stock portfolio are
perfectly positively correlated.
d. The rates of return of the futures contract and of the underlying stock index portfolio are serially
uncorrelated.49
e. The non-contemporaneous rates of return of the futures contract and the underlying stock portfolio
are uncorrelated.
48
Cornell and French (1983), Figlewski (1984a), Gastineau and Madansky (1983), Modest and Sundaresan (1983), Peters (1985),
Stoll and Whaley (1986b), MacKinlay and Ramaswamy (1988), Kleidon (1991), Kleidon and Whaley (1991), and Miller,
Muthuswamy, and Whaley (1991) examine the arbitrage process and consider possible explanations for observed deviations from
theoretical prices. Other papers, notably Garcia and Gould (1987), Gould (1988), and Brennan and Schwartz (1990), analyze
strategies for trading on mispricing.
49
Technically speaking, more than an assumption of market efficiency is needed to ensure serially uncorrelated rates of return. It
must also be the case that the expected rates of return of the futures and stock index are constant. [See Fama (1976, pp. 149-151).]
Such an assumption is reasonable since the rate of return series that we will examine below are intraday.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 89/358
Naturally, all of the above implications are based on the assumption that the cost-of-carry relation (7.9) holds
at all points in time. It has been shown, however, that (7.9) does not hold exactly; indeed, one of the puzzles
in stock index futures is the frequency with which deviations from the cost-of-carry relation are observed.
Stoll and Whaley (1986b, Table 23A), for example, report frequent violations of the cost-of-carry relation in
excess of transaction costs using hourly S&P 500 index and index futures data during the period April 1982
through December 1985. The frequency of violation is nearly eighty percent for the June 1982 futures
contract. For more recent contract maturities, however, the frequency falls below fifteen percent. MacKinlay
and Ramaswamy (1988, Table 6) report similar results for the S&P 500 futures contracts expiring in
September 1983 through June 1987. Using fifteen-minute price data, they find that the cost-of-carry relation
is violated 14.4 percent of the time on average.
Violations of the cost-of-carry relation may appear for a variety of reasons. Some, like transaction
costs, were discussed in the last section. The presence of transaction costs tends to introduce noise in the rate
of return relation (7.14). An important reason not mentioned in the last section is the infrequent trading of
stocks within the index. Markets for individual stocks are not perfectly continuous. Consequently, stock
index prices, which are averages of the last transaction prices of component stocks, lag actual developments
in the stock market. Fisher (1966) describes this phenomenon. Cohen, et (1986, Ch. 6) give a more general
discussion of serial correlation of stock index returns in terms pf delays in the price adjustment of securities.
Lo and MacKinlay (1988) model the effects of infrequent trading on index returns under certain restrictive
assumptions. Assuming that the index futures prices instantaneously reflect new information, observed
futures returns should be expected to lead observed stock index returns because of infrequent trading, even
though there is no economic significance to this behavior whatsoever.
Stoll and Whaley (1990b) use five-minute, intraday rate of return data for the S&P 500 index and the
nearby S&P 500 futures contracts to (a) model and purge the effects of infrequent trading in the stock index
portfolio, and (b) assess the degree of simultaneity between returns in the index futures and stock markets.
The effects of infrequent trading are shown in Table 7.6. Note that, while the S&P 500 futures contract
returns have virtually no serial correlation, the returns of the S&P 500 index portfolio are strongly positively
serially correlated. The first-order serial correlation in the S&P 500 index returns exceeds 0.5. Because not
all stocks within the S&P 500 index portfolio trade in every five-minute interval, a market movement within
this interval may not be recorded in the price of less actively traded stocks until some time later when the
stock finally trades. The effect of this phenomenon is positive serial correlation in the portfolio return series.
The serial correlation does not disappear until lag 4 or 5 using five-minute returns.
The effects of infrequent trading on observed stock index returns are modeled theoretically and
estimated empirically in Stoll and Whaley (1990b). The residuals (return innovations) from the estimated
model are examined to assess the degree of any remaining positive serial correlation. The last pair of
columns in Table 7.6 show these results. With the effects of infrequent trading modeled and purged, the
return innovations of the S&P 500 index are virtually white noise. None of the estimated serial correlation
coefficients exceed 0.02 in absolute magnitude.
Finally, to assess the degree of simultaneity between the S&P 500 index futures and stock market returns, the
return innovations of the S&P 500 index are regressed on lag, contemporaneous, and lead futures returns,
S ,t k 3 k RF ,t k ut .
3
(7.15)
The regression results are shown in Table 7.7. In addition, for purposes of comparison, the regression results
of observed S&P 500 index returns regressed on lag, contemporaneous, and lead futures returns are also
reported.
The return innovation regression results in Table 7.7 indicate that the dominant relation between the
two markets is contemporaneous. The estimated coefficient of the contemporaneous futures return, ˆ0 , in
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 90/358
the return innovation regression is 0.1338, higher than any of the leading or lagged coefficients. The
estimated coefficient of the lag one futures return, ̂1 , is 0.1015, showing that there is a tendency for the
futures market to lead the stock market. All other coefficients in the return innovation regression are
indistinguishably different from zero in an economic sense. When stock index returns are used as the
dependent variable, the leading effect of the futures market appears considerably longer, but most of this is
illusion attributable to infrequent trading in the stock market. Overall, the evidence supports the notion that
futures markets tend to play a price discovery role in the marketplace.
TABLE 7.6 Estimated serial correlation coefficients of observed returns of the S&P 500 index ( RSo,t )
and the S&P 500 index futures contract ( RFo ,t ) for the 1249-day period April 21, 1982, through March
31, 1987.50
k ( RSo,t , RSo,t k ) k ( RFo ,t , RFo ,t k ) k ( S ,t , S ,t k )
Lag k No. of ̂ k 52 t ( ˆ k ) 53 ̂ k 18 t ( ˆ k ) 19 No. of ̂ k 18 t ( ˆ k ) 19
Obs.51 Obs.54
1 86,952 0.5117 175-61 0.0229 6.77 84,454 0.0071 2.06
2 85,703 0.2654 80.60 0.0265 7.76 83,205 0.0053 1.52
3 84,454 0.1312 38.46 0.0015 0.45 81,956 0.0068 1.95
4 83,205 0.0759 21.96 -0.0137 -3.96 80,707 0.0050 1.41
5 81,956 0.0460 13.17 -0.0222 -6.36 79,458 0.0052 1.48
6 80,707 0.0199 5.64 -0.0108 -3-06 78,209 -0.0042 -1.18
7 79,458 0.0077 2.18 -0.0087 -2.46 76,960 -0.0119 -3.30
8 78,209 0.0154 4.32 -0.0015 -0.42 75,711 0.0017 0.46
9 76,960 0.0195 5.42 0.0039 1.07 74,462 -0.0005 -0.15
10 75,711 0.0110 3.04 -0.0030 -0.83 737213 -0.0082 -2.22
11 74,462 0.0018 0.49 0.0047 1.29 71,964 -0.0163 -4.37
12 73,213 0.0019 0.51 0.0002 0.07 70,715 -0.0067 -1.77
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 91/358
TABLE 7.7 Parameter estimates from regressions of S&P 500 index returns/return innovations on lag,
contemporaneous, and lead nearby S&P 500 futures returns for the 1249-day period April 21, 1982
through March 31, 1987.55
Returns : RS ,t k 3 k RF ,t k ut and Return Innovations S ,t k 3 k RF ,t k ut
3 3
Suppose that your research director has informed you that the market (as reflected by the S&P 500)
will drop by sixteen percent over the next three months. You have a great deal of confidence in his prediction
so you decide to hedge the market risk of your portfolio. One option that you have is to liquidate the stock
portfolio and buy T-bills, however this strategy would not allow you to capture the non-market returns that
your portfolio of “winners” is expected to earn over the next three months. Selling S&P 500 futures
contracts, on the other hand, allows you to hedge the market risk of the stock portfolio without selling your
stocks.
55
The numbers in this table are taken from Stoll and Whaley (1990b, Table 5).
56
Parameter estimates obtained from times series regression across all five-minute returns in all days of the period, excluding
overnight returns and the first two returns each trading day.
57
The t - ratio corresponding to the null hypothesis that the respective coefficient equals zero.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 92/358
~ ~
~ S S 0 (1 r * d * ) RS r* d *
RF T .
S 0 (1 r * d * ) (1 r * d * ) (1 r * d * )
~ ~ ~
Rearranging to isolate RS , we get: RS RF (1 r * d * ) r * d * .
Substituting for the stock index return, the expression for the stock portfolio beta becomes
~ ~ ~ ~
Cov[ R p , RF (1 r * d * ) r * d * ] Cov[ R p , RF ]
p ~ ~ .
Var[ RF (1 r * d * ) r * d * ] Var[ RF ](1 r * d * )
The remaining step in showing the relation between the stock portfolio beta and the hedge ratio involves
~ ~
substituting the relations between returns and price changes. These relations are R p p / p0
~ ~
and RF F / F0 . Hence, the stock’s rate of return beta p is.
1 ~ ~
~ ~ Cov( p , F ) ~ ~
Cov( p / p0 , F / F0 ) p0 F0 F0 Cov( p , F )
p ~ ~
Var ( F / F0 )(1 r * d * ) 1 ~ p Var ( F )(1 r d )
* *
2
Var ( F )(1 r d )
* * 0
F0
Using the definition of the optimal hedge ratio given in Chapter 4 and assuming the cost-of-carry relation,
S
(4.9), p can be written as p h* 0 .
p0
Finally, the initial investment in the stock portfolio and the cash index with respect to which S 0 is calculated
are the same, so S 0 p0 . This implies that p h*
In other words, the optimal hedge ratio is the negative of the stock portfolio beta. In the case of the example,
the optimal hedge ratio is
h* 1.2
The optimal number of futures contracts to sell is therefore the stock portfolio beta times the number of units
50,000,000
of the stock portfolio, 1.2 321.17
373.63(500)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 93/358
changes and the futures price changes. Over the life of the futures contract, the futures and stock index price
changes are perfectly correlated and the standard deviation of the futures price change equals the standard
deviation of the stock index price change, so the correlation coefficient may be written
p ,S
p , F p ,S .
p S
On the basis of the given values ( p 1.20, S 0.25 and p 0.40 ); the correlation coefficient,
ρ p,F , is 0.75 . The R-squared is thus 0.5625 and the proportion of the stock portfolio return variance that is
unrelated to the return variance of index futures is 1 - 0.5625 = 0.4375. The remaining variance of the rate of
price change on the hedged portfolio is therefore
h2 0.4375(0.16) 0.07
TABLE 7.8 Hedging market risk of a stock portfolio that has a 1.2 and a three-month dividend yield of
0.8 percent.
Cash Market December Futures
Value of
Value of Stock Value of Hedged
Index Level Futures Price Future’s
Portfolio Portfolio
Position58
Sept 15 373.63 50,000,000 375.50 -60,300,000
Dec 15 298.90 401400,00059 298.90 -48,000,000
Gain -9,600,000 12,300,000 2,700,000
60 26
Return (%) -20.00 -19.20 -20.40 -24.60 5.4026
The overall rate of return on the unhedged portfolio can be measured easily by focusing on the price
appreciation and dividend yield components of total return, that is,
40,000,000 400,000
Rp 1 19.20%
50,000,000 50,000,000
To find the hedged portfolio return, we must also compute the rate of return on the futures. At the outset, the
S&P 500 index level was 373.63 and the three-month S&P 500 futures price was 375.50. If the S&P 500
58
The optimal hedge involves selling 321.17 futures contracts, with each contract valued at 500 times the index futures price.
59
Includes dividends of $400,000.
60
Dollar gain divided by the initial stock portfolio value, $50,000,000.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 94/358
index level fell by twenty percent over the three-month period, the new index level and futures price (recall
that futures had three months to expiration when they were sold) are 373.63(0.80) or 298.90. The rate of
return on the index futures over the period was therefore
298.90
RF 1 20.40%
375.50
Thus, the total return of the hedged portfolio over the three-month period is
F 375.50
Rh R p h 0 RF - 0.1920 - 1.2 ( - 0.2040) 5.40%
S
0 373.63
The alternative to hedging in this example is to liquidate the stock portfolio and buy three-month T-bills.
Such an action would have produced a 1.5 percent return, given our assumption that the T-bill rate is 1.5
percent. The riskless rate and abnormal performance components of the hedge portfolio return in this
illustration are therefore 1.5 percent and 3.9 percent, respectively. In other words, the 3.9 percent return was
the abnormal or extra-market rate of return arising from the fact that the portfolio of “winners” outperformed
the market on a risk-adjusted basis.
The hedged portfolio would also have earned 1.5 percent if the stock portfolio had declined exactly
according to its beta of 1.20, without an abnormal return. In that case, the return would have been
where Rm is the return on the stock index including the dividends, or -0.19. That implies a value for the
stock portfolio, including dividends, of $38,450,000, instead of the value of $40,400,000 shown in Table 7.8.
The values in Table 7.8 for the cash index and the futures market would remain the same. The dollar gain on
the hedged portfolio becomes $750,000, and the hedged return becomes 1.5 percent, exactly the same as the
riskless rate.
It is worth noting that the hedged stock portfolio has basis risk because the portfolio’s return is not
perfectly correlated with the index futures return. If, for example, the stock portfolio had a negative abnormal
return, the hedged portfolio would have earned less than the riskless rate.
7.6 SUMMARY
In this chapter, stock index futures contract index and the composition of stock indexes underlying futures
contracts are described. The cost-of-carry relation for stock indexes is derived, and the role of index arbitrage
in maintaining the link between stock index futures and cash prices is explained. Evidence on the short-run
behavior of the returns of index futures and of the cash index is presented. Finally, the use of stock index
futures to hedge the market risk in a stock or a portfolio of stocks is illustrated in detail.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 95/358
8.9 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 96/358
Like stock index futures contracts, interest rate futures have been extremely successful financial futures
contracts. Interest rate futures provide a means of trading future loan commitments and are important and
useful risk management tools for fixed-income portfolio managers. This chapter begins with a description of
interest rate futures markets and the specifications of the contracts traded in those markets. Section 2
analyzes the instruments underlying interest rate futures contracts. In section 3, a key measure of interest rate
risk, duration, is explained. In section 4, the term structure of interest rates is examined. We focus on the
effect of the term structure on the value of coupon and non-coupon bonds. Spot rates, forward rates, and
yield to maturity are defined. Examples illustrate most of the important concepts. Sections 5 and 6 describe,
in detail, the interest rate futures contracts corresponding to short-term and long-term interest rates. The most
active short-term interest rate futures contracts in the U.S. are the Chicago Mercantile Exchange’s Treasury
bill and Eurodollar contracts, while the most active long-ten-n interest rate futures contract is the Chicago
Board of Trade’s Treasury bond contract. Contracts in the intermediate-term range, the CBT’s five-year and
ten-year Treasury note futures, are also discussed. Section 7 describes the cost-of-carry equilibrium in the T-
bond futures market. Section 8 contains three illustrations of the use of interest rate futures contracts, and
section 9 contains a summary.
TABLE 8.1 Interest rate futures contracts specifications (most active contracts in U.S. markets).
Units/Minimum
Security Trading Contract Last Day of
Price Deliverable Grade62
(Exchange) Hours Months61 Trading
Fluctuation
Nominal 8% coupon
Seven business days
T-Bond 8:00-2:00 $100,000/ with 15 years to
3,6,9,12 prior to last business
(CBT) (CST) 1/32 ($31.25) maturity or first call
day of month
date
10-yr. Seven days prior to Nominal 8% coupon
8:00-2:00 $100,000/
T-Note 3,6,9,12 last business day of with 6.5 to 10 years to
(CST) 1/32 ($31.25)
(CBT) month business maturity
Seven business days Any of the four most
5-yr. T-Note 8:00-2:00 $100,000/
3,6,9,12 prior to last business recently auctioned 5-
(CBT) (CST) 1/32 ($31.25)
day of month year Treasury notes
Business day
91-day T-Bill 7:20-2:00 $1,000,000/ Any of the three T-bills
3,6,9,12 preceding issue date
(CME) (CST) 0.01 ($25.00) 91-days from maturity
of new 91-day T-bill
3,6,9,12
Eurodollar Second London
7:20-2:00 plus $1,000,000/
Time Deposit business day before Cash settled
(CST) current 0.01 ($25.00)
(CME) third Wednesday
month
61
The notation used in this column corresponds to the month of the calendar year (e.g., 1 is January, 2 is February, and so on).
62
All interest rate futures contracts other than the Eurodollar contract call for delivery of the underlying security.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 97/358
Unlike stock index futures contracts, most interest rate futures contracts call for the delivery of the
underlying interest rate instrument. The CBT’s T-bond futures contract, for example, requires the delivery of
a nominal eight-percent coupon, $100,000 face value, U.S. Treasury bond. Delivery may take place at any
time during the delivery month, at the discretion of the short. The last day of trading of the futures contract is
the eighth-to-last business day of the contract month.
Table 8.2 shows prices of these contracts as of the close of trading on November 13, 1991. The open
interest figures in Table 8.2 show that the T-bill, Eurodollar, and T-bond futures contracts are the most active
interest rate futures contracts presently trading. These contracts are the focus of the dominant part of this
chapter.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 98/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 99/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 100/358
The T-bill and Eurodollar futures contracts are on short-term debt instruments and are discussed first. The T-
bond (and T-note) futures are written on long-term interest rates and are discussed second. Prior to beginning
either discussion, however, it is useful to review the pricing mechanics of the fixed-income securities that
underlie these futures contracts.
Fn
Bd (8.1)
(1 y ) n
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 101/358
T-bill quotes are unusual in that their prices are not reported directly. Table 8.3 contains market
information about the T-bills that were active on November 13, 1991. Note that, in this table, bid and ask
“discounts” appear. These are bank discount price quotations. The definition of a bank discount is
where n is the number of days to maturity of the bill, and Bd is the bill’s price expressed as a percentage of
par. (Traditionally, bankers have assumed a 360-day year.) To compute the price of the T-bill, simply
rearrange equation (8.2) to isolate Bd , that is,
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 102/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 103/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 104/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 105/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 106/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 107/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 108/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 109/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 110/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 111/358
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones &
Company, Inc. All Rights Reserved Worldwide.
To illustrate the bank discount convention, consider the T-bill in Table 8.3 that matures on February 13,
1992. If this bill is purchased, it would be at the ask price,63 as reflected by the ask discount of 4.61. The
number of days to maturity of this bill is 92. Therefore, the ask price of the bill is
percent of par. If the par value is $10,000, the price of the T-bill is $9,882.20. To compare the rate of return
on the T-bill with the rate of return on other instruments, the effective annual rate of return is often
computed. If the price of a 92-day T-bill is 98.822, the effective annual rate of return compounded on a daily
basis64 is
Eurodollar certificates of deposit (CD’s) are also discount bonds. Eurodollar deposits are U.S. dollar deposits
in any foreign bank, although Eurodollar deposits are most typically thought of as being U.S. dollar deposits
in London. The rate on these CD’s is referred to as the London Interbank Offer Rate (LIBOR). The usual
denomination of Eurodollar CD’s is $1,000,000. The usual maturities are in the three- to six-month range,
however, maturities as long as five years are not uncommon. Simple interest on Eurodollar deposits is
calculated for the actual number of days on a 360-day year basis and is paid at maturity on deposits with
terms of less than one year. For longer-term deposits, interest is paid annually.65
Sample Eurodollar rates are contained in Table 8.4. The rates reported are averages of the rates
quoted by five major banks in London. The three-month rate, for example, is reported to be 51/8 percent.
That implies that a $1,000,000 Eurodollar deposit on November 13, 1991, will be worth $1,000,000
63
The bid/ask spread is the cost of immediate trade execution. A market-maker stands ready to buy immediately at her bid price
and sell immediately at her ask price. The bid/ask spread is her compensation for providing market liquidity.
64
The effective annual return compounded continuously is r ln (100/98.822)(365/92) 4.70% .
65
For more information on Eurodollar certificates of deposit, see Stigum (1990).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 112/358
+1,000,000(0.05125)(90/360), or $1,012,812.50, on February 11, 1992.66 The effective annual rate of return
on this investment may be computed as follows:
Note that although this Eurodollar deposit has approximately the same maturity as the T-bill in the previous
example, the effective annual rate of return is 49 basis points higher.67 This reflects the higher default risk of
these deposits vis-à-vis U.S. government securities.
T-bills and CD’s are short-term discount bonds. Table 8.3 also contains prices of Treasury strips, which are
long-term discount bonds. Strips get their name from the fact that, for each year, coupons of several
government coupon bonds are stripped” from the principal and combined to create a discount bond that
makes only one payment. Table 8.3 provides the prices and yields to maturity for these bonds.
PRIME RATE: 71/2%. The base rate on corporate loans at U.S. money market banks.
FEDERAL FUNDS: 51/e% high, 2% low, 2% near closing bid, 21/2% offered. Reserves traded among
commercial banks for overnight use In amounts of $1 million or more. Source: Babcock Fulton Prebon
(U.S.A.) Inc.
DISCOUNT RATE: 4.50%. The charge on loans to depository Institutions by the Federal Reserve Banks.
CALL MONEY-. 63/d% to 7%. The charge on loans to brokers on stock exchange collateral.
COMMERCIAL PAPER placed directly by General Electric Capital Corp.: 4.85% 15 to 36 days; 4.70% 37
to 59 days; 4.93% 60 to 89 days; 4.90% 90 to 149 days; 4.87% 150 to 179 days; 4.70% 180 to 189 days;
4.87% 190 to 270 days. Commercial Paper Placed directly by General Motors Acceptance Corp.: 4.85% 30
to 44 days;- 4.80% 45 to 59 days; 5% 60 to 270 days.
COMMERCIAL PAPER: High-grade unsecured notes sold through dealers by mayor corporations In
multiples of $1,000: 5% 30 days; 5.10% 60 days; 5.02% 90 days.
CERTIFICATES OF DEPOSIT: 4.48% one month; 4.61% two months; 4.60% three months; 4.62% six
months; 4.98% one year. Average of too rates paid by mayor New York banks on Primary new Issues of
negotiable C.D.s, usually on amounts of $1 million and more. The minimum unit is $100,000. Typical rates
in the secondary market: 4.90% one month; 5% three months; 5% six months.
BANKERS ACCEPTANCES: 4.80% 30 days; 5.03% 60 days; 4.91% 90 days; 4.87% 120 days; 4.85% 150
days; 4.81% 180 days. Negotiable, bank-backed business credit instruments typically financing an Import
order.
LONDON LATE EURODOLLARS: 5% - 47/8% one month; 51/2% - 51/4% two months; 5 3/16% - 5
1/16% three months; 5 3/16% - 5 1/16% four months; 5 3/16% - 5 1/16% five months; 51/8% - 5% six
months.
LONDON INTERBANK OFFERED RATES (LIBOR): 4 15/16% one month; 51/8% three months; 51/8%
six months; 51/4% one year. The average of Interbank offered rates for dollar deposits In the London market
based on quotations at five mayor banks. Effective rate for contracts entered into two days from date
appearing at top of this column.
FOREIGN PRIME RATES: Canada 8.50%; Germany 11.50%; Japan 7%; Switzerland 10%; Britain 10.50%.
These rate Indications aren’t directly comparable; lending practices vary widely by location.
66
The number of days from November 13, 1991, to February 11, 1992, is 90.
67
A basis point is 1/100 of one percent.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 113/358
TREASURY BILLS: Results of the Tuesday, November 12, 1991, auction of short-term U.S. government
bills, sold at a discount from face value in units of $10,000 to $1 million: 4.64% 13 weeks; 4.71% 26 weeks.
FEDERAL HOME LOAN MORTGAGE CORP. (Freddie Mac): Posted Yields on 30-year mortgage
commitments. Delivery within 30 days 8.54%, 60 days 8.60%, standard conventional fixed-rate mortgages;
6%, 2% rate capped one-year adjustable rate mortgages. Source: Telerate Systems Inc.
FEDERAL NATIONAL MORTGAGE ASSOCIATION (Fannie Mae): Posted yields on 30-year mortgage
commitments for delivery within 30 days (Priced at par). 8.47%, standard conventional fixed rate-mortgages;
6.10%, 6/2 rate capped one-year adjustable rate mortgages. Source: Telerate Systems Inc.
MERRILL LYNCH READY ASSETS TRUST: 4.90%. Annualized average rate of return after expenses for
the past 30 days; not a forecast of future returns.
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
Coupon Bonds
The price of a coupon-bearing bond, B , is computed by summing the present values of the fixed periodic
coupon payments, Ct , and the present value of the terminal face value, Fn , that is,
Ct Fn
B t 1
n
(8.4)
(1 y ) t
(1 y ) n
In general, the coupon payments are the same fixed amount each period, so the bond price may also be
written as the present value of an annuity plus the present value of the final amount:
B (C / y )[1 (1 y ) n ] F (1 y ) n . (8.5)
The Treasury bonds and notes in Table 8.3 are coupon-bearing bonds. Treasury notes are issued with
maturities of two to ten years and Treasury bonds with maturities longer than ten years. The denominations
of bonds and notes range from $1,000 to $1 million. Their prices are quoted as a percentage of par, so a
reported price of 96:00 for a $100,000 face-value bond is actually 96 percent of $100,000 or $96,000. In
addition, the decimal part of a T-bond or T-note price is the number of 32nds, so a reported price of 94:8 is
actually 948/32, or 94.25 in decimal form.
Finally, all T-bonds and T-notes have semiannual coupon payments. The “9’s of November 2018” in
Table 8.3, for example, pay coupon interest at a rate of 4.5 percent of par in May and November of each
year. The last coupon is paid with the repayment of the face value in November 201868. The fact that
Treasury bonds pay semiannual coupons implies that the yield to maturity, y , in (8.4) and (8.5) is an
effective rate over a six-month period. To annualize this rate, the effective annual yield to maturity, y A , may
be computed as
y A (1 y ) 2 1 (8.6)
While (8.6) is technically correct, some people prefer to use the bond equivalent yield, which is simply
2 y , to measure of the expected annual rate of return on the bond.69
One last convention about Treasury bonds and notes must be discussed. The bond price reported in
the financial press (i.e., the quoted bond price) excludes the accrued interest of the current coupon period.
68
Fabozzi and Fabozzi (1989, pp. 83-84) provide a summary of the various types of bill, note, and bond issues of the U.S.
government.
69
Yet others rely on coupon yield, which is simply the annual coupon amount divided by the bond price.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 114/358
The price that you would pay for the bond if you decided to buy it would be the quoted price plus the coupon
interest that has accrued on a straight-line basis during the current coupon period. That is, accrued interest
equals the proportion of the current coupon period that has elapsed times the amount of the coupon,
To illustrate, consider the 9s of November 2018, whose price is reported in Table 8.3. The bond is reported
to have a bid price of 111:21 and an ask price of 111:23. Treasury bonds pay coupon interest every six
months and do so on the 15th of the month. For the 9s of November 2018, coupon interest of 4.5 percent of
par is paid on May 15th and November 15th. On November 13, 1991, the number of days since the last
coupon payment is 182. The total number of days from May 15, 1991, to November 15, 1991, is 184. The
accrued interest on this bond as of November 13 is, therefore, 4.5(182/184) or 4.45. If this bond were bought
at the ask price, we would pay the reported ask price of 11123/32 plus accrued interest of 4.45, or a total
amount 116.17.
To verify that 116.17 is actually the ask price of the 9s of November 2018, compute the present value
of the promised coupons and face value as of November 13, 1991, using the reported bond equivalent yield
of 7.94 percent. We cannot apply the bond valuation formulas (8.4) and (8.5) directly because those formulas
assume that the next coupon payment is in exactly six months. For the 9s of November 2018, we are part
way through the coupon period, so the bond valuation formula needs to be modified:
n Ct Fn
B (1 y ) p t 1 , (8.8)
(1 y ) (1 y ) n
t
where p is the ratio of the number of days elapsed in the current coupon period to the total number of days
during the current coupon period. The term in bracket represents the present value, just after the last coupon,
of all future payments to the bondholder. This amount is then capitalized forward p coupon periods to the
current date. Substituting the example parameters,
n Ct Fn 55 4.5 100
B (1 y ) p t 1 n
(1.0397) p t 1 116.17
(1 y ) (1 y )
t
(1.0397) (1.0397) n
t
Subtracting the accrued interest yields the quoted ask price of the bond reported in Table 8.3.
Finally, some Treasury bonds do not have a fixed maturity date. The issues denoted by the hyphenated
maturity date in Table 8.3 are callable bonds, which the Treasury has the right to “call,” or redeem, at any
3/4
time during a prespecified period in the future. The 11 of August 2009-14, for example, are callable bonds
that may be redeemed during the period November 2009 through 2014. Callable bonds may be delivered on
certain futures contracts, as long as they satisfy some minimum amount of time before the first call date.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 115/358
the risk of the default-free fixed income securities (such as U.S. Treasury securities) that these institutions
hold.70
Duration
One measure of a fixed income security’s interest rate risk is its duration. Duration specifies the sensitivity of
the bond price to movements in yield. A specific formula for computing duration may be obtained by taking
the derivative of B with respect to y in equation (8.4). First, for the sake of mathematical convenience,
rewrite (8.4)
B t 1 Ct (1 y ) t . (8.9)
n
Under this specification, we let Ct C for t 1, , n 1 and Cn C Fn for t n . Now, take the derivative
of B with respect to y ,
dB t 1
t 1 tCt (1 y ) .
n
(8.10)
dy
Multiply (8.10) by (1 y ) ,
dB t
t 1 tCt (1 y ) .
n
(1 y ) (8.11)
dy
dB / B C (1 y ) t
t 1 t t t 1 twt D , (8.12)
n n
dy /(1 y ) B
D t 1 twt . (8.13)
n
Equations (8.12) and (8.13) tell us two important things about duration. First, as the left-hand side of (8.12)
indicates, minus duration, D , can be interpreted as the percentage change in the bond price, dB / B ,
induced by a change in the bond’s yield to maturity, dy , scaled by 1 /(1 y ) . Second, duration is the
weighted average time to maturity of a bond, where the weights, wt , are the present values of the payments
in each period.71 For coupon bonds, duration is less than time to maturity because some of the bond
payments –the coupons– are received in years prior to maturity of the bond. For non-coupon bonds, duration
equals time to maturity.
The scale factor 1 /(1 y ) on the left-hand side of (8.12) is cumbersome to account for. Most fixed
income portfolio managers prefer to use a measure of duration that is simply the percentage change in bond
price, dB / B , induced by a change in yield, dy . To find this expression, divide (8.12) by (1 y ) ,
70
Default risk is also an important consideration in fixed-income security risk management. Recall that on Chapter 4, we showed
how to hedge both the interest rate and default risk exposure of Mobil Oil bonds using T-bond and stock index futures contracts.
t
n C (1 yt ) B
Note that by virtue of equation (8.9), the weights, wt , sum to one, that is, t 1 wt t 1 t
n
71
1
B B
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 116/358
dB / B C (1 y ) t
t 1 t t t 1 twt D . (8.14)
n n
dy /(1 y ) B
is called modified duration and is more commonly used in interest risk management strategies. It is simply
the duration, as defined in (8.13), divided by (1 y ) . We use modified duration in a hedging application later
in this chapter. We now turn to a numerical example that uses the duration formulas (8.13) and (8.14).
EXAMPLE 8.1
What are the duration and the modified duration of an 8 percent, ten-year, $1000 bond, assuming annual
coupon payments and a required yield to maturity of 7 percent? From equation (8.4), we know the bond price
80 1,000
is B t 1
10
$1,070.24
t
(1.07) (1.07)10
This present value consists of the present value of 10 payments. In tabular form,
t Ct (1 y ) t Ct (1 y ) t wt twt
1 80.00 0.9346 74.77 0.0699 0.0699
2 80.00 0.8734 69.88 0.0653 0.1306
3 80.00 0.8163 65.30 0.0610 0.1831
4 80.00 0.7629 61.03 0.0570 0.2281
5 80.00 0.7130 57.04 0.0533 0.2665
6 80.00 0.6663 53.31 0.0498 0.2989
7 80.00 0.6227 49.82 0.0466 0.3259
8 80.00 0.5820 46.56 0.0435 0.3480
9 80.00 0.5439 43.51 0.0407 0.3659
10 1080.00 0.5083 549.02 0.5130 5.1299
Total 1,070.24 1.0000 7.3466
The duration of the bond is 7.3466, and the modified duration is 7.3466/1.07 = 6.8660. The modified
duration figure computed in the example predicts that if interest rates increase by 100 basis points, the bond
price will change by
Conversely, a decrease in yield of 100 basis points implies a 6.8660% increase in bond price.
The formula for the duration of a bond shows that duration –price sensitivity– depends on the
maturity of the bond, on the coupon level, and on the yield to maturity. First, the greater the maturity, the
greater the duration, holding constant other characteristics of the bond. Second, the larger a bond’s coupon,
the smaller the duration. Coupon payments cause weight to be put on the early years in the duration formula.
In the case of a zero-coupon bond, duration equals maturity. Third, duration decreases with increases in the
yield to maturity. This is because an increase in the yield has a greater effect on the present value of a distant
coupon than on the present value of a nearby coupon.
Convexity
Modified duration is only an approximation of the percentage change in bond price for a given change in
yield. In fact, it is accurate only for infinitesimal shifts in yield. To assess the degree of error that may be
introduced, reconsider the results of Example 8. 1. The level of modified duration predicts that the bond
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 117/358
price will increase to $1,070.24(l.068660) = $1,143.72 if the yield drops to 6 percent. By using the bond
valuation formula (8.4), however, we know that the bond value is exactly $1,147.20 at a yield of 6 percent.
The difference between these prices is attributable to the fact that bond price is a nonlinear function of yield
to maturity.
Figure 8.1 illustrates the approximation error in this example. At a yield of 7 percent, the bond’s price
is $1,070.24. Modified duration depends on the derivative, dB / dy , evaluated at 7 percent. The derivative is
depicted in the figure by the straight-line tangent to the bond price curve at 7 percent. To estimate the change
in bond price due to a 100 basis point drop in yield, we draw a vertical line from 6 percent on the horizontal
axis to the straight line depicting the derivative, and then draw across horizontally to the vertical axis. The
estimated value of the bond based on modified duration is $1,143.72.. If the vertical line is continued upward
to the bond price curve, we find the exact value of the bond at a 6-percent yield, $1,147.20. The pricing
error, $3.48, is attributable to the failure of modified duration to account for the convex nature of the bond
pricing function.
We can achieve greater precision in measuring the bond’s responsiveness to yield shifts by also
accounting for the bond’s convexity. To understand convexity, expand the bond price as represented by (8.9)
into a Taylor series:
dB 1 d 2B
dB dy 2
(dy ) 2 . (8.15)
dy 2 dy
$1,147.20
$1,143.72
$1,070.24
6% 7% Yield to maturity, y
The error term, , recognizes the fact that we have used only the first two terms of the Taylor series
expansion. Hence, our approximation for bond price changes will remain only approximate. Now, we drop
the error term and divide both sides by B ,
dB dB / B 1 d 2B 1
dy 2
(dy ) 2 . (8.16)
B dy 2 dy B
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 118/358
1 d 2B 1
convexity , (8.17)
2 dy 2 B
and rewrite (8.16), the percentage change in bond price, using (8.14) and (8.17), as
dB
Dm dy Convexity(dy ) 2 . (8.18)
B
The value of the second derivative of the bond price with respect to a change in yield is
d 2B Ct
t 1 t (t 1)
n
t 2
. (8.19)
(1 y )
2
dy
EXAMPLE 8.2
What is the convexity of an 8 percent, ten-year, $1000 bond, assuming annual coupon payments and a
required yield to maturity of 7 percent? From Example (8. 1), we know the bond price is $1,070.24. The
components of convexity (8.19) are
Ct t (t 1)Ct
t t (t 1)
(1 y ) t 2 (1 y ) t 2
1 2 65.3038 130.61
2 6 61.0316 366.19
3 12 57.0389 684.47
4 20 53.3074 1,066.15
5 30 49.8200 1,494.60
6 42 46.5607 1,955.55
7 56 43.5147 2,436.82
8 72 40.6679 2,928.09
9 90 38.0074 3,420.67
10 110 479.5329 52,748.62
Total 67,1231.77
The value of d 2 B / dy 2 for this bond is 67,231.77, so the convexity is 1/2 x 67,231.77 1/1,070.24 31.4098 .
To complete our illustration, we now use convexity in conjunction with modified duration to arrive at
a more accurate estimate of the percentage change in bond price attributable to a 100 basis point fall in yield.
dB
Dm dy Convexity(dy ) 2 6.8662 (0.01) 31.4098 (0.01) 2 7.180% .
B
In other words, based upon modified duration and convexity, the bond price is expected to rise to
$1,147.08 if yield falls by 100 basis points. Note that the error in prediction has been reduced from $3.48 to
$0.12 by accounting for convexity.
All bonds with fixed payment schedules plot as convex curves in y , as shown in Figure 8.1, although
different bonds have different degrees of convexity. Mortgage backed securities (MBS’s), however,
sometimes exhibit “negative convexity.” That is, they plot as curves that are concave to the origin. This is
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 119/358
because their payment schedules are uncertain. The underlying mortgages have the option to prepay before
maturity, and they will choose to prepay just when it is most inconvenient for the holder of the MBS. When
interest rates fall, mortgages are prepaid, so the duration and price sensitivity of the MBS are reduced.
Consequently, while fixed-payment bonds increase in price, the price of MBS’s increases little and
may even decline. When interest rates rise, mortgage prepayments slow, resulting in duration and price,
sensitivity increases for the MBS. Consequently, the price decline resulting from an increase in interest rates
is relatively greater for a MBS than for a fixed-payment bond.
Yield Curve
The simplest way to consider the relation between expected bond return and time to maturity is to plot the
yield to maturity of U.S. Treasury obligations versus the term to maturity (making sure that the bonds have
the same default risk, no imbedded options such as callability, and no differential tax privileges). This
relation is referred to as the yield curve. The yield curve is usually plotted using Treasury bonds and notes.
The yield to maturity differs by maturity because expected future interest rates are different for different
maturities and because risk premia differ by maturity.
The yield to maturity of a coupon-bearing bond can be a misleading reflection of its expected rate of
return, however. To see this, consider the following situation. Suppose there exist two discount Treasury
bonds, one with a one-year maturity and a price of 90.91, and the other with a two-year maturity and a price
of 81.16. Assuming each of these bonds is redeemed at a par value of 100 at their respective maturities, the
expected yield on the one-year bond is 100/90.91 - 1 10.00% , and the expected yield on the two-year bond
is (100/81 - 16)1/2 - 1 11.00% .
Note that if the Treasury decided to issue other one- and two-year discount bonds, they must have the
same prices as the existing issues, otherwise, costless arbitrage profits could be earned.
Now, suppose that the Treasury also has a two-year, 12-percent coupon bond. The 12-percent issue
pays 12 at the end of one year, and 112 at the end of two years. The price of this coupon-bearing Treasury
bond has to be
12 112
B 101.81 ,
1.10 (1.11) 2
otherwise, costless arbitrage profits are possible.72 But, if the two-year coupon-bearing bond has a price of
101.81, its yield to maturity is 10.94 percent, that is, the solution to
12 112
101.81
(1 y ) (1 y ) 2
On the surface, it might appear that the two-year coupon bond expected yield to maturity of 10.94 percent
conflicts with the two-year discount bond expected yield to maturity of 11 percent. However, arbitrage
profits are not possible. The discrepancy between the rates arises because the two-year coupon bond has an
72
Note that buying 0.12 units of the one-year discount bond and 1.12 units of the two-year discount bond produces a cash flow
stream exactly the same as the two-year coupon-bearing bond at a cost of 0.12(90.91) + 1.12(81.16) = 101.81.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 120/358
“actual” term to maturity of less than two years. If you buy the coupon bond, you will receive 12 at the end
of one year and 112 at the end of two years; so, in essence, you have a portfolio of two discount bonds. The
first discount bond is worth
12
Bd1 10.91
1.10
112
Bd 2 90.90
(1.10) 2
The average term to maturity of the bond portfolio (or the coupon-bearing bond) is, therefore,
10.91 90.90
Average term to maturity 1 2 1.893 years .
101.81 101.81
The 10.94-percent yield is not an expected return on a bond with two years to maturity, but rather on a bond
with a maturity of approximately 1.893 years.
Two principles underlie this discussion. First, to value coupon-bearing bonds precisely, all cash flows
of the bond should not be discounted at the same rate. Instead, each cash flow should be discounted at the
zero-coupon or discount bond rate, rt , which coincides with the timing of the cash flow,
n Ct Fn
B t 1
(1 rt ) (1 rn ) n
t
The rate rt is called the spot rate of interest on a t - period loan. Second, to estimate the relation between
expected bond returns and time to maturity, only zero-coupon or spot rates of interest should be used. The
yields to maturity on coupon bonds should not be used because the term to maturity of the bond overstates its
true economic life. The economic life of a coupon bond is a weighted average life of its constituent discount
bonds, that is,
1 n Ct Fn
Average term to maturity t 1 t n .
n
(8.21)
(1 rt ) (1 rn )
t
B
Note that this weighted average term to maturity is the same as the bond’s duration defined in (8.12).73 In
practice, the relation between zero-coupon or spot rates of interest and time to maturity is referred to as the
zero-coupon yield curve. In this section, we refer to the relation as the term structure of interest rates. The
term structure can be read directly from the yields to maturity of Treasury strips. For example, Table 8.3
shows that the term structure was upward sloping on November 13, 1991. The one-year, two-year, five-year,
and ten-year spot rates were 5.05, 5.68, 6.84, and 7.78, respectively.
The shape of the term structure of spot rates affects the relative values of bonds with different
coupons. We have already illustrated this point when we showed that a two-year coupon bond would yield
10.94 percent to maturity while a two-year discount bond yielded 11 percent. In general, for bonds of the
73
Duration and maturity are the same only for a discount bond because the coupon terms, Ct , in the summation of (8.21) are zero.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 121/358
same maturity, when the term structure is upward sloping, the yield to maturity decreases as the coupon
payment increases; when the term structure is downward sloping, the yield to maturity increases with a larger
coupon. This result can be explained in two ways. First, a coupon bond has a duration that is less than its
maturity. Its yield to maturity corresponds to its, duration, not its maturity. Second, when the term structure
is upward sloping, investors prefer coupon bonds because it is expected that the coupon payments can be
invested at higher future rates. They, therefore, bid up prices of coupon bonds and lower their yield to
maturity. Correspondingly, when the term structure is downward sloping, investors prefer low coupon bonds
because it is expected that the coupons can be invested only at lower future rates. Only when the term
structure of interest rates is flat do all bonds have the same yield to maturity.
(1 rnt ) nt
(1 n f t ) t . (8.22)
(1 rn ) n
(1 r31 ) 31
(1 3 f1 )1 .
(1 r3 ) 3
EXAMPLE 8.3
Suppose the one-year spot rate of interest is 10 percent and the two-year spot rate is II percent. What is the
implied one-year forward rate in one year?
Substituting the parameters of the example into equation (8.22), we find that
(1 r2 ) 2
(1 1 f1 )
1
1 f1 12.01% .
(1 r1 )1
EXAMPLE 8.4
Using the values reported in Table 8.3, find the 91-day forward rate implied by selling the December 19,
1991 T-bill and buying the March 19, 1992 T-bill. The effective annual rate of return of the December 19th
365 / 360
100
T-bill implied is r 1 4.562 .
100 4.39(36 / 360)
365 / 127
100
r 1 4.858
100 4.64(127 / 360)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 122/358
Given these two spot rates of interest, the implied forward rate of interest on a 91-day T-bill in 36 days is the
solution to
(1.04858)127 / 365
(1 36 f 91 )
91/ 365
36 f 91 4.975%
(1.04562)36 / 365
The implied forward rate of interest is 4.975 percent.
365 / 91
100
1 4.866% .
98.822
It is interesting to note that the implied forward rate of interest from the futures contract, 4.868 percent, is
less than the 4.975 percent rate implied by the T-bills in Example 8.4. Both rates apply to the 91-day period
beginning on December 19, 1991. If one could borrow at 4.868 percent and lend at 4.975 percent, an
arbitrage profit could be earned. In effect, this can be done by the following arbitrage transactions: (a) sell
the December 19, 1991 T-bill for $9,956.10; (b) buy the March 19, 1992 T-bill for $9,836.31; and (c) sell the
December 1991 T-bill futures contract at $9,882.20. Transactions (a) and (b) have the net effect of lending
money over the period December 19, 1991 to March 19, 1992; transaction (c) commits the investor to
borrowing money for the same period. The borrowing and lending, however, may never occur because the
transactions can be closed out on December 19, 1991. On that date, the arbitrageur covers the short position
in the maturing December 19 T-bill by paying $10,000, thereby incurring a cost of $43.90. The March 19 T-
bill purchased on November 13 now has 91 days to maturity and can be delivered against the futures contract
for $9,882.20, a net gain of $45-89. Finally, the net proceeds from the T-bill transactions on November 13,
1991, $119.79, have earned interest at, say, 5 percent and, after 36 days, are now worth $120.37. The
arbitrage profit realized on December 19, therefore, totals
The Eurodollar futures contract is a commitment to transact a $ 1,000,000, three-month Eurodollar deposit.
Delivery never takes place since a Eurodollar futures contract is cash settled. Cash settlement occurs on the
second London business day before the third Wednesday of the contract month. The settlement price of the
Eurodollar futures contract on the expiration day is computed in such a way as to minimize the variation in
the quoted Eurodollar deposit rates. During the last day of trading, a random sample of approximately twelve
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 123/358
rates are taken from the twenty-plus approved banks in the London Euro-dollar market. The rates are then
ranked from highest to lowest, and the highest and lowest rates are discarded. The remaining ten rates are
averaged, and the average rate is subtracted from 100 to determine the settlement price.
As noted in the previous paragraph, quoted Eurodollar futures prices are actually index values, that is,
the value reported in the financial press is 100 less the Eurodollar interest rate. Thus, if we buy the March
1992 Eurodollar futures at the price reported in Table 8.2, 94.94, the implicit agreement that we are entering
into is to buy a $1,000,000 three-month Eurodollar certificate of deposit on March 16 1992 (the second
London business day before the third Wednesday of the future’s contract month), where the stated interest
rate on the deposit is 100 - 94.94, or 5.06 percent. The effective three-month forward interest rate on such a
deposit is, therefore,
365 / 92
100 5.06(92 / 360)
r 1 5.23% .
100
Note that the three-month interval from March 16, 1992, through June 16, 1992, has 92 days.
8 100
B t 1
15
100.00 .
(1.08) (1.08)15
t
Now, consider the price of a 12-percent, fifteen-year bond with fifteen years to maturity and the same 8-
percent yield. The bond price is
12 100
B t 1
15
134.24 .
(1.08) (1.08)15
t
Note that the only difference between the two bonds is that the second bond has higher coupon payments.
Owning the 12-percent coupon bond is like owning 1.3424 8-percent bonds. Since the futures price is based
74
For an interesting analysis of why this contract supplanted the earlier GNMA contract, see Johnston and McConnell (1989).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 124/358
on an 8-percent coupon bond, the futures price is multiplied by a conversion factor of 1.3424 to compute the
amount paid (delivery price) by the long to the short if the short delivers the 12-percent coupon issue.
The actual formula for computing the conversion factor is slightly more complex than what is demonstrated
in the above example because coupons are paid on a semi-annual basis, and, in general, the next coupon
payment is made in less than six months (i.e., we are part of the way through the current coupon period). The
actual formula for the conversion factor, CF , is
C C C (6 X )
CF (1 y / 2) X / 6 [1 (1 y / 2) 2 n ] 1(1 y / 2) 2 n , (8.23)
2 y 2 6
where C is the annual coupon rate of the bond in decimal form, y 0.08 , n is the number of whole years to
first call, if the bond is callable, or the number of years to maturity, and X is the number of months that the
maturity exceeds n , rounded down to the nearest quarter (e.g., X 0, 3, 6, 9 ). Note that if X 0, 3, 6 , the
formula (8.23) is used directly. If X 9 set 2n 2n 1, X 3 , and calculate as above.75 Computer programs
are available to perform the computation (8.23). The values in Table 8.5, for example, were generated using
a program called OPTVAL. Alternatively, the CBT and others publish and distribute conversion factor
tables.
TABLE 8.5 Conversion factors for the U.S. Treasury Bonds eligible for delivery on the CBT’s T-bond
futures contract. These factors convert different coupon issues to yield 8 percent.
1/8 1/4 3/8 1/2 5/8 3/4 7/8
Years-Months 9% 9 % 9 % 9 % 9 % 9 % 9 % 9 %
C
75
Note that if X 0 the formula (8.23) reduces to CF 1 (1 y / 2)2n (1 y / 2)2n
y
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 125/358
TABLE 8.5 Cont. Conversion factors for the U.S. Treasury Bonds eligible for delivery on the CBT’s T-
bond futures contract. These factors convert different coupon issues to yield 8 percent.
1/8 1/4 3/8 1/2 5/8 3/4 7/8
Years-Months 9% 9 % 9 % 9 % 9 % 9 % 9 % 9 %
To illustrate the use of the conversion factor system, suppose that we are considering delivery of the 9s of
November 2018 on the March 1992 T-bond futures contract. This bond is eligible for delivery because on
March 1, 1992, it has more than fifteen years to maturity. Specifically, on March 1, 1992, the 9s of
November 2018 have 26.50 years to maturity (rounded down to the nearest quarter). Using Table 8.5, the
conversion factor of this bond is 1.1094. In other words, in place of delivering the hypothetical 8-percent,
fifteen-year bond on the March 1992 futures contract, we can deliver the 9s of November 2018, but the buyer
is going to have to pay 1.1094 times the prevailing futures price.
On the delivery date, the seller of the T-bond futures delivers an eligible T-bond to the buyer of the
T-bond futures contract. In return, the buyer must pay the invoice price to the bond seller. The amount of the
invoice price will be the sum of the futures price times the conversion factor of the delivered bond and the
accrued interest on the delivered bond. For example, suppose that on March 15, 1992, the March 1992
futures contract is priced at 96-18. Like the underlying bonds, the decimal part of the price is the number of
32nds, so the futures price is 96.5625. If we sell the futures and promptly deliver the 9s of November 2018 to
the futures contract buyer, the invoice price paid by the buyer equals .965625 (the futures price in decimal
form) times 100,000 (the denomination of the futures contract) times 1. 1094 (the conversion factor of the 9s
of November 2018 as of March 1, 1992), or $107,126.44, plus the accrued interest on the 9s of November
2018 as of March 15, 1992, $2,991.76 [i.e., 0.045(121/182)(100,000)]. The total invoice price is
Cheapest to Deliver
In principle, the system of conversion factors is intended to make the short indifferent about which bond he
delivers. This means that if we are at time T –the end of the futures contract life– the profits from selling the
futures and buying and delivering bond i , computed as
i ,T FT (CFi ) AI i ,T Bi ,T AI i ,T FT (CFi ) Bi ,T ,
where FT (CFi ) AI i ,T , is the invoice price received from delivering bond i and Bi ,T AI i ,T , is the price paid
for the purchase of bond i , should equal zero.
In practice, however, one of the eligible delivery bonds is “cheapest to deliver” because the system of
conversion factors is not exact. Each bond will have a different value of i,T , the bond with the highest i,T
is the cheapest to deliver.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 126/358
Its value of i ,T , however, will be equal to zero. If it were positive, costless arbitrage profits could be
earned by the short at the expense of the long. If it were negative, costless arbitrage profits could be earned
by the long at the expense of the short.
The computed profits for all other deliverable bonds will be negative. However, negative profits for
these bonds do not imply arbitrage opportunities. To capture these “gains,” the long would need to take
delivery of a bond issue that is not the cheapest to deliver. Since it is not rational for the short to deliver any
issue other than the cheapest to deliver, no arbitrage gains are possible.
A cheapest-to-deliver issue arises because the conversion factors are derived by discounting the cash
flows of all bonds at 8 percent. Using 8 percent assumes that coupon payments can be reinvested at 8
percent. If the average future interest rate at which the coupons can be reinvested exceeds 8 percent,
however, investors prefer high-coupon bonds over low-coupon bonds (when each bond is valued by
discounting at 8 percent). The cheapest-to-deliver bond is, therefore, the low-coupon bond. If the average
future interest rate at which coupons can be reinvested is less than 8 percent, investors prefer low-coupon
bonds over high-coupon bonds, so the cheapest to deliver is the high-coupon bond. Only when the term
structure of interest rates is flat at 8 percent, will all bonds be equally desirable for delivery. If the yield
curve is above 8 percent, low-coupon bonds are the cheapest to deliver, and if the yield curve is below 8
percent, high-coupon bonds are the cheapest to deliver.
F0 . (8.24)
CFi
The left-hand side of the equation is the futures price at time 0. The maturity of the futures contract is T
periods hence. The first term in the numerator of the right-hand side of the equation, ( Bi , 0 AI i ,0 )e rT . is the
time 0 cost of the bond taken forward to time T at the riskless rate of interest (i.e., the bond purchase is
financed at the short-term riskless interest rate). The second and third terms in the numerator represent
interest earned on the bond-accrued interest, AI i ,T , received when the bond is delivered against the futures
T
contract and the future value of the coupons received (if any), t 0
Ci ,t e r (T t ) . The conversion factor CFi in
the denominator “converts” bond i into the hypothetical 8-percent coupon issue upon which the futures
contract is designed.
Prior to discussing the cost-of-carry relation in more detail, it is worth noting that the “short-term
riskless rate of interest” used to finance the purchase of bonds required in the arbitrage transactions that drive
(8.24) is usually the rate on a repurchase agreement or “repo.”. Repurchase agreements are collateralized
loans. They involve a commitment to sell and then later to buy back a specific bond issue (presumably at the
maturity of the futures contract).76 The agreement specifies the date on which the bond will be repurchased,77
as well as the interest rate that will be paid on the loan. The dollar interest paid on the loan is computed as
76
Repurchase and reverse repurchase agreements are discussed at length in Stigum (1990).
77
When the term of the loan is one day, it is called an overnight repo. Terms of greater than one day are term repos.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 127/358
Note that, under this arrangement, the lender has committed to consummating actions opposite the borrower,
that is, the lender has entered into an agreement to buy and then later to sell the underlying bond. For this
reason, the lender is said to have a reverse repurchase agreement, a “reverse repo” or, simply, a “reverse.”
Both the borrower and lender gain from these agreements. The borrower gets a lower rate than he might
otherwise get at the bank, and the lender gets a higher rate than he might otherwise get on short-term, highly
liquid investments.
Before maturity, as at maturity, the futures price is based on the price of the cheapest to deliver, and
the cheapest to deliver is determined by finding the bond with the highest “cash and carry” portfolio profit,
i ,0 F0 (CFi ) AI i ,T t 0 Ci ,t e r (T t ) ( Bi , 0 AI i ,0 )e rT .
T
(8.25)
Again, the highest value of profit equals zero; otherwise arbitrage profits are possible. This identifies
the cheapest-to-deliver issue. All other profits will be less than zero. In other words, the futures price will be
less than in (8.24) for all bonds other than the cheapest to deliver.
Although (8.25) allows us to identify which bond is cheapest to deliver at time 0, there is no assurance that
this bond will also be cheapest to deliver at time T. Since the identity of the cheapest to deliver at time T is
uncertain, (8.24) does not hold as an equality, even for the bond that is currently cheapest to deliver. Indeed,
the short has a valuable quality option that gives him the right to choose which bond to deliver at time T.
Although the short may have entered a cash-and-carry position when bond i was cheapest to deliver, if, at
maturity, bond j is cheapest, the short can profit by selling bond i, buying bond j, and then delivering bond j
on his short futures commitment. Because this option to switch bonds is valuable, the investor doing cash-
and-carry arbitrage is willing to sell futures at a price below the price specified by the cost of carry on the
right-hand side of (8.24), that is,
( Bi ,0 AI i , 0 )e rT AI i ,T t 0 Ci ,t e r (T t )
T
F0 , (8.26)
CFi
or, alternatively,
( Bi ,0 AI i ,0 )e rT AI i ,T t 0 Ci ,t e r (T t )
T
where bond i is the current cheapest to deliver. The value of the quality option embedded in the T-bond
futures contract is estimated in Chapter 11.
The short futures also has a timing option that allows a choice about when during the contract month to
deliver. The most valuable element in the timing option is called the wildcard option. In the delivery month,
the futures price at which delivery is made is the settlement price established when the market closes at 2:00
PM. The short has until 8:00 PM to declare delivery. Obviously, if news arrives that justifies a decline in
bond prices, the short will choose to make delivery at the already established settlement price.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 128/358
Under this notation, the dollar change in the value of the unhedged fixed-income portfolio is DP P times the
interest rate change. If we buy h futures contracts against his fixed-income investment, the dollar change in
the overall portfolio is ( DP P hDF F ) times the interest rate change, which we equate to the dollar change in
the hedged portfolio at the desired duration level, D* P , that is,
D * P DP P hDF F , (8.28)
in order to determine the optimal hedge ratio. Rearranging to isolate h , we get
P( D * DP )
h . (8.29)
DF F
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 129/358
EXAMPLE 8.5
Suppose that the cheapest-to-deliver bond (and hence the T-bond futures) has a duration of 12.50. Suppose
also that the duration of the bond portfolio that we are managing is 10.00 and that the market value of the
portfolio is $50,000,000. If today’s date is November 13, 1991, and we wish to hedge completely against
movements in the level of long-ten-n rates until the end of February 1992, how many March 1992 T-bond
futures contracts should we sell?
50,000,000(0 - 10.00)
Substituting the example values into equation (8.29), we get h - 403.15
12.50 0.9921875 100, 000
Asset Allocation
At its most basic level, portfolio management involves a decision concerning what types of assets should be
purchased. For example, a fund manager might choose to invest 40 percent of the fund in stocks, 40 percent
in bonds, and 20 percent in real estate. Deciding what proportion of fund wealth to place in each asset
category is called the asset allocation decision.
Once the asset allocation decision is made and fund wealth is invested, dramatic changes to the
allocation are usually avoided because the transaction costs of liquidating assets in one category and buying
assets in another are excessive. Instead, fund managers use futures contracts to change the asset allocation
indirectly. To demonstrate, assume that a fund consists of S in stocks and B in bonds, for a total value of
V S B . Now, suppose the fund manager wants to change the amount invested in long-term bonds from
B to B * . The bond portfolio has a modified duration of DB . Rather than selling (buying) stocks to buy (sell)
bonds, the portfolio manager can effectuate the change by buying (selling) T-bond futures contracts. She
wants her bond portfolio to have income DB B * if interest rates fall 1 percent. She plans on generating that
amount with income from the current bond portfolio, DB B , and income from a T-bond futures position,
hDF F , that is,
DB B* DB B hDF F . (8.30)
Note that if the dollar investment in bonds is to be reduced (i.e., B B * ), T-bonds futures contracts are sold,
and, if the dollar investment in bonds is to be increased (i.e., B B * ), T-bond futures contracts are
purchased. Presumably, the reduction (increase) in bond investment is then transferred to stocks through
buying (selling) stock index futures contracts.
EXAMPLE 8.6
A fund manager currently has $50,000,000 in a stock portfolio whose composition matches the S&P 500 and
$50,000,000 in a bond portfolio whose modified duration is 12.00. Believing that stocks are going to do
extraordinarily well over the next three months, the fund manager wants to take advantage of the impending
stock market rise and to eliminate his interest rate risk exposure. Unfortunately, liquidating bonds and buying
stocks is expensive, particularly if, at the end of the three months, the manager wants to return to his fifty-
fifty portfolio mix. How can the fund manager use T-bond and S&P 500 futures to carry out his plans?
Assume that the cheapest-to-deliver bond (and hence the T-bond futures) has a duration of 9.00 and that the
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 130/358
price of a three-month T-bond futures contract is 96.00. Also, assume that the three-month S&P 500 futures
is priced at 325.
First, with respect to eliminating the interest rate exposure, the number of T-bond futures to sell is 694.44:
This action is tantamount to liquidating the bond investment. Second, to take a long position of $50,000,000
in stocks using the S&P 500 futures, the number of contracts to buy is
50,000, 000
h 694.44 .
325.00 500
8.9 SUMMARY
Following an introduction to the particular interest rate futures markets in the U.S., specific pricing and yield
conventions governing the trading of fixed-income securities are discussed. The major focus of this chapter
is interest rate risk management. To this end, the notions of modified duration and convexity are introduced
and applied. Following that, the relation between short-term and long-ten-n rates, that is, the term structure
of spot interest rates, is presented. From the spot rates, forward rates of interest are derived, and the forward
rates implied by cash and futures prices are compared. A detailed discussion of the specifics of T-bond
futures contract delivery and pricing is provided. The chapter concludes with two interest rate risk
management illustrations.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 131/358
9.7 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 132/358
Currency futures contracts were developed in response to the shift from fixed to flexible exchange rates in
1971. At that time, the United States ceased redeeming dollars for gold. By 1973, most countries stopped
maintaining the price of their currencies with respect to the dollar and allowed market forces to determine
their exchange rates. The increased volatility of exchange rates created a demand for currency futures
markets both as speculative and hedging vehicles.78
Futures Market
The U.S. futures market in currencies operates like futures in any other item. Table 9.1 gives contract terms
for the major foreign currency contracts, all of which are traded on the Chicago Mercantile Exchange's
International Monetary Market. Prices are quoted in U.S. dollars per unit of foreign currency. An example of
currency futures price quotes is contained in Table 9.2. While Table 9.1 indicates that eight maturity months
are available, only the nearby months are actively traded. Unlike other futures contracts, delivery of
currencies has relatively few complications. Only one "grade" of currency is available, delivery takes place
on a specific date during the delivery month, and no transportation costs are incurred.
TABLE 9.1 Currency futures contracts specifications (most active contracts in U.S. markets).
Trading
Contract Months80 Last Trading Day Delivery
Hours
7:20-2:00 1,3,4,6,7,9,10,12 Two business days prior to third
Third Wednesday
(CST) current Wednesday
Security (Exchange) Units/ Minimum Price Fluctuation
Canadian Dollar (CME) 100,000Can$/0.0001 ($10)
Deutsche Mark (CME) 125,000DM/0.0001 ($12.50)
French Franc (CME) 250,000FF/0.00005 ($12.50)
Japanese Yen (CME) 12,500,000yen/0.000001 ($12.50)
British Pounds (CME) 62,500pounds/0.0002 ($12.50)
Swiss Franc (CME) 125,000/0.0001 ($12.5.0)
78
While flexible exchange rate systems have short run variability, they do not have the periodic large price changes and trade
dislocations that were typical of fixed exchange rate systems.
79
See Grabbe (1986) and Solnik (1988) for greater detail on currency markets and international bond and stock markets.
80
The notation used in this column corresponds to the month of the calendar year (e.g., 1 is January, 2 is February, and so on).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 133/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 134/358
OTHER FUTURES
Settlement prices of selected contracts. Volume and open Interest of all contract months.
Forward contracts are subject to two kinds of default risk –credit risk and country risk. Credit risk is the risk
that the contra bank has poor credit. This risk is low because only the most creditworthy international banks
participate in the interbank market. In the futures market, clearinghouse margin requirements serve to limit
credit risk. Country risk is the risk that a country will impose restrictions on the transfer of currencies and
thereby make it impossible for two banks to carry out the terms of their forward contract. Such restrictions
were quite frequent under fixed exchange rate systems, but are less frequent today.
81
In principle, the price of wheat could be quoted both as the dollar price of wheat, say four dollars per bushel, or the wheat price
of dollars, which would be 1/4 bushels per dollar.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 135/358
be 0.6120 - 0.5877 = 0.0243 dollars per mark. To avoid triangular arbitrage opportunities, the following
condition must be met for all trios of currencies
where Si , j is the number of units of the i - th currency required to purchase one unit of the j - th currency.
TABLE 9.3 Spot and forward currency prices from the Interbank market.
EXCHANGE RATES: Wednesday, November 13, 1991. The New York foreign exchange selling
rates below apply to trading among banks In amounts of $1 million and more, as quoted at 3 p.m.
Eastern time by Bankers Trust Co. and other sources. Retail transactions provide fewer units of foreign
currency Per-dollar.
U.S. $ equiv. Currency per U.S. $
Country Wed. Tues. Wed. Tues.
Argentina (Austral) 0.0001008 0.0001008 9918.67 9918.67
Australia (Dollar) 0.7860 0.7870 1.2723 1.2706
Austria (Schilling) 0.08681 0.08681 11.52 11.52
BMrsin (Dlnar) 2.6539 2.6539 0.3768 0.3768
Belgium (Franc) 0.02966 0.02966 33.72 33.72
Brazil (Cruzeiro) 0.00144 0.00146 694.71 695.60
Britain (Pound) 1.7730 1.7725 0.5640 0.5642
30-Day Forward 1.7648 1.7640 0.5666 0.5669
90-Day Forward 1.7504 1.74% 0.5713 0.5716
180-Day Forward 1.7299 1.7291 0.5781 0.5783
Canada (Dollar) 0.8842 0.8838 1.1310 1.1315
30-Day Forward 0.8815 0.8814 1.1344 1.1346
90-Day Forward 0.8784 0.8779 1.1384 1.1391
180-Day Forward 0.8737 0.8733 1.1445 1.1451
Chile (Peso) 0.002844 0.002780 351.56 359.65
China (Renmlnbi) 0.185642 0.185642 5.3867 5.3867
Colombia (Peso) 0.001753 0.001753 570.38 570.38
Denmark (Krone) 0.1573 0.1573 6.3570 6.3555
Ecuador (Sucre) Floating rate 0.000966 0.000966 1035.00 10335.00
Finland (Markka) 0.24984 0.24941 4.0025 4.0095
France- (Franc) 0.17881 0.17879 5.5925 5.5930
30-Day Forward 0.17813 0.17808 5.6140 5.6156
90-Day Forward 0.17690 0.17695 5.6529 5.6545
180-Day Forward 0.17510 0.17504 5.7110 5.7130
Germany (Mark) 0.6112 0.6111 1.6362 1.6365
30-Day Forward 0.6090 0.6088 1.6421 l.6426
90-Day Forward 0.6045 0.6044 1.6543 1.6544
180-Day Forward 0.5982 0.5902 1.6717 1.6718
Greece (Drachma) 0.005405 0.005405 185.00 185.00
Hong Kong (Dollar) 0.12884 0.12884 7.7615 7.7615
India (Rupee) 0.03890 0.03890 25.77 25.77
Indonesia (Rupiah) 0.0005056 0.0005056 1978.00 1978.00
Ireland (Punt) 1.6330 1.6318 0.6124 0.6128
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 136/358
TABLE 9.3 Cont. Spot and forward currency prices from the Interbank market.
Israel (Shekel) 0.4308 0.4321 2.3215 2.3142
Italy (Lira) 0.0006121 0.0006117 1231.41 1232.01
Japan (Yen) 0.007686 0.007707 130.10 129.75
30-Day Forward 0.007678 0.007696 130.24 129.90
90-Day Forward 0.007666 0.007686 130.45 130.10
180-Day Forward 0.007656 0.007677 130.62 130.26
Jordan (Dinar) 1.4500 1.4500 0.6897 0.6897
Kuwait (Dinar) 3.4965 3.4965 0.2860 0.2860
Lebanon (Pound) 0.001134 0.001134 881.50 881.50
Malaysia (Ringgit) 0.3650 0.3647 2.7400 2.7400
Matta (Lira) 3.1250 3.1250 0.3200 0.3200
Mexico (Peso) Floating rate 0.0003254 0.0003254 3073.01 3073.01
Netherland (Guilder) 0.5423 0.5423 1.9445 1.9445
New Zeland (Dollar) 0.5610 0.5620 1.7825 1.7794
Norway (Krone) 0.1558 0.1558 6.4175 6.4195
Pakistan (Rupee) 0.0405 0.0405 24.72 24.72
Peru (New Sol) 1.0152 1.0051 0.99 0.99
Philippines (Peso) 0.03839 0.03839 26.05 26.05
Portugal (Escudo) 0.007067 0.007063 141.50 141.59
Saudi Arabia (Rlyal) 0.26663 0.26663 3.7505 3.7505
Singapore (Dollar) 0.5958 0.5959 1.6795 1.6780
South Africa (Rand) Commercial rate 0.3568 0.3574 2.8023 2.7961
Financial rate 0.3240 0.3240 3.0790 3.0790
South Korea (Won) 0.0013310 0.0013310 751.30 751.30
Spain (Peseta) 0.009723 0.009699 102.85 103.10
Sweden (Krona) 0.1673 0.1672 5.9775 5.9615
Switzerland (Franc) 0.6888 0.6892 1.4517 1.4510
30-Day Forward 0.6872 0.6675 1.4552 1.454
90-Day Forward 0.6835 0.6839 1.4631 1.4621
120-day Forward 0.6788 0.6792 1.4732 1.4724
Taiwan (Dollar) 0.038650 0.037906 25.74 26.38
Thailand (Baht) 0.03926 0.03926 25.47 25.47
Turkey (Lira) 0.0002044 0.0002020 4892.01 4950.00
United Arab (Dirham) 0.2723 0.2723 3.6725 3.6725
Uruguay (New Peso) Financial 0.000425 0.000425 2352.94 2352.94
Venezuela (Bolivar) Floating rate 0.01695 0.01661 59.00 60.20
SDR 1.38023 1.38189 0.72452 0.72365
ECU 1.24952 1.25088
Special Drawing Rights (SDR) are based on exchange rates for the U.S., German, British, French and
Japanese currencies. Source: International Monetary Fund.
European Currency Unit (ECU) Is based on a basket of community currencies. Source: European
Community Commission.
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 137/358
TABLE 9.4 Key currency cross rates. Late New York Trading Nov. 13, 1991
Dollar Pound Sfranc Guilder Yen Lira D-Mark Ffranc CdnDol
Canada 1.1313 2.0098 0.78021 0.61434 0.00870 0.00092 0.69235 0.20261 …
France 5.5835 9.919 3.8507 3.0320 0.04292 0.00454 3.4171 … 4.9355
Germany 1.6340 2.9028 1.1269 0.88732 0.01256 0.00133 … 0.29265 1.4444
Italy 1229.3 2183.8 847.76 667.53 9.449 … 752.29 220.16 1086.6
Japan 130.10 231.12 89.724 70.649 … 0.10584 79.621 23.301 115.00
Netherlands 1.8415 3.2714 1.2700 … .01415 0.00150 1.1270 0.32981 1.6278
Switzerland 1.4500 2.5759 … 0.78740 0.01115 0.00118 0.88739 0.25969 1.2817
U.K 0.56290 … 0.39821 0.30568 0.00433 0.00046 0.34449 0.10082 0.49757
U.S … 1.7765 0.68966 0.54304 0.00769 0.00081 0.61200 0.17910 0.88394
Source: Telerate
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
where rd (rf ) is the continuously compounded, riskless rate of interest in the domestic (foreign) currency and
T is the maturity of the futures contract.
In international finance, the relation (9.2) is called the interest rate parity (IRP) relation. It is
instructive to derive IRP in a somewhat different way. Consider an investor who has one U.S. dollar to
invest. If the dollar is invested domestically at the riskless rate, the value at maturity is e rd (T t ) .
On the other hand, the dollar may also be used to purchase foreign currency that is then invested at
the foreign riskless rate of interest. If the proceeds at maturity of the foreign investment are sold in the
futures market, a dollar return can be guaranteed at time t . At maturity, the dollar cash proceeds of a hedged
r (T t )
foreign investment are (1 / St )e f Ft
For example, consider an investment in Germany. Suppose St 0.57 , rf 0.05 per year,
T t 0.25 years, and Ft 0.58 . The hedged dollar proceeds on the investment in Germany are 1.03034
dollars.
In equilibrium, the two ways of investing the dollar-directly in the U.S. or indirectly in a foreign
country-must have the same value at maturity, assuming that both investments offer a riskless return and that
there is no default risk on the futures or forward contract. In other words, the absence of costless arbitrage
opportunities in the marketplace ensures that
r (T t )
e rd (T t ) (1 / St )e f Ft .
This expression is easily manipulated to give (9.2). Dividing by S, and subtracting one from each side,
equation (9.2) can also be written as
Ft St ( r r )(T t )
e d f 1 . (9.3)
St
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 138/358
The left-hand side of this expression, which in other futures markets we call the percentage basis, is called
the forward premium or the swap rate in the currency markets. The term swap rate comes from the fact that
investors frequently buy a foreign currency and agree to swap it back for dollars. The swap rate specifies the
gain or loss on such a transaction. The right-hand side is the interest differential between the two countries
over the time period T t .
The IRP relation can also be derived for the case in which interest is earned discretely over the life of
the futures contract. Suppose rd* and rf* are the U.S. and foreign riskless rates of interest for the time over
which funds are invested, T t .
Then an investment in the U.S. is worth $( 1 rd* ) at maturity, and a hedged investment in the foreign country
is worth $(1 / St )(1 rf* ) Ft at maturity. To eliminate riskless arbitrage opportunities, these outcomes must be
equal, which implies
F S rd rf
* *
. (9.4)
S 1 rf*
This equilibrium relation is plotted in Figure 9.1 as the 45-degree line. Arbitrage opportunities arise if an
observation is not on the line.
For example, consider point A. At A, exchange rates are S 0.50 and F 0.51 , yielding a forward
premium ( F S ) / S 0.02 ; and three-month interest rates are rd* 0.03 and rf* 0.02 , yielding an interest
differential of (rd* rf* ) /(1 rf* ) 0.01 . Although the interest differential is in favor of the U.S., it is profitable
to borrow in the U.S. and invest in the foreign country because the profit on the foreign exchange transaction
exceeds the loss on the interest differential. An arbitrageur could borrow one million dollars in the U.S.,
thereby incurring an obligation to repay (1 rd* ) 1.03 million dollars at maturity; and she would invest in
the hedged foreign investment, which pays $(1 / S )(1 rf* ) F 1.0404 million dollars at maturity. The
arbitrage profit is 1.0404 million minus 1.03 million or 10,400 dollars. The activity of arbitrageurs results in
a capital outflow from the U.S. that raises U.S. interest rates, raises the spot exchange rate, lowers foreign
interest rates, and lowers the futures price, until equilibrium is restored. In fact, all points above the line in
the figure imply a capital outflow from the U.S. Points below the line, such as point B, imply a capital
inflow into the United States. At point B, the interest differential is in favor of the U.S., but the forward
premium is zero. Arbitrage profits could be earned by borrowing in the foreign country at rf* , selling the
foreign currency at S , investing in the U.S. at rd* , and entering into a futures contract to purchase the foreign
currency at F to repay the loan. The dollar value of the face amount of the foreign loan is $(1 / S )(1 rf* ) F ,
and the value of the dollar investment is (1 rd* ) . Assuming that S 0.50; rd* 0.03; rf* 0.02 and
F 0.50 , the profit is 0.01 for every dollar borrowed and invested in the United States.
In today's highly efficient international financial markets, arbitrage is instantaneous, and deviations
from IRP are rarely observed except for very small deviations caused by transaction costs. Apparent
deviations may be observed for certain interest rates. Each country has only one spot rate and one futures rate
for any maturity, but each country has many short-term interest rates that reflect different degrees of risk.
Thus, it is possible to find two interest rates that cause IRP to be violated. Differences in risk or differences
in transaction costs, however, usually explain deviations from IRP.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 139/358
F S
S
FIGURE 9.1 Interest Rate Parity
A
0.02
Gd SG f
where Gd and G f are the U.S. and foreign prices of the commodity and S is the spot exchange rate (the
dollar price of the foreign currency). If commodity arbitrage takes place for every commodity, the LOP must
hold for bundles of the same commodities in the two countries. Changes in the price of the bundle in one
country due to inflation without a corresponding change in the price of the bundle in the other country must
imply a change in the exchange rate. In other words, under PPP, the exchange rate is the dependent variable,
and the prices of commodities in the two countries are independent variables.
Pd ,t
St (9.5)
Pf ,t
The variable Pd ,t , is the dollar price of a bundle of commodities in the U.S. and Pf ,t is the foreign currency
price of the same bundle. For example, at t 1991 , Pd ,t 500 dollars and Pf ,t 2,000 units of foreign
currency. That implies an exchange rate of 0.25 according to PPP. The relative version of PPP looks at the
change in the exchange rate as a function of the relative changes in the prices of the bundle of commodities
82
For detailed discussions of factors determining exchange rates, see Lessard (1985).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 140/358
in each country. We can derive the relative version of PPP by writing equation (9.5) for a base period, time
0. Dividing the base period equation into (9.5) yields
St P /P P /P St I d ,t
d ,t f ,t d ,t d ,0 or (9.6)
S0 Pd ,0 / Pf ,0 Pf ,t / Pf ,0 S0 I f ,t
where I d ,t and I f ,t , are the U.S. and foreign country price indexes with a time 0 base year. Equation (9.6) is
the relative version of PPP. Suppose the base year is 1967 and that in that year the bundle costs Pd ,0 200 in
the U.S. and Pf ,0 400 in the foreign country. That implies an exchange rate of .50 dollars in 1967. Using
the 1991 bundle prices assumed above, the price indexes are I d ,0 2.50 , in the U.S. and I f ,0 5.00 in the
foreign country. The ratio of these indexes implies a ratio of exchange rates in 1991 versus 1967 of 0.5,
which, according to PPP, means the exchange rate is 0.25 in 1991.
In practice, PPP does not seem to hold as well as in the example we have just presented. A difficulty arises
from measurement problems. Price indexes in different countries do not include the same commodities or
have the same weights when the same commodities are included. Difficulties also arise in measuring prices
of commodities accurately and at the same time in the different countries. Another difficulty arises from the
fact that transportation costs are high for many commodities, so that the law of one price cannot be
established. Indeed, transport costs are prohibitive for certain commodities. Moreover, many items such as
services are simply not traded. Services (say, haircuts) can only be traded by moving labor, but restrictions
on international migration prevent arbitrage of services. Neither of these difficulties is severe if the source of
a change in currency values is inflation that affects the prices of all commodities-traded and non-traded-in
the same way. In that case, any index will be representative. PPP does not hold in practice, however, because
the prices of different commodities move in different ways in different countries. It is possible, for example,
for a country's traded goods not to increase in price (because of cost cutting measures in the traded goods
sector of the economy), while the prices of non-traded goods increase considerably more. The exchange rate
of that country will not depreciate as much as would be predicted by the aggregate inflation in the country
because it has remained competitive in those goods that are traded internationally.
Despite these difficulties, PPP remains an important determinant of exchange rate changes over longer
periods of time, particularly when comparing countries with significantly different inflation rates. Table 9.5
presents some data for the exchange rates and inflation of seven industrial nations for the period 19.67 to
1983 in which substantial differences in inflation arose. The second column gives the ratio of the dollar price
of the foreign currency in 1983 to the dollar price in 1967.
During this period, the dollar appreciated relative to every country except West Germany and Japan.
For example, the price of the British pound fell by 44.9 percent, whereas the price of the German mark rose
by 56.4 percent. The third column shows that much of the change in exchange rates can be explained by
differences in inflation rates. Typically, when U.S. inflation is less than foreign inflation, the foreign
currency falls, and when U.S. inflation is greater than foreign inflation, the price of the foreign currency
rises. Japan is the only exception. According to the consumer price indexes for the U.S. and Japan, inflation
was about equal in the two countries, yet the Japanese yen appreciated by 52.6 percent. Part of the
explanation for this discrepancy lies in the fact that the consumer price index is not representative of the
price of Japanese traded goods. Consumer goods within Japan have risen in price, but goods traded
internationally have not risen in price to the extent implied by the Japanese consumer price index.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 141/358
The second column is the ratio of the dollar price of the foreign currency in 1983 to the dollar price of the
foreign currency in 1967. The third column is the ratio of the consumer price index in the United States to
the consumer price index in the foreign country. The base year for both indexes is 1967. The fourth column
is the ratio of the second column to the third column. A number greater than (less than) one indicates that the
currency had a real appreciation (depreciation) relative to the dollar.
mt v I t yt , (9.7)
where mt , is the money supply at time t divided by the money supply in the base period, v is the velocity of
money, I t is the price index, and yt is real income at time t divided by real income in the base period. The
price index can be written as
I t (mt v) / yt , (9.8)
which shows that, under the assumption that the velocity of money is stable, the quantity theory states that
inflation occurs if the money supply expands faster than real income.
Under the quantity theory, the price level may be written for both the domestic and foreign countries:
I d ,t (md ,t v) / yd ,t (9.8a)
I f ,t (m f ,t v) / y f ,t (9.8b)
The base period for the variables is assumed to be the same in the two countries. Substituting (9.8a) and
(9.8b) in (9.6) yields
St I d ,t md ,t vd y f ,t
. (9.9)
S0 I f ,t m f ,t v f yd ,t
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 142/358
Equation (9.9) says that the change in the exchange rate depends on relative money supply growths and
relative real income growths in the two countries, assuming constant velocities. Suppose that velocities in
two countries are one, that real income grew 100 percent in the foreign country ( y f ,t 2 ) and 50 percent in
the domestic economy ( yd ,t 1.5 ), that the money supply grew 300 percent in the foreign country ( m f ,t 4 )
and 100 percent in the domestic economy ( md ,t 2 ). That would imply St / S0 0.67 , a 33 percent decline in
the price of the foreign currency. The monetary approach can become a good deal more complicated as other
factors that affect the impact of money supply changes on the economy are considered.
Like PPP, the monetarist approach is less successful in explaining short-run changes in exchange
rates than in explaining long-run changes. Using the monetarist approach to predict exchange rates requires
predicting money supply growths in the two countries and other variables that affect inflation, something that
is not an easy task over short intervals.
Part of the problem is that the current account is endogenous, that is, it depends on fundamental forces that
also affect the exchange rate. For example, the current account depends on monetary and fiscal policy that
also have a direct effect on the exchange rate. A country might import more because its real income has
grown. A growth in real income could be consistent with a decline in the price of foreign currencies (a
strengthening of the domestic exchange rate) as shown above for the monetarist approach.
Analysts often emphasize the current account, but the balance of payments also includes a capital
account. Overall, the balance of payments must balance. A current account deficit must be matched by a
capital account surplus (assuming government reserves and borrowings do not change), and a current
account surplus must be matched by a capital account deficit. If a country, imports more than it exports, the
cost of the net imports must be financed by borrowing from abroad. If a country exports more than it
imports, the foreign currency earnings must be invested abroad (or used to reduce foreign debt). Some
analysts argue that capital flows are exogenous and that the current account is endogenous. Under this
argument, the U.S. trade deficit results from a large capital inflow to the U.S. in response to higher U.S.
interest rates and other factors. The capital account surplus, in turn, made resources available to the U.S.,
some of which were spent on imports, thereby generating the current account deficit. Under this scenario, the
price of foreign currencies declined even though the U.S. ran a current account deficit.
Because of capital flows and other macroeconomic factors, the balance of payments approach, with
its focus on the current account, has not proved adequate to explain the behavior of exchange rates.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 143/358
asset approach thus focuses on capital flows and the factors that determine capital flows. In addition to
expected real returns, capital flows respond to risk.
In summary, a variety of complicated forces affect exchange rates. PPP is important, but short-run deviations
from PPP are prevalent. These short-run deviations depend on monetary and fiscal policy, exogenous
changes in the demand for imports and exports, and exogenous changes in capital flows.
We noted in Chapter 4 that the trading pressures of hedgers might cause speculators to demand a risk
premium that brings about a divergence between the futures price and the expected spot price. In currency
markets, the risk premium could easily be positive or negative since hedgers could be buying foreign
currency or selling foreign currency. In the absence of a reason to assume the risk premium is a particular
sign, we shall assume that condition (9.10) is met and that no risk premium exists.
A speculator who expects a foreign currency to appreciate could also borrow dollars and buy the foreign
currency. The speculation is profitable if the currency appreciates more than the cost of holding the currency.
The mechanics of spot speculation are as follows: borrow a dollar, buy 1 / St , units of foreign currency, invest
~
the foreign currency at the foreign interest rate, rf* , sell the foreign currency at the future spot rate, ST , and
pay back the dollar plus interest of rd* . The expected profit is
~
(1 / St )(1 rf* ) E ( ST ) (1 rd* )
In the absence of a risk premium, equilibrium requires a zero-expected profit, which implies
(1 rd* ) ~
St E ( ST ) . (9.11)
(1 rf )
*
It is worth noting that the left-hand sides of (9. 1 0) and (9.1 1), taken together, yield IRP. Under IRP, spot
speculation and futures speculation are equivalent.
The International Fisher Effect
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 144/358
Relation (9.1 1) is often called the international Fisher effect (IFE). The IFE is usually written as
~
E ( ST ) (1 rd* )
St (1 rf* )
The IFE says that the expected change in the exchange rate equals the interest rate differential between the
countries. For example, if the one-year interest rate in the U.S. is 8 percent and the one-year interest rate in
Brazil is 80 percent, the IFE says the price of the Brazilian cruzeiro is expected to be 60 percent of its
current value.
The IFE takes its name from Irving Fisher, who argued that the domestic interest rate is
approximately equal to the real rate of interest plus the expected rate of inflation. The IFE assumes that
international differences in interest rates reflect differences in expected inflation rates. According to Fisher,
where ad* and a*f are the domestic and foreign real rates of interest, and d* and *f ; are the U.S. and foreign
inflation rates. If ad* a*f , the IFE becomes
~
E ( ST ) 1 E (~d* )
. (9.12)
St 1 E (~ *f )
In this form, the IFE is very much like the relative PPP equation (9.6), except that PPP is an ex post relation
between the realized exchange rate change and realized inflation while the IFE is an ex ante relation
between the expected exchange rate change and expected inflation.
83
Evidence on forecasting ability is provided by Levich, “Evaluating the Performance of the Forecasters,” in Lessard (1985).
84
For greater detail on hedging and on other financial issues in managing international operations see Shapiro (1989) and Eiteman
and Stonehill (1986).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 145/358
Transactions Risk
Transactions risk refers to the foreign currency risk of a particular future transaction denominated in a
foreign currency. For example, a U.S. company sells a product to a German importer who agrees to pay for
the product in three months, when shipment is received. Payment is specified in German marks. The dollar
value of that account receivable for the U.S. Company is uncertain. This transaction can be hedged by selling
forward the German marks to be received in three months. Such a forward sale is usually done through a
bank because the quantity and other terms of the forward contract can be tailored to the specific transaction
incurring currency risk. Futures contracts could also be used, but the size and maturity of a futures contract
may not match the hedging need exactly.
An example of transactions risk might be a U.S. airplane manufacturer who has committed to deliver
a jet to Lufthansa in one year for 40,000,000 marks, with payment to be made in one year. The spot price of
the mark is assumed to be St $0.6000 and the one-year forward price, Ft $0.61132075 . If left unhedged,
the dollar value of the contract is subject to fluctuations in the value of the German mark. At the current
forward price, the contract is worth $24,452,830, but if the spot price should fall two percent below its
current value to $0.5880, the contract would be worth $23,520,000, a loss of nearly one million dollars. The
manufacturer can hedge foreign exchange risk by entering a forward contract today to sell 40,000,000 marks
a year from now at the forward price. At maturity, when payment is made, the German mark proceeds from
the export of the jet are delivered to the bank in return for dollars. The manufacturer has a problem if there
are delays in delivery and payment. In that case, the forward contract has to be rolled over at some cost.
An alternative to the forward market hedge is a money market hedge. Under interest rate parity, a
money market hedge is equivalent to a forward market hedge. U.S. and German interest rates consistent with
the spot, and forward rates in the example are rd* 0.08 and rf* 0.06 . Under a money market hedge, the
U.S. jet manufacturer borrows German marks against the proceeds of the sale of the jet, which amounts to
40,000,000/(1.06) = 37,735,849 marks, and converts the marks to dollars at the current spot rate to get
(0.60)(37,735,849) = 22,641,509. If this amount is invested at 8 percent, the proceeds at maturity are
(22,641,509)(1.08) = 24,452,830, the same as under the forward contract. In practice, the two approaches
may not be the same since the short-term interest rates that cause interest rate parity to hold may not be the
rates at which the manufacturer can borrow and lend. If the manufacturer is in need of funds, the money
market hedge is preferable to the forward market hedge.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 146/358
An example of a balance sheet hedge of a non-contractual item is hedging a finished goods inventory
position valued at 5,000 pounds at the current price of the product in Britain. This hedge is more complicated
because a decline in the value of the British pound that would reduce the dollar value of the inventory might
be offset by an increase in the price of the finished good in the United Kingdom. Indeed, under PPP, that is
exactly what would be expected. Suppose declines in the dollar price of the British pound can be partially
offset by increasing product prices by 50 percent of the amount they would increase under PPP. If repricing
offsets half the decline in the price of the pound, the optimal hedge is to sell short about 2,500 pounds. In
effect, the ability to reprice inventory in the U.K. provides a partial natural hedge (for 50 percent of the
decline). The futures market hedge hedges the other half of the exchange rate decline. Table 9.6 details this
example.
9.7 SUMMARY
This chapter begins by describing currency markets and currency quotations. The cost-of-carry model for
currency futures is then derived and shown to be the same as the well-known interest rate parity condition.
Theories of the determinants of exchange rates-the purchasing power parity approach, the monetarist
approach, the balance of payments approach, and the asset approach-are described.
We distinguish futures speculation and spot speculation. Competition among speculators in the futures
market implies that the futures price is a good estimate of the expected spot price. Similarly, competition in
spot speculation leads to the international Fisher effect-that the expected change in the exchange rate equals
the interest differential between two countries.
Currency futures may be used to hedge against fluctuations in exchange rates. The chapter concludes
with descriptions of hedging transaction risk and balance sheet risk.
TABLE 9.6 Hedging foreign exchange risk of 1000 units of finished goods inventory when 50 percent
of the exchange rate change can be offset by raising the sale price.
Inventory Futures Market
Futures
Value Value Futures Value
Date 85 Exch. Rate Position
(pounds) (dollars) Price (dollars)
(pounds)
Sept. 1 5,000 1.60 8,000 -2,500 1.60 -4,000
Dec. 1 5,833 1.20 7,000 -2,500 1.20 -3,000
Gain -1,000 1,000
85
The initial price of inventory is assumed to be 5 pounds per unit. The old exchange rate is 1/3 greater than the new exchange
rate. The table assumes the selling price can be raised by half the amount that would offset the change in the exchange rate, or by
0.5(1/3)(5).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 147/358
10.9 RESUMEN
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 148/358
Mercando de Mercado de
Mercancías Opciones
Mercado de Mercado de
Futuros Opciones de
Futuros
Un contrato que provee a su tenedor el derecho a comprar la mercancía subyacente se llama opción de
compra (call); un contrato que provee el derecho de vender se llama opción de venta (put). En el contrato de
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 149/358
opciones, el precio especificado al cual se puede comprar o vender la mercancía se llama precio de ejercicio
o precio de cierre de una opción. Si el precio actual de la mercancía excede el precio de ejercicio de la
opción, la call es in-the-money y la put es out-of-the-money. Si el precio actual de la mercancía está por
debajo del precio de ejercicio de la opción, la call esta out-of-the-money y la put está in-the-money. Cuando
el precio actual de la mercancía es aproximadamente igual al precio de ejercicio de las opciones, tanto la call
como la de venta están at-the-money.
Dos tipos diferentes o estilos de contratos de opciones se negocian, las opciones Europeas y las
Americanas. Estos contratos de opciones son parecidos en muchos aspectos, excepto que las opciones
Europeas pueden ejercerse sólo al vencimiento, mientras que las opciones Americanas pueden ejercerse en
cualquier momento incluyendo el día de vencimiento.
La notación más comúnmente utilizada para representar los parámetros relacionados con el precio de
las opciones es la siguiente:
S ( ST ) precio actual (terminal aleatorio) de la mercancía
F ( FT ) precio actual (terminal aleatorio) de los futuros
X precio de ejercicio o strike
T tiempo hasta el vencimiento de la opción
c(S, T, X) Call europea con precio de ejercicio X y tiempo hasta el vencimiento T .
p(S, T, X) Put europea con precio de ejercicio X y tiempo hasta el vencimiento T .
C(S, T, X) Call americana con precio de ejercicio X y tiempo hasta el vencimiento T .
P(S, T, X) Put americana con precio de ejercicio X y tiempo hasta el vencimiento T .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 150/358
incurre en ningún costo para almacenar la mercancía y que no se dan dividendos ni ningún otro pago a los
propietarios de la mercancía. La notación vectorial para la posición larga de futuros es 1,1 y para la
posición corta de futuros es 1,1 . La notación vectorial es útil para determinar la ganancia de
combinaciones de posiciones establecidas en un solo precio de ejercicio dado que se pueden agregar las
posiciones correspondientes en los vectores. Por ejemplo, el equivalente de una posición larga en futuros de
una mercancía puede establecerse comprando una call y vendiendo una put : 0,1 1,0 1,1 . Esta es
una posición larga sintética en futuros. Otra manera de ver esto es sumar verticalmente las ganancias de las
posiciones de compra de la call y la venta de la put en la Figura 10.2. Se puede establecer el equivalente de
una posición corta en futuros, una posición corta sintética en futuros, vendiendo una call y comprando una
put: 0,1 1,0 1,1 .
FIGURA 10.2a Flujo de caja al Vencimiento de las Posiciones
Larga y Corta de las Calls (Ignorando la Prima Inicial).
Utilidad en el momento T
Comprar Call 0,1
SX
Precio de la Mercancía, S
Utilidad en el momento T
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 151/358
Ganancia en T
SX F
Precio de Commodity, S
Así como se pueden utilizar las opciones para sintetizar las posiciones en futuros para la mercancía
subyacente, se puede utilizar una posición en la mercancía subyacente combinada con una opción para
obtener otra opción. Una posición larga en una call 0,1 , por ejemplo, se puede repetir comprando futuros de
la mercancía y comprando una put: 1,1 1,0 0,1 . Similarmente, una posición larga en una put, 1,0 se
puede repetir vendiendo corto los futuros de la mercancía y comprando una call: 1,1 0,1 1,0 . Una
estrategia de inversión frecuente es vender una call contra una posición larga en futuros de la mercancía
subyacente, 0,1 1,1 lo cual produce la misma ganancia que vender una put, 0,1 .
Una posición que produce dinero sin importar la dirección en que cambie el precio es un “straddle”
(call y put al mismo precio de ejercicio). Un “straddle” consiste en comprar una call y una put:
0,1 1,0 1,1 . Por supuesto, esta posición tiene un precio alto porque el vendedor del straddle debe ser
compensado por las pérdidas esperadas por vender el straddle. Alguien que piensa que es probable que se
eleve el precio de la mercancía subyacente podría comprar dos calls. La notación vectorial para eso es 0, 2
porque el valor de las opciones sube dos dólares por cada dólar de incremento en el precio de la mercancía
por encima del precio de ejercicio.
Opciones Europeas
Al vencimiento, los límites inferiores de las opciones Europeas son dados por el diagrama de ganancia de la
Figura 10.2. La call se vende por su valor de ejercicio, esto es, c max[0, ST X ] y la put se vende por su
valor de ejercicio, p max[0, X ST ] . Antes del vencimiento, una opción europea se vende en por lo menos
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 152/358
el valor actual de su valor de ejercicio más una asignación por el valor actual del costo de almacenar la
mercancía. El precio de la call debe satisfacer la siguiente condición para una opción in-the-money:
S (ebT 1) S X
c( S , T ; X ) rT (10.1)
e rT e
donde S es el precio del activo subyacente en algún momento antes del vencimiento y T es el momento
hasta el vencimiento.86 Si el precio de la mercancía al vencimiento fuese S , el valor de ejercicio de la call
sería S X . El valor antes del vencimiento es el valor descontado, el segundo término en (10.1). La call
también debe reflejar los costos de almacenamiento asociados con una posición larga en la mercancía ya que
al comprar una opción se evitan estos costos de almacenamiento. El valor actual de los costos de
almacenamiento, incluyendo el costo del interés de los fondos atados en S , es el primer término de (10.1).
Dado que la opción podría estar out-of-the-money, condición completa del límite inferior, la cual se muestra
en la Figura 10.3, es
S (ebT 1) S X
c( S , T ; X ) max 0, rT . (10.2a)
e rT e
La condición establece que una call europea nunca se puede vender por menos de la cantidad dada en el lado
derecho de (10.2a). Correspondientemente, el límite inferior para una put europea es
S (ebT 1) S X
p ( S , T ; X ) max 0, (10.2b)
e rT e rT
Precio de "Commodity"
X/e bT S
S (ebT 1) S X
Para S X / ebT , c(S , T ; X ) rT
erT e
Para los valores de put que no son cero, el límite inferior es el valor actual del valor de ejercicio de la put
más una ajuste por costos de almacenamiento.
Opciones Americanas
Una característica singular de las opciones Americanas, a diferencia de las opciones Europeas, es que se
pueden ejercer antes del vencimiento. Dado que el derecho de ejercer antes no puede tener un valor negativo
(es decir, a usted no hay que pagarle para inducirlo a asumir un privilegio), se aplican las siguientes dos
condiciones:
86
El lado derecho de (10.1) se puede simplificar a Se ( br )T Xe rT , lo cual se hará más adelante en el capítulo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 153/358
C ( S , T ; X ) c( S , T ; X ) (10.3a)
y P( S , T ; X ) p( S , T ; X ) (10.3b)
Estas condiciones no quieren decir que las opciones Americanas tengan mayor valor que las Europeas –sólo
que no pueden tener valores menores. El derecho de ejercer antes puede no tener valor positivo, pero nunca
tendrá valor negativo.
Dado que las opciones Americanas no se pueden vender a menos que las opciones Europeas, el límite
inferior de las opciones Americanas es el límite inferior de la opción europea correspondiente al mismo
precio de ejercicio y vencimiento. Una opción americana tiene el beneficio adicional que puede ser ejercido
inmediatamente para recibir el valor de ejercicio S X para la call americana y X S para la put americana.
Esto significa que el límite inferior para la opción americana es el límite inferior para la opción europea o el
valor de ejercicio actual, el que fuese mayor. Los límites inferiores son
S (ebT 1) S X
C ( S , T ; X ) max 0, rT
rT , S X (10.4a)
e e
S (e bT 1) S X
P ( S , T ; X ) max 0, rT
rT
, X S (10.4b)
e e
Call Europea
El límite de precio inferior de una call europea se puede determinar considerando el valor inicial y final de
una cartera que consiste de una posición larga en una call europea c( S , T ; X ) , una posición larga de
Xe rT bonos libres de riesgo, y una posición rollover corta en la mercancía Se(b r )T como se muestra en la
Tabla 10.1.87 La cantidad de préstamo libre de riesgo se determina por el hecho que necesitamos disponible
X en el momento T para ejercer la call. Si, al vencimiento de una opción, el precio de la mercancía excede
el precio de ejercicio, la call es ejercida y la unidad de la mercancía subyacente recibida es utilizada para
cubrir la posición corta de la mercancía. Para pagar el precio de ejercicio de la call, se utilizan exactamente
los bonos libres de riesgo. Por lo tanto, el valor final neto de la cartera es cero. Por otro lado, si al
vencimiento el precio de la mercancía cae por debajo del precio de ejercicio, la call vence sin valer nada y el
valor de los bonos libres de riesgo excede el valor necesario para cubrir la posición corta de la mercancía. En
87
Noten que por comodidad en la tabla se han suprimido los términos de la notación de precio de opción.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 154/358
este caso, el valor final de la cartera es positivo. Dado que el tenedor de la cartera se asegura de no tener un
valor final negativo por su cartera, el valor inicial debe ser no-positivo, esto es,
c( S , T ; X ) Se( b r )T Xe rT 0
o
c( S , T ; X ) Se( b r )T Xe rT (10.5a)
de lo contrario, son posibles ganancias de arbitraje. Esta condición es la misma que (10.1), la cual se puede
observar sumando y restando Se rT en el lado derecho de (10.5a).
La condición (10.5a) muestra sólo uno de los límites de precio inferiores de la call europea. En un
mercado racional, el precio de la opción nunca será negativo ya que es un derecho y no una obligación. Por
lo tanto, la condición completa del límite de precio inferior para una call europea es
c( S , T ; X ) max[0, Se( b r )T Xe rT ] (10.6a)
TABLA 10.1 Operaciones de Arbitraje para establecer el límite inferior de precio de una call europea
Valor Final
Posición Valor Inicial ~ ~
ST X ST X
c ~
Comprar Call europea 0 S X T
rT
Comprar bonos libres de riesgo Xe X X
~ ~
Vender posición rollover en mercancía Se(b r )T ST ST
~
Valor neto de la cartera c Se (br )T Xe rT X ST 0
Call Americana
Dado que la opción americana siempre se vende por más que la correspondiente opción europea, la call
americana está limitada desde abajo por (10.6a). También es limitada desde abajo por los ingresos de ejercer
inmediatamente la call, S X . De lo contrario, se podrían obtener ganancias libres de riesgo al comprar la
call y ejercerla inmediatamente. Por lo tanto, la condición completa del límite de precio inferior para la
opción americana es
C ( S , T ; X ) max[0, Se( b r )T Xe rT , S X ] (10.7a)
Si b r el segundo término en los corchetes de (10.7a) excede al tercero, lo cual significa que el límite
inferior de la call americana es Se(b r )T Xe rT , el límite inferior europeo. En otras palabras, cuando b r ,
una call americana se comporta como una call europea; no se ejercerá hasta el vencimiento. Uno puede
escribir la condición de manera que el segundo término en corchetes de (10.7a) exceda el tercer término en
corchetes:
1 e rT
SX (10.8)
1 e ( b r )T
Si se satisface la condición b r , la call americana tendrá el mismo límite inferior que la call europea. La
condición, (10.8) se cumplirá si b r (teniendo en mente que estamos buscando los valores de S para los
cuales el límite inferior de la call exceda a cero). Por otro lado, para los valores b r , el segundo término en
los corchetes de (10.7a) puede ser menor que el tercer término con lo cual no se cumple (10.8). Esto implica
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 155/358
que el límite de la call americana es S X . Las opciones que no satisfacen (10.8) pueden ejercerse antes del
vencimiento y tendrán un valor mayor que la opción europea correspondiente.
Put Europea
El límite de precio inferior para una put europea se puede derivar considerando el valor inicial y final de una
cartera que consiste de una posición larga en una put europea, p( S , T ; X ) , una posición rollover larga en la
mercancía subyacente, Se(b r )T y una posición corta de Xe rT bonos libres de riesgo, como se muestra en la
Tabla 10.2. La cantidad de préstamo libre de riesgo se determina por el hecho que el ejercicio de la put al
vencimiento proporciona X . Por lo tanto, si al vencimiento de la put el precio de la mercancía cae por debajo
del precio de ejercicio, la mercancía disponible es vendida por el precio de ejercicio de la put al ejercer la
put. Los ingresos del ejercicio son entonces utilizados para pagar el préstamo libre de riesgo. El efecto neto
es que el valor final de la cartera será igual a 0. Por otro lado, si al vencimiento el precio de la mercancía
excede el precio de ejercicio, el valor neto de la cartera será positivo porque la put vence sin valer nada y el
precio de la mercancía excede la cantidad necesaria para pagar el préstamo libre de riesgo. Dado que la
cartera proporciona un valor final no-negativo, entonces es valor inicial debe ser no-positivo. Si el valor
inicial neto de la cartera es no-positivo, esto es si p( S , T ; X ) Se( b r )T Xe rT 0 , Entonces
p( S , T ; X ) Xe rT Se( b r )T (10.5b)
TABLA 10.2 Operaciones de Arbitraje para establecer el límite inferior de precio de una put europea
Valor Final
Posición Valor Inicial ~ ~
ST X ST X
p ~
Comprar Put europea X ST 0
~ ~
Comprar posición rollover en mercancía Se( b r )T ST ST
Pedir prestado Xe rT Xe rT X X
~
Valor neto de la cartera p Se( b r )T Xe rT 0 ST X
Put Americana
Naturalmente, la condición del límite de precio inferior para la put europea también se aplica a la put
americana. Pero, los ingresos del ejercicio de la put americana, X S , pueden ser mayores que
Xe rT Se( b r )T y sabemos que en un funcionamiento racional la put americana está limitada desde abajo por
X S . De otra manera, se podrían obtener ganancias de arbitraje libres de costo al comprar una put y
ejerciéndola inmediatamente. De manera que la condición completa del límite inferior para la put americana
es
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 156/358
Opciones de Futuros
La Tabla 10.3 contiene un resumen de las condiciones de límite de precio inferior para opciones de
mercancías y futuros. Recuerden que anteriormente en este capítulo vimos que un contrato de futuros es una
mercancía con una tasa de costo de acarreo cero. Los límites de precio en la última columna realizan esta
sustitución.
TABLA 10.3 Resumen de los límites de precio inferior para las opciones Europeas y Americanas
de mercancías y futuros
Tipo de Opción Opción de Mercancía Opción de Futuros
( b r )T rT
Call europea max[0, Se Xe ] max[0, ( X F )e rT ]
Call americana max[0, Se( b r )T Xe rT , S X ] max[0, F X ]
Put europea max[0, Xe rT Se (b r )T ] max[0, ( F X )e rT ]
Put americana max[0, Xe rT Se( b r )T , X S ] max[0, X F ]
Call Americana
El que la call americana escrita sobre una mercancía puede ser ejercida temprano depende de la tasa del
costo de acarreo, b . Si b r , la call americana no se ejercerá temprano. Para ver esto, consideren la
condición de límite inferior (10.7a) o, alternativamente, la Figura 10.4a. Si b r , los ingresos del ejercicio
de la call, S X , son siempre menos que el valor mínimo por el cual se negocia la call en el mercado,
Se(b r )T Xe rT .
Dado que la call americana vale más cuando no se ejerce o “viva” que ejercida o “muerta”, nunca se ejercerá
antes del vencimiento. Por lo tanto, si b r , la prima de ejercicio temprano de la call americana no vale
nada, esto es
C (S ,T ; X ) 0 (10.10)
y, de la ecuación (10.9a), la call americana tiene un valor igual a la call europea.
C ( S , T ; X ) c( S , T ; X ) . (10.11)
La intuición del hecho que la opción americana no será ejercida temprano cuando b r se pude desarrollar
más fácilmente considerando la cantidad mínima perdida por el ejercicio temprano, esto es
S X [ Se( b r )T Xe rT ] (10.12)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 157/358
La expresión (10.13) dice lo siguiente: Si la call americana se ejerce ahora en lugar de al vencimiento, el
tenedor de la call americana pierde de dos maneras. Primero, incurre en el valor actual de los costos de
almacenamiento que tendrá que pagar como resultado de aceptar la entrega de la mercancía subyacente,
S [1 e( b r )T ] 88. Al mantener la opción, el tenedor de la opción no tiene inversión directa en la mercancía
subyacente –sólo el derecho de compra la mercancía en el futuro. Si acepta la entrega, ahora él enfrenta el
prospecto de almacenar la mercancía, asegurarla, etc. Segundo, él incurre en el valor actual del interés sobre
el precio de ejercicio de la opción, X [1 e rT ] . Si la opción se ejerce ahora, el tenedor de la opción está
obligado a realizar el pago de la cantidad X ahora en lugar de más adelante, renunciando así al ingreso por
interés que hubiese podido ganar sobre el precio de ejercicio de la opción. La Figura 10.4a resume
convenientemente estos efectos mostrando que el límite de precio inferior de la call europea excede los
ingresos del ejercicio de la call americana para todos los valores posibles del precio de la opción.
FIGURA 10.4a Precios de call europea y americana como función del precio
del subyacente cuando el costo de acarreo (b) excede la tasa de interés (r)
( b r ) de manera que la call americana no se ejerza óptimamente antes del
vencimiento
Precio de la call,
c(S,T; X) and C(S,T; X)
Precio de la
X/e bT
X Mercancía, S
En el caso donde la tasa del costo de acarreo, b , es menor que la tasa de interés libre de riesgo, r ,
puede ser óptimo ejercer temprano la call americana, tal como se ve al examinar la expresión (10.13).
Cuando b r , existen influencias compensadoras que afectan la decisión sobre el ejercicio temprano. Por un
lado, el diferir el ejercicio temprano permite al tenedor de la call obtener implícitamente interés sobre el
precio de ejercicio de la opción, como vimos anteriormente. Por otro lado, b r significa que el tenedor
obtiene alguna forma de rendimiento sobre la mercancía subyacente. Por ejemplo, supongamos que la call es
escrita sobre un índice de acciones y que la cartera del índice de acciones paga dividendos a una tasa
conocida.89 Diferir el ejercicio significa que se está obteniendo un interés, pero que se renuncia al dividendo.
Noten que mientras mayor es el valor del precio actual de la mercancía, S , mayor será el valor de la
88
Dado que b r , este término es no-positivo.
89
Recuerden que la tasa de costo de acarreo, b , consiste de la tasa de interés, r , más el costo de almacenamiento, seguro, etc.
Dado que el único “costo” aparte del interés involucrado en el mantenimiento de una cartera de acciones el es pago de dividendos
(es decir, un costo negativo), b r .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 158/358
expresión (10.13) y mayor será la rentabilidad del ejercicio temprano. El hecho se refleja en la Figura 10.4b,
mediante la creciente distancia entre el valor de ejercicio de la call americana y el límite de precio inferior de
la call europea conforme aumenta el precio de la mercancía subyacente. Sin embargo, en cualquier nivel de
precio de la mercancía existe un beneficio no-cero para el ejercicio temprano si b r , de manera que la
prima de ejercicio temprano tiene un valor positivo,
C (S ,T ; X ) 0 (10.14)
y la call americana tiene un mayor valor que la call europea,
C ( S , T ; X ) c( S , T ; X ) (10.15).
FIGURA 10.4b Precio de la call europea y americana como función del precio
del subyacente si el costo de acarreo (b) es menor que la tasa de interés (r)
b r de manera que la opción americana pueda ser ejercida temprano.
C(S * ,T;X)
C(S,T;X)
Valor de Ejercicio de la
Call Americana, S X
c(S,T;X)
C (S,T;X)
Precio de la Mercancía, S
X(e -rT -1 )
X/e bT X S*
e(b-r)T -1
La Tabla 10.4 contiene un ejemplo de una call para la cual el ejercicio temprano puede ser óptimo. La
mercancía subyacente en este ejemplo es una moneda extranjera. La tasa de interés de Estados Unidos se
asume como 8 porciento anual y la tasa de interés extranjera se asume como 12 porciento anual. Como
resultado, el costo de acarreo es negativo, esto es, -4 porciento anual. El precio de ejercicio de la call es 150
centavos y el vencimiento es en 30 días. En este ejemplo, el ejercicio temprano ocurre si el precio de la
moneda alcanza los 165 centavos. En este punto, el precio de la opción americana es igual al límite
americano porque la opción es valorada sobre la premisa que será ejercida. El ejercicio temprano es deseable
porque el tomar posesión de la moneda e invirtiéndola en el país extranjero proporciona una tasa
relativamente alta de ingreso por interés (12 porciento versus 8 porciento localmente). Si la opción se
mantuviese hasta el vencimiento, su valor es sólo 14.5692 (es decir, el valor de la opción europea). No
importa que, para precios entre 161 y 164, el límite inferior americano exceda el precio de la opción europea
porque el ejercicio temprano no es óptimo. En este rango, el precio de mercado de la opción americana
excede el límite inferior de la opción americana, por lo tanto el ejercicio no es óptimo.
Put Americana
La condición (10.7b), así como la Figura 10.5, muestran que siempre existe una oportunidad para que
cualquier put americana sea ejercida óptimamente antes del vencimiento. Siempre hay alguna región de los
precios de las mercancías sobre el cual los ingresos del ejercicio de la put serán mayores que la condición del
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 159/358
límite de precio inferior de la put europea. Para lograr una intuición sobre la naturaleza del intercambio
involucrado aquí, veamos la diferencia de los ingresos del ejercicio de la put americana y el límite de precio
inferior de la put europea, como lo hicimos con la call americana en la expresión (10.13). La diferencia es
X S [ Xe rT Se( b r )T ] (10.16)
o, simplemente
X [1 e rT ] S [1 e (b r )T ] (10.17)
TABLA 10.4. Límites Inferiores y Precios para las Calls Europeas y Americanas de moneda
extranjera. El precio de ejercicio de la opción ( X ) es 150, y el tiempo al vencimiento ( T ) es 30 días
(0.08219 años). La tasa de interés libre de riesgo doméstica ( rd ) es 8 porciento y la tasa de interés libre
de riesgo extranjera ( r f ) es 12 porciento. La tasa de costo de acarreo es por lo tanto –4 porciento. (La
moneda subyacente tiene una tasa de volatilidad anual de 20 porciento).
Límite Inferior Límite Inferior Precio de Opción Precio de Opción
Precio Actual
Europeo Americano Europea Americana
150 0.0000 0.0000 3.1637 3.1991
151 0.5011 1.0000 3.6704 3.7125
152 1.4913 2.0000 4.2224 4.2726
153 2.4814 3.0000 4.8784 4.8189
154 3.4716 4.0000 5.4581 5.5288
155 4.4618 5.0000 6.1381 6.2219
156 5.4520 6.0000 6.8564 6.9556
157 6.4422 7.0000 7.6103 7.7276
158 7.4324 8.0000 8.3969 8.5355
159 8.4226 9.0000 9.2132 9.3767
160 9.4127 10.0000 10.0560 10.2489
161 10.4029 11.0000 10.9225 11.1496
162 11.3931 12.0000 11.8097 12.0769
163 12.3833 13.0000 12.7149 13.0289
164 13.3735 14.0000 13.6355 14.0041
165 14.3637 15.0000 14.5692 15.0000
166 15.3538 16.0000 15.5138 16.0000
167 16.3440 17.0000 16.4677 17.0000
168 17.3342 18.0000 17.4291 18.0000
169 18.3244 19.0000 18.3967 19.0000
170 19.3146 20.0000 19.3693 20.0000
El primer término de (10.17) es el valor actual del interés que se puede obtener si se ejerce la opción
inmediatamente. Si el tenedor de la opción ejerce su put, él recibe X y entrega la mercancía subyacente que
vale S .
Los ingresos del ejercicio se puede invertir inmediatamente para obtener un interés. El efecto neto del
segundo término puede ser positivo o negativo, dependiendo de si b r o b r . En el primer caso, ejercer la
opción significa que el tenedor de la opción puede entregar la mercancía subyacente y renunciar a los costos
de almacenamiento involucrados con la mercancía que actualmente le está proporcionando algún tipo de
rendimiento. Puede estar renuente a hacerlo, pero la expresión (10.17) será positiva para los casos donde el
rendimiento sobre la mercancía es menor que la tasa de interés libre de riesgo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 160/358
La posibilidad de ejercicio temprano de una put americana también se puede ver geométricamente en
la Figura 10.5. Independientemente del valor de b , siempre existe un rango de precios de la mercancía por
encima de los ingresos del ejercicio de la put americana, X S , en el cual son mayores que aquellos de la
put europea; por lo tanto, siempre existe la posibilidad que la put americana sea ejercida antes del
vencimiento. Para ver si este es el caso, consideren qué sucede si el precio de la mercancía cae por debajo de
cero. El valor de la put americana es igual al precio de ejercicio de la put dado que el tenedor de la put
americana tiene el derecho a vender una mercancía con precio cero por X en cualquier momento.
FIGURA 10.5 Precios de Puts Europea y Americana como Función del precio del
Subyacente
Precios de Puts,
p(S,T; X) y P(S,T; X)
X
Limite inferior del Precio de
P(S**,T;X)
Put Europea, Xe-rT - Se(b - r)T
Xe -rT
p(S,T;X) P(S,T;X)
De hecho, en el caso que el precio de la mercancía caiga por debajo de cero, el tenedor de la put
americana ejerce su opción inmediatamente porque (a) él puede comenzar a ganar un interés sobre los
ingresos del ejercicio, y (b) el precio de la mercancía puede elevarse en cuyo caso el precio de la put caerá.
Sin embargo, en S 0 , la put europea tiene un valor de Xe rT . Para reconocer esto, consideren las
condiciones de límite impuestas sobre el precio de la put. El límite de precio inferior está dado por la
condición (10.2b). En S 0 , el valor mínimo para la put es Xe rT . Por otro lado, dado que el precio de la
mercancía no puede ser menor que cero al vencimiento de la opción, el valor actual de los ingresos máximos
del ejercicio es Xe rT . Si el precio de la put europea está limitado por arriba y por abajo por Xe rT se deduce
que el precio es Xe rT . Dado que la put americana puede ejercerse inmediatamente para ingresos iguales a
X mientras que la opción europea tiene un valor menor, la prima de ejercicio temprano debe ser positiva
para S 0 . En general, mientras exista alguna posibilidad de ejercicio temprano, la prima de ejercicio
temprano tiene valor positivo.
P (S ,T ; X ) 0 (10.18)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 161/358
S (ebT 1)
c( S , T ; X ) p( S , T ; X ) 0 o
e rT
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 162/358
S (ebT 1)
c( S , T ; X ) p ( S , T ; X ) (10.20)
e rT
Si los precios de las puts fuesen altos con relación a los precios de las calls, los convertidores venderían puts,
1,0 , comprarían calls, 0,1 y asumirías posiciones cortas en futuros sobre la mercancía subyacente,
1,1 , para establecer una cobertura perfecta. Bajo la premisa que la posición corta puede obtener ingresos
de almacenaje que paga la posición larga, se da la misma relación de equilibrio.
Se requiere un ligero ajuste a la relación de paridad de put y call si el precio en el cual se establece la
mercancía subyacente difiere del precio de ejercicio de las opciones. Si S X , la relación de paridad put y
call las opciones Europeas es
S (ebT 1) S X
c( S , T ; X ) p( S , T ; X ) rT (10.21)
e rT e
donde el último término da cuenta que la call o la put está in-the-money. Aunque la ecuación de la paridad
put y call (10.21) se ve complicada, la idea básica es muy simple. La relación de paridad put y call
simplemente dice que el precio de la call menos el precio de la put es el valor actual del costo de mantener la
mercancía subyacente hasta el vencimiento de las opciones más el valor actual de la cantidad por la cual el
precio de la mercancía excede el precio de ejercicio. Nosotros asumimos que los costos de almacenamiento
son incurridos a una tasa continua, b , pero se puede realizar otras suposiciones o premisas. Por ejemplo, los
costos de mantenimiento podrían ser pagados al inicio como una suma abultada, B . En ese caso, el primer
término en el lado derecho de (10.21) sería B .
Para ilustrar la paridad put y call para las opciones Europeas, consideren el ejemplo de la Tabla 10.4.
Específicamente, cuando el precio de la moneda subyacente es 155, el precio de la call europea es 6.1381.
Sobre la base de este precio, el precio de la put europea se puede calcular como
155(e 0.04 ( 0.08219 ) 1) 155 150
6.1381 p ( S , T ; X ) 4.4618
e0.08( 0.08219 )
Por lo tanto, el valor de la put europea dado por la paridad put y call es 1.6763.
La relación put y call arriba establecida puede no mantenerse exactamente para las opciones
Americanas porque el ejercicio temprano de una opción americana puede romper la cobertura libre de riesgo.
Por ejemplo, el convertidor que vende puts, compra calls y se mantiene corto en la mercancía subyacente,
puede ejercer la put. La mercancía entregada al convertidor al precio de ejercicio puede utilizarse para pagar
la mercancía prestada para la venta corta, pero el convertidor también debe liquidar la inversión de los
ingresos de la venta corta (los cuales son necesarios para pagar la mercancía entregada) y eso se podría hacer
a pérdida. Como resultado, para las opciones Americanas, se podrían establecer límites sobre la diferencia
entre los precios de las calls y las puts, algo que se hace en la siguiente sección.
Opciones Europeas
La relación de paridad put y call, (10.21) establecida anteriormente para las opciones de mercancías
Europeas, también se puede escribir como:
c( S , T ; X ) p( S , T ; X ) Se( b r )T Xe rT (10.22)
Para entender como se deriva esta relación. Consideren una cartera de inversión que consiste de vender una
call europea, comprar una put europea al mismo precio de ejercicio, comprar una posición rollover en la
mercancía subyacente comenzando con e(b r )T unidades, y pidiendo prestado Xe rT a la tasa de interés libre
de riesgo. El valor inicial y final de esta cartera se presentan en la Tabla 10.5. Noten que los valores finales
son iguales a cero, independientemente de si el precio final de la mercancía está por encima o por debajo del
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 163/358
precio de ejercicio de las opciones. Si al vencimiento la call está in-the-money, se requiere que entreguemos
una unidad de la mercancía y recibamos X . Por virtud de la posición rollover, tenemos disponible una
unidad de la mercancía para realizar la entrega. Los ingresos del ejercicio son utilizados para cubrir los
préstamos libres de riesgo, y la put vence sin valer nada. Si al vencimiento la put está in-the-money,
ejercemos la put, entregando la mercancía y recibiendo X . Las transacciones restantes son como las arriba
descritas. Dado que los valores finales son con seguridad iguales a cero, debe ser el caso que nadie pagará un
precio distinto a cero por asumir la cartera. El establecer el valor neto inicial de la cartera igual a cero
produce la ecuación (10.22).
TABLA 10.5 Operaciones de Arbitraje para establecer la paridad put y call para las opciones
Europeas
Valor Terminal
Posición Valor Inicial ~ ~
ST X ST X
~
Vender call europea c ( ST X )
p ~
Comprar put europea X ST
~ ~
Comprar posición rollover en la mercancía Se( b r )T ST ST
Pedir prestado Xe rT Xe rT X X
Valor neto de la cartera c p Se(b r )T Xe rT 0 0
La relación de paridad put y call para opciones de futuros Europeas es una caso especial de (10.22) donde la
tasa de costo de acarreo, b , es igual a cero. (Recuerden que las posiciones en futuros no requieren de ningún
desembolso de inversión inicial.) Le relación es
c( F , T ; X ) p( F , T ; X ) e rT ( F X ) (10.23)
Opciones Americanas
La característica del ejercicio temprano de las opciones Americanas causa que la especificación de la
relación de paridad put y call sea diferente a la de las opciones Europeas. Las relaciones que vinculan el
precio de la mercancía y los precios de las opciones Europeas de mercancías,
S X C ( S , T ; X ) P( S , T ; X ) Se( b r )T Xe rT , if b r (10.24a)
y
( b r )T
Se X C ( S , T ; X ) P( S , T ; X ) S Xe rT , if b r , (10-24b)
deben desarrollarse mediante dos conjuntos separados de transacciones de arbitraje. Nosotros consideramos
una desigualdad a la vez.
El lado izquierdo de la condición (10.24a) puede ser derivado considerando los valores de una cartera
que consiste de comprar una call, vender una put, prestar X libre de riesgo, y vender una posición rollover en
una mercancía comenzando con una unidad y disminuyendo la posición por el factor e ( br ) cada día y
prestando X libre de riesgo. La Tabla 10.6 contiene estos valores de cartera. Noten que ahora en la tabla
existe una columna adicional con el título “Valor Intermedio”. Dado que el tenedor de la cartera está corto en
una opción, corre el riesgo que se le asigne la entrega sobre la opción antes del vencimiento. Debemos tener
en cuenta esta posibilidad al derivar los límites de precio racionales de las opciones Americanas.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 164/358
En la Tabla 10.6, se puede ver que, si todas las posiciones del título valor se mantienen abiertas hasta
el vencimiento, el valor final de la cartera será positivo, independientemente de si el precio final de la
mercancía está por encima o por debajo del precio de ejercicio de las opciones. Si el precio final de la
mercancía está por encima del precio de ejercicio, se ejerce la call, y la mercancía adquirida al precio de
ejercicio X se utiliza para la entrega, en parte, contra la posición corta en la mercancía. Si el precio final de la
mercancía está por debajo del precio de ejercicio, el tenedor de la put ejerce su opción de vendernos la
mercancía subyacente al precio de ejercicio X . A su vez, utilizamos la mercancía para entregarla contra
nuestra posición corta en la mercancía establecida desde el inicio. Por lo tanto, si se mantienen las posiciones
de las opciones hasta el vencimiento, el valor final de la cartera es con seguridad positivo.
En el caso que el tenedor de la put decida ejercer su opción antes del vencimiento en un momento t ,
la inversión en bonos libres de riesgo es más que suficiente para cubrir el pago del precio del ejercicio al
tenedor de la put, y la mercancía recibida del ejercicio de la put es utilizada para cubrir la mercancía vendida
cuando se formó la cartera. Además, aún mantenemos la call, la cual puede tener un valor significativo. En
otras palabras, formando una cartera de títulos valores en las proporciones antes mencionadas, hemos
formado una cartera que nunca tendrá un valor futuro negativo. Si el valor futuro se asegura como no-
negativo, el valor inicial se asegura como no-positivo, o C ( S , T ; X ) P( S , T ; X ) S X 0
TABLA 10.6. Operaciones de Arbitraje para establecer la paridad put y call para las opciones
Americanas, donde b r .
Put Ejercida
Put Ejercida Temprano
Temprano
Valor Inicial Valor Intermedian Valor Final
~ ~
Posición ST X ST X
~ ~
Comprar call americana C Ct ST X
~ ~
Vender put americana P ( X St ) ( X ST )
Vender posición rollover en ~ ~ ~
S S t e ( br ) t ST e (b r )T ST e (b r )T
mercancía
Prestar X X Xe rt Xe rT Xe rT
~
Ct X [e rt 1] X [e rT 1] X [e rT 1]
Valor neto de la cartera C P S X
St [1 e ( b r ) t ] ST [1 e ( b r )T ] ST [1 e (b r )T ]
Despejando se obtiene el lado izquierdo de la ecuación (10.24a). El lado izquierdo de la ecuación (10.24b)
puede establecerse utilizando operaciones de arbitraje y argumentos similares a aquellos en la Tabla 10.6,
excepto que la posición rollover en la mercancía comienza con una inversión de e( b r )t unidades.
El lado derecho de (10.24a) se puede derivar considerando la cartera utilizada para probar la paridad put y
call europea. Cambiando la notación para reflejar el hecho que estamos discutiendo opciones Americanas e
introduciendo la columna “Valor Intermedio” para reflejar el prospecto de un ejercicio temprano, la tabla del
valor de cartera se convierte en la Tabla 10.7. Aquí, el valor final de la cartera es con seguridad igual a cero,
si las posiciones de la opción se mantienen abiertas hasta ese momento. Las posiciones de la opción son
compensadas por la posición de la mercancía y los préstamos libres de riesgo son compensados por los
precios de ejercicio de las opciones. En el caso que el tenedor de la opción americana decida ejercer su
opción antes del vencimiento, el tenedor de la cartera utiliza su posición larga en la mercancía para cubrir su
obligación de mercancía en la call ejercida y utiliza los ingresos del ejercicio para retirar su deuda pendiente.
Después que se realizan estas acciones, aún tiene una posición abierta larga en la put, el efectivo en la
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 165/358
cantidad de X [1 e r (T t ) ] , y una posición en la mercancía que vale S [e( b r )(T t ) 1] . Dado que es seguro que
la cartera tendrá resultados no-negativos, el valor inicial debe ser no-positivo o
C ( S , T ; X ) P( S , T ; X ) Se( b r )T Xe rT 0 . (10.25)
Despejando se obtiene el lado derecho de la condición (10.24a) de la relación de paridad put y call
americana. El lado derecho de (10.24b) se establece considerando la cartera en la Tabla 10.7, con la
excepción que la posición rollover de la mercancía se inicia con una unidad en lugar de e( b r )T .
La relación de paridad put y call para las opciones Americanas es un caso especial de (10.24b). Dado
que la tasa de acarreo, b , es igual a cero, la relación se convierte en
Fe rT X C ( S , T ; X ) P( S , T ; X ) F Xe rT (10.26)
La Tabla 10.8 contiene un resumen de las relaciones de paridad put y call desarrolladas en esta sección.
TABLA 10.7 Transacciones de arbitraje para establecer la paridad put y call Americanas,
cuando b r. C ( S , T ; X ) P( S , T ; X ) Se( b r )T Xe rT .
Call Ejercida Call Ejercida al
Temprano Vencimiento
Valor Inicial Valor Intermedian Valor Final
~ ~
Posición ST X ST X
~ ~
Vender call americana C ( St X ) ( ST X )
~ ~
Comprar put americana P Pt X ST 0
Comprar posición rollover ~ ~ ~
Se ( b r )T
St e(b r )(T t ) ST ST
en mercancía
Pedir prestado Xe rT Xe rT Xe r (T t ) X X
~ r (T t )
C P Se ( br )T
Pt X [1 e ]
Valor neto de la cartera 0 0
Xe rT S [e ( br )(T t ) 1]
TABLA 10.8 Resumen de las relaciones de paridad de puts y calls para Commodities y Futuros.
Tipo de Opción Opciones sobre Subyacente
Europea c( S , T ; X ) p( S , T ; X ) Se ( br )T Xe rT
S X C ( S , T ; X ) P( S , T ; X ) Se( b r )T Xe rT , if b r
Americana
Se(b r )T X C ( S , T ; X ) P( S , T ; X ) S Xe rT , if b r
Opciones sobre Futuros del Subyacente
Europea c( F , T ; X ) p( F , T ; X ) e rT ( F X )
Americana Fe rT X C ( F , T ; X ) P( F , T ; X ) F Xe rT
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 166/358
concentrarnos en las relaciones de arbitraje que existen entre las opciones sobre el subyacente y las opciones
sobre futuros del subyacente.90
Opciones Europeas
La relación entre las opciones Europeas sobre el subyacente y las opciones Europeas sobre el futuro es muy
simple porque a término el subyacente y el futuro tienen el mismo precio. La call (put) sobre el futuro tendrá
exactamente el mismo valor que la call (put) europea sobre el subyacente. Así que
c( S , T ; X ) c( F , T ; X ) (1 0.27a)
y
p( S , T ; X ) p( F , T ; X ) . (10.27b)
En el caso de opciones Europeas las opciones sobre el subyacente y los futuros son sustitutos perfectos.
Opciones Americanas
La igualdad de las opciones Europeas se debe a que no pueden ser ejercidas antes del vencimiento y a que
entonces los valores del subyacente y los futuros son iguales. Para opciones Americanas, que si se pueden
ejercer antes, el precio de los futuros y el subyacente puede ser diferente. Cuando el precio de los futuros es
por lo menos el precio del subyacente (es decir F SebT , F S if b 0 ), la call americana escrita sobre los
futuros vale por lo menos lo que vale la call sobre el subyacente.
C ( F ,T ; X ) C (S , T ; X ) (10.28a)
Para ver esto considere los valores iniciales, intermedio y final de una cartera que consiste en una posición
larga en la call sobre futuros y una posición corta en la call sobre el subyacente como se ve en la tabla 10.9.
Si ambas opciones se mantienen hasta termino el valor neto de la cartera es cero. Si las opciones están out-
of-the-money terminan sin valor y si están in-the-money los flujos de caja se cancelan mutuamente. Si la
~ ~
opción americana se ejerce antes contra el tenedor de la cartera, el valor de la cartera es Ct St X . Pero el
~
límite inferior del precio de la call es Ft X . Como hemos asumido que Ft St , el valor intermedio de la
cartera es no-negativo. Para eliminar oportunidades de arbitraje el valor inicial de la cartera debe ser no-
positivo, así que la condición (10.28a) debe cumplirse.
Un argumento similar se pude desarrollar para las puts Americanas. Como las puts son derechos de venta del
subyacente el instrumento con el precio más bajo representa el mayor precio de las opciones. Por lo tanto
F S,
P( S , T ; X ) P( F , T ; X ) . (10.28b)
Las condiciones (10.28a) y (10.28b) son las relaciones de precios entre opciones Americanas sobre el
subyacente y los futuros del subyacente cuando el costo de acarreo, b , es positivo o igual a cero. In algunos
mercados b puede ser negativo. Por ejemplo en monedas b 0 cuando la tasa de interés foránea es mayor
que la doméstica. Cuando esto sucede se puede demostrar fácilmente que las condiciones (10.28a) y (10.28b)
serán invertidas. La tabla 10.10 nos da un resumen de las relaciones de precios desarrolladas en esta sección.
90
En el Capitulo 1 indicamos que estos contratos existen para muchas monedas como el Marco Alemán y para índices bursátiles
como el S&P 500.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 167/358
TABLA 10.9 Operaciones de Arbitraje para establecer la relación entre los precios de las opciones
Americanas escritas sobre el subyacente y sobre los futuros cuando b 0 . C ( F , T ; X ) C ( S , T ; X )
Call Ejercida Temprano Call Ejercida a la Expiración
Valor Terminal
Posición Valor Inicial Valor Intermedio ~ ~
ST X ST X
Compre opción ~ ~
C (F ,T ; X ) C ( Ft , T t ; X ) 0 ST X
de futuros
Venda opción
~ ~
sobre el C (S , T ; X ) ( St X ) 0 ( ST X )
subyacente
Valor Neto de ~ ~
C (S ,T ; X ) C (F ,T ; X ) C ( Ft , T t ; X ) ( St X ) 0 0
la Cartera
TABLA 10.10. Resumen de las relaciones de precios entre opciones sobre subyacente y futuros.
Tipo de Opción Call Put
Europea c ( S , T ; X ) c ( F , T ; X ) p ( S , T ; X ) p ( F , T ;X)
C ( S , T ; X ) C ( F , T ; X ), si b 0 P ( S , T ; X ) P( F , T ; X ), si b 0
Americana
C ( S , T ; X ) C ( F , T ; X ), si b 0 P ( S , T ; X ) P( F , T ; X ), si b 0
10.9 RESUMEN
Primero mostramos como posiciones de opciones se pueden representar usando notación de vectores. Luego
presentamos y explicamos los limites inferiores de los precios de las opciones.
La diferencia entre opciones Americanas y Europeas es el derecho de ejercer temprano que tienen las
opciones Americanas. Discutimos luego las condiciones bajo las cuales este derecho tiene valor. Primero
ejercer temprano tiene sentido si la opción esta substancialmente in-the-money. Segundo, para calls ejercer
temprano requiere que el costo de acarreo sea “pequeño” comparado con la tasa de interés; para puts ejercer
temprano requiere que el costo de acarreo sea “alto” comparado con la tasa de interés.
La relación de paridad de puts y calls se derivo para opciones Americanas y Europeas y para opciones
sobre le subyacente y sobre futuros del subyacente. Finalmente, se derivo la relación entre los precios de las
opciones sobre le subyacente y sobre futuros del subyacente.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 168/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 169/358
11 VALUACIÓN DE OPCIONES
Evaluación de E (cT )
Valor Actual de la Call
11.9 RESUMEN
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 170/358
11 VALUACIÓN DE OPCIONES
En el último capítulo examinamos los precios de las opciones implícitos en la ausencia de oportunidades de
arbitraje. El enfoque de ese capítulo proporcionó relaciones de precios muy interesantes, pero los resultados
tomaron la forma de límites en los precios de las opciones en lugar de las ecuaciones de valuación.
El enfoque utilizado aquí asume que las opciones son valorizadas como si todas las personas en la
economía fuesen neutrales al riesgo. Esta premisa es razonable porque el valor de la opción no depende de la
tasa de retorno esperada de la mercancía subyacente.
En la sección 1 se explica el concepto de valuación neutral del riesgo y su equivalencia a las
valuaciones adversas al riesgo. La sección 2 examina las implicancias de la premisa que los precios de las
mercancías están distribuidos lognormalmente. En la tercera sección las premisas de la lognormalidad y la
neutralidad ante el riesgo son utilizadas para poner precio a una call europea y en la cuarta sección a la put
europea. La sección 5 describe la sensibilidad de los precios de las opciones ante los cambios en los
determinantes subyacentes de las opciones. La sección 6 presenta la ecuación de valuación para una opción
que permite al tenedor intercambiar una mercancía riesgosa por otra. Esta opción, llamada opción de
intercambio, se introduce en muchos tipos de contratos de futuros. Los métodos de aproximaciones de las
valuaciones para las opciones americanas se discuten brevemente en la sección 7. La sección 8 describe
cómo se pueden calcular los parámetros de la ecuación de valuación, y la sección 9 concluye con un breve
resumen.
91
Para un recuento histórico del desarrollo del modelo de precios de opciones Black-Scholes, ver Black (1989).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 171/358
personas son neutrales al riesgo. Las variaciones de las dos distribuciones son las mismas, pero el valor
esperado de la distribución neutral al riesgo es menor que el valor esperado de la distribución adversa al
riesgo. Según en enfoque Cox-Ross, el valor esperado de la distribución neutral al riesgo se descuenta a la
tasa de interés libre de riesgo, y según el enfoque de Samuelson, el valor esperado de la distribución adversa
al riesgo se descuenta a la tasa de retorno ajustada al riesgo. Al final, ambos enfoques proporcionan el mismo
valor actual para la call.
Probabilidad
X ST
Figura 11.2 Distribuciones del Precio de la Mercancía para las Personas Neutrales al Riesgo y las Adversas
al Riesgo
Probabilidad
Neutral al Riesgo
Adversa al Riesgo
ST
En este capítulo, utilizamos el enfoque de valuación neutral al riesgo por su ductilidad matemática. Sin
embargo, antes de hacerlo, mediante una ilustración utilizando un modelo binomial simple demostraremos
que los dos enfoques producen el mismo resultado. Primero, demostraremos el concepto de cobertura libre
de riesgo. Segundo, mostramos la valuación neutral al riesgo. Finalmente, mostramos la equivalencia de la
valuación adversa al riesgo y la valuación neutral al riesgo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 172/358
concepto de cobertura libre de riesgo, consideren el siguiente problema numérico simple. Supongamos que el
precio actual de la mercancía es $40 y que al final de tres meses el precio de la mercancía será $45 o $35. La
figura a continuación ilustra los posibles movimientos en el precio de la mercancía.
$45
$40
$35
Ahora, consideren una call europea escrita sobre esta mercancía. Esta call tiene un precio de ejercicio de $40
y vence en exactamente tres meses. Al vencimiento, esta call tendrá un valor de $5 o $0, dependiendo de si el
precio de la mercancía es $45 o $35, como vemos en la figura a continuación.
$5
$c
$0
Ahora, supongamos que vamos a comprar una unidad de la mercancía y vender nc calls . El valor final de
esta cartera es $45 - 5nc si el precio de la mercancía se eleva y $35 si el precio de la mercancía baja. La
incertidumbre del valor final de la cartera puede eliminarse completamente estableciendo nc de manera que
$45 - 5nc 35 or nc 2 .
En otras palabras, si compramos una unidad de la mercancía y vendemos dos calls, el valor final de la cartera
es con seguridad $35. Este es el concepto de una cartera libre de riesgo. Debido a la existencia de la cartera
libre de riesgo, podemos poner precio a la call europea en el ejemplo anterior. El costo de formar esta cartera
de cartera libre de riesgo en el momento 0 es $40-$2 c . Dado que la inversión de $40 - 2nc proporciona un
valor final seguro de $35, entonces si invertimos alternativamente los $40 - 2nc en bonos libres de riesgo
también obtendríamos un valor final de $35. Si la tasa de interés libre de riesgo durante el intervalo de tres
meses es 2 porciento, entonces la ausencia de oportunidades de arbitraje libre de costo en el mercado
requiere que $(40-2 c )(1.02) = $35.
El hecho que se pueda formar una cartera libre de riesgo entre la opción y la mercancía subyacente tiene una
implicancia importante –el precio de la call riesgosa se puede derivar sin conocer la tasa de retorno esperada
sobre la mercancía. Aún cuando en el ejemplo anterior no se conocen las probabilidades que el precio de la
mercancía suba a $45 o baje a $35, podemos poner precio a la opción. En otras palabras, el valor de la call
con relación a la mercancía no se ve influenciado por las preferencias del inversionista. No importa si una
persona es adversa al riesgo o neutral al riesgo, ambas están dispuestas a pagar $2.84 por la call en el
ejemplo anterior.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 173/358
Finalmente, el valor actual de la call es simplemente el valor actual del valor final esperado. Según la
premisa de neutralidad ante el riesgo, la tasa de descuento es la tasa de interés libre de riesgo, de manera que
el precio actual de la call es,
$2.90
c $2.84 ,
1.02
Exactamente el resultado que obtuvimos utilizando la cartera de cobertura libre de riesgo. Es importante
recordar que este enfoque pone precio a la opción con relación al precio actual de la mercancía, el cual se
asume es “correcto”.
esto es, p ' es 66 porciento. La mayor probabilidad de un alza refleja el hecho que la persona adversa al
riesgo exige una mayor recompensa por asumir el riesgo que una persona neutral al riesgo. El precio
esperado de la opción al vencimiento es, por lo tanto,
E (cT ) $5(0.66) $0(0.34) $3.30 .
El siguiente paso en la valuación es determinar la tasa de descuento ajustada al riesgo apropiada para la call.
En una economía neutral al riesgo, la tasa es simplemente la tasa de interés libre de riesgo dado que las
personas son indiferentes ante el riesgo. Sin embargo, las personas adversas al riesgo se preocupan por el
riesgo y exigen tasas de retorno mayores por activos o mercancías más riesgosas. Por ejemplo, según el
modelo de precios de activos de capital (CAPM), la tasa de retorno esperada sobre la mercancía es
ES r ( EM r ) S ,
Dado que S representa el cambio porcentual en el precio de la mercancía con relación al cambio porcentual
en la cartera de mercado, podemos multiplicar S por el cambio porcentual en el precio de la call con
respecto a un cambio porcentual en el precio de la mercancía para obtener el beta de la call y, por lo tanto, la
tasa de retorno esperada. Esto es,
dc / c
Ec r ( E M r ) c , Ec r ( E M r ) S
dS / S
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 174/358
E (ct ) 3.30
Por lo tanto, el valor actual de la call es c
1 Ec 1 0.02 0.40 / c
3.30 0.40
de manera que el precio de la call es c 2.84 .
1.02
exactamente el mismo resultado obtenido para la economía neutral al riesgo.
La tasa de retorno aleatoria sobre la mercancía durante T períodos se define como el precio relativo menos
uno, esto es
~
ST / S 0 1
La tasa de retorno aleatoria y compuesta continuamente durante los T períodos es
~ ~
x ln( ST / S0 ) 1 (11.2)
Noten que el retorno compuesto continuamente de T períodos es la suma de los retornos periódicos
compuestos continuamente, esto es
~ ~ ~
ln( ST / S0 ) t 1 ln(St / St 1 ) (11.4)
T
Una premisa que se utiliza comúnmente en el desarrollo de modelos financieros es que los retornos de los
activos son distribuidos independiente e idénticamente en cada período. Por lo tanto, el retorno periódico
esperado compuesto continuamente es
E[ln( St / St 1 )] (11.5)
y, por la ecuación (11.4), el retorno esperado compuesto continuamente de 0 a T es
~ ~ ~
E[ln(ST / S 0 )] t 1 E[ln(S t / St 1 )] T
T
(11.6)
92
Estos precios de la mercancía son observados a intervalos iguales a lo largo del tiempo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 175/358
El primer y segundo término en (11.8) son iguales por la independencia entre los retornos en períodos
diferentes. La desviación estándar del retorno compuesto continuamente de 0 a T es T . La segunda
premisa que invocamos es que las tasas de retorno periódicas compuestas continuamente están distribuidas
normalmente con una media y una varianza 2 . En este caso, el retorno compuesto continuamente de 0 a
T también está distribuido normalmente con una media T y una varianza 2T . También implica que la
distribución de los precios de los activos es lognormal con una media
~
E ( ST ) S 0 eT , (11.9)
donde 2 / 2 (11.9a)
0.4
Pr
ob 0.3
abi
0.2
lid
ad 0.1
0
-5 -3 -1 1 3 5
Retornos del Subayacente
1 log( x / m) 2
Figura 11.3(b) Distribución Lognormal: f ( x; m, ) 2
exp , x 0
2 L2
L
x L 2
2 2 2
m e 0.5 , L2 e 2 (e 1); , los parametros de la distribución normal.
Pr 0.5
ob 0.4
abi 0.3
lid
0.2
ad
0.1
0
0 5 10 15 20 25 30 35
Las figuras 11.3a y 11.3b contienen ilustraciones de las dos distribuciones que implícitamente estamos
utilizando. La primera es la distribución normal para ~
x , que tiene una media T y una varianza 2T . La
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 176/358
~
segunda es una distribución lognormal de ST , que tiene una media de S 0 eT . Noten que la distribución del
precio tiene la característica intuitivamente atractiva que el precio final de la mercancía no puede caer por
debajo de cero. Si los precios finales se asumen como distribuidos normalmente, existiría una posibilidad que
el precio de la mercancía caiga por debajo de cero. Nuestro uso de la distribución normal se facilita más
transformando el retorno compuesto continuamente x o ln( ST / S0 ) , en una variable estándar distribuida
normalmente, ~ z , que tiene una media de cero y una varianza de uno, esto es
~ ~
~ x T ln( ST / S 0 ) T
z (11.10)
T T
que también se puede escribir en términos del precio final de la mercancía
~ ~
ST S 0 e T T z (11.11)
~
La variable z tiene la función de densidad
2
e z / 2
n( z ) (11.12)
2
La probabilidad que un resultado de esta distribución normal unitaria produzca un valor menor que la
constante, d , es
z2 / 2
d e
~
Pr( z d ) N (d ) dz (11.13)
2
Para evaluar la probabilidad N (d ) en (11.13), se puede utilizar una variedad de modelos. Las
aproximaciones polinómicas son populares porque son simples de programar. El Apéndice 11.2 contiene dos
de estas aproximaciones y sus niveles de exactitud. Otra alternativa es utilizar los valores de las
probabilidades normales tabuladas en los libros de texto de estadística y otras publicaciones. El Apéndice
11.3 contiene las probabilidades normales tabuladas en rango de d , desde –4.99 hasta +4.99.
Dos propiedades de la función de densidad normal unitaria serán útiles más adelante en este capítulo.
Primero, la probabilidad de obtener un valor mayor que d de una distribución normal unitaria es igual a uno
menos la probabilidad de obtener un valor menor que d ,93 esto es,
Pr( ~
z d ) 1 N (d ) . (11.14)
EJEMPLO 11.1
Calcule las probabilidades que el resultado de una distribución normal estándar proporcione un valor que
esté (a) dentro de una desviación estándar de la media, (b) dentro de dos desviaciones estándar de la media y
(c) dentro de tres desviaciones estándar de la media. Primero, se debe notar que cualquier variable distribuida
normalmente, ~ x , puede transformarse en una variable unitaria distribuida normalmente (es decir, una
variable con una media de cero y una varianza de uno) al aplicar la transformación (11.10). Segundo,
nosotros evaluamos las probabilidades utilizando los valores tabulados para la distribución normal unitaria
acumulada. (Ver Apéndice 11.3)
93
Este resultado se deduce simplemente de Prob (z d) Prob (z d) 1 y (11.13).
2 2 2 2
d e z /2 e z / 2 e z / 2 d e z /2
94
Este resultado se deriva de la siguiente manera: N (d ) dz dz dz dz 1 N (d )
2 d 2 2 2
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 177/358
Pr(1 ~
z 1) Pr( ~
z 1) Pr( ~
z 1) 1 2 Pr( z 1) = 0.84134 - 0.15866 = 0.68268
Pr(2 ~
z 2) Pr( ~
z 2) Pr( ~
z 2) 1 2 Pr( z 2) = 0.97725 - 0.02275 = 0.95450
Pr(3 ~
z 3) Pr( ~
z 3) Pr( ~
z 3) 1 2 Pr( z 3) = 0.99865 - 0.00135 = 0.99730
EJEMPLO 11.2
Asuman que el precio actual de la mercancía es $50 y que la tasa de retorno compuesta continuamente tiene
una media anualizada de 16 porciento y una desviación estándar de 32 porciento. Calcular la probabilidad de
que el precio de la mercancía exceda 75 al final de tres meses.
Primero, utilizamos la ecuación (11.10) para transformar el precio final lognormal en un valor
variable normal unitario. Específicamente,
ln(75 / 50) 0.16(0.25)
d 2.28416
0.32 0.25
Segundo, redondeamos d al centésimo más cercano y utilizamos las tablas de probabilidades:
~
Pr( ST 75) Pr( ~
z 2.28) 0.98870 .
Noten que estamos evaluando la probabilidad que el precio final de la mercancía sea menos que 75 porque
las tablas encuentran el área bajo la función de densidad normal unitaria menos infinito hasta el límite d .
Para calcular la probabilidad que el precio final de la mercancía sea mayor que 75, simplemente tomamos el
~ ~
complemento o Pr( ST 75) 1 Pr( ST 75) 1 Pr( ~
z 2.28) 1 0.98870 0.01130 .
Noten que estamos introduciendo cierto error como resultado de redondear el límite integral superior
d a la centésima más cercana cuando utilizamos las tablas. Podríamos interpolar entre los valores de las
tablas para lograr una mayor exactitud, o podríamos utilizar una de las aproximaciones polinómicas del
Apéndice 11.2. Utilizando la segunda aproximación polinómica del apéndice obtenemos
N (2.28416) 0.98882 .
EJEMPLO 11.3
Utilizando los parámetros del Ejemplo 11.2, calcular la probabilidad que el precio de la mercancía caiga
entre 40 y 60 al final de seis meses.
Una vez más, utilizamos la ecuación (11.10) para transformar el precio final de la mercancía a la
variable unitaria distribuida normalmente. Los límites de la integración son
una vez más las probabilidades fueron calculadas utilizando la segunda aproximación polinómica del
Apéndice 11.2. El paso final involucra diferenciar las probabilidades, esto es,
~
Pr(40 ST 60) Pr(1.33972 ~
z 0.45220) Pr(~
z 0.45220) Pr(~z 1.33972)
0.67444 0.09017 0.58427 =58.427%
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 178/358
EJEMPLO 11.4
Utilizando los parámetros del Ejemplo 11.2, calcular el rango del precio de la mercancía en tres meses
asumiendo que estará entre dos desviaciones estándar de su nivel actual.
Utilice la ecuación (11.11) y establezca z igual a 2 . Los dos precios finales de las mercancías son
ST 1 50e 0.16 ( 0.25)0.32 0.25 ( 2 )
37.78919
ST 2 50e 0.16 ( 0.25)0.32 0.25 ( 2 )
71.66647 .
EJEMPLO 11.5
Supongamos que existe una call europea a tres meses sobre la mercancía en el Ejemplo 11.2 y su precio de
ejercicio es 50. Calcular la probabilidad que la call esté in-the-money al vencimiento. El límite integral
superior d es
ln(50 / 50) 0.16(0.5)
d1 0.25000
0.32 0.25
La probabilidad que el precio final de la mercancía sea menor que el precio de ejercicio es
Pr( ST 50) Pr( zT 0.25000) N (0.25000) 0.40129
EJEMPLO 11.6
Calcular la tasa de retorno esperada sobre la mercancía durante un intervalo de tres meses y el precio
esperado de la mercancía en ese momento en el tiempo. Por la ecuación (11.9a), sabemos que la tasa de
retorno esperada sobre la mercancía es igual a la media más la mitad de la varianza de la distribución del
~
logaritmo del ratio del precio de la mercancía, ln(ST / S ) , esto es, 2 / 2 0.16 0.32 2 / 2 0.2112
~
Por lo tanto, el precio final esperado de la mercancía es, E ( ST ) S 0 eT 50e 02112(0.25) 52.71094
Para evaluar el precio final esperado de la Call, asumimos que la tasa de retorno esperada sobre la mercancía
es igual a la tasa de interés libre de riesgo (neutralidad ante el riesgo) y los precios de las mercancías están
distribuidos de manera lognormal al vencimiento de la opción. Para descontar el precio final esperado de la
call al presente, asumimos que la tasa de retorno esperada sobre la call es igual a la tasa de interés libre de
riesgo (neutralidad ante el riesgo).
Para poder hacer que la ecuación (11.16) sea operacional, necesitamos evaluar el término E (c~T ) , el
~
valor final esperado de la call. Si se asume que ST está distribuido lognormalmente, la distribución del
precio final de la Call, c~T , es conocido dado que c~T es simplemente
~ ~
~ ST X for ST X
cT ~ (11.17)
0 for ST X .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 179/358
Con el precio final de la mercancía con una distribución lognormal, la condición (11.17) implica que el
precio final de la call tiene una distribución lognormal truncada y que la media de la distribución del precio
de la call es:
~ ~
E (c~T ) E ( ST X | ST X ) E (0 | ST X ) E ( ST X | ST X )
~ ~
E ( ST | ST X ) E ( X | ST X ) E ( ST | ST X ) X Pr( ST X )
Evaluación de X Pr( ST X )
Sea L( ST ) la función de densidad lognormal de ST , el término X Pr( ST X ) , es
X Pr( ST X ) L( ST )dST
X
La manera más fácil de evaluar el integral es realizar un cambio de variables en ST . La ecuación (11.10) nos
muestra la transformación que aplicamos a ST . El límite inferior y superior de la integración para la nueva
variable z se obtienen sustituyendo y X por ST en (11.10). Por lo tanto, los límites son y
[ln( X / S ) T ] / T , respectivamente. Por lo tanto,
[ln( X / S ) T ]
X Pr( ST X ) [ln( X / S )T ] n( z )dz T n( z )dz XN (d 2 ) (11.19)
T
donde d1 T [ln( X / S ) T ] / T d 2 T . Los pasos en (11.20) son los siguientes: (a) el valor
esperado condicional se expresa en forma de integral donde L( ST ) es la función de densidad lognormal para
ST ; (b) se realiza un cambio de variables en ST , y se elimina la función de densidad de la variable normal
estandarizada, z ; (c) Se T se factoriza fuera de la integral y se completa la raíz en el exponente dentro del
2
integral; (d) e T / 2 se factoriza fuera del integral y se simplifica la expresión restante en el exponente dentro
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 180/358
del integral; (e) se realiza un cambio de variables y T z y se redefinen los límites de integración;95 y
(f) la expresión se simplifica.
Evaluación de E (c~T )
Para resumir, estamos tratando de valorizar una call europea según la premisa que los precios de la
mercancía están distribuidos lognormalmente y que las personas son neutrales al riesgo. Estamos en el
proceso de valorizar el valor final esperado de la call, E (c~T ) . Sustituyendo las ecuaciones (11.20) y (11.19)
en la ecuación (11.18), ahora tenemos
E (c~T ) Se T T N (d1 ) XN (d 2 )
2
(11.21)
donde
d 2 d1 T (11.21a)
ln( S / X ) T
d1 (11.21b)
T
Sin embargo, no nos detendremos aquí. La tasa de retorno esperada de la mercancía en los límites del
~ ~
integral d1 y d 2 es la media de un logaritmo de los ratios del precio de la mercancía – E[ln(St / St 1 )] .
~
Quisiéramos expresar la tasa de retorno de la mercancía en términos de los precios relativos – E ( ST / S ) .
Sabemos que
~ 2
E ( ST / S ) eT e ( / 2)T (11.22)
Ahora, recuerden que hemos invocado una premisa de neutralidad ante el riesgo. El valor de en (11.22) es
la tasa de retorno esperada sobre la mercancía y, en un mundo neutral al riesgo, la tasa de retorno esperada
sobre la mercancía es igual a la tasa del costo de acarreo, b (es decir, el costo del interés más cualquier costo
adicional de almacenamiento). Sustituyendo b por en (11.9a) y aislando , obtenemos
b 2 / 2 . (11.23)
95
Donde y = -z, se mantiene la siguiente propiedad:
La propiedad se utiliza al simplificar (11.20).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 181/358
ln( S / X ) (b 0.5 2 )T
donde d1 (11.25a)
T
d 2 d1 T . (11.25b)
La interpretación de los términos en (11.25) es bastante directa dado nuestro enfoque de valuación neutral al
riesgo. El término Se ( br )T N (d1 ) es el valor actual del beneficio esperado al ejercer la call al vencimiento,
condicional a que el precio final de la mercancía sea mayor que el precio de ejercicio al vencimiento de la
opción. El término N (d 2 ) es la probabilidad que el precio de la mercancía sea mayor que el precio de
ejercicio al vencimiento. El valor actual del costo esperado de ejercer la condicional de la call siempre y
cuando la call esté in-the-money al vencimiento es Xe rT N (d 2 ) .
EJEMPLO 11.7
Calcular el precio de una call europea de moneda extranjera a tres meses con un precio de ejercicio de 40. El
tipo de cambio spot actualmente es 40, la tasa de interés doméstica es 8 porciento anual, la tasa de interés
extranjera es 12 porciento anual, y la desviación estándar del retorno compuesto continuamente es 30
porciento anual. Noten que la tasa de costo de acarreo, b , por lo tanto es –4 porciento.
Los valores de N ( d1 ) y N ( d 2 ) son 0.5033 y 0.4437, respectivamente, de manera que el precio de la call
europea es c 38.818(0.5033) - 39.208(0.4437) 2.14 .
Sustituyendo la ecuación de la call europea (11.25) por el término c( S , T ; X ) , encontramos que la ecuación
de la valuación de la put europea sobre una mercancía es
p( S , T ; X ) Se ( br )T N (d1 ) Xe rT N (d 2 ) Se( br )T Xe rT
p( S , T ; X ) Xe rT [1 N (d 2 )] Se ( br )T [1 N (d1 )]
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 182/358
donde
ln( S / X ) (b 0.5 2 )T
d1 (11.28a)
T
d 2 d1 T (11.28b)
Por lo tanto, la valuación de la put europea se deduce directamente de la paridad put-call europea y la
valuación de la call europea.
La interpretación de los términos en (11.28) es igual a la interpretación neutral al riesgo para la call.
El término Xe rT N (d 2 ) es el valor actual del beneficio esperado del precio de ejercicio de la put al
vencimiento, condicionado a que el precio final de la mercancía sea menor que el precio de ejercicio al
vencimiento de la opción. Recuerden que la put proporciona el derecho a vender la mercancía de manera que
el beneficio de mantener la opción es el efectivo que nosotros recibimos cuando ejercemos la opción, esto es,
X . N ( d 2 ) es la probabilidad que el precio de la mercancía sea menor que el precio de ejercicio al
vencimiento. Noten que es el complemento de N (d 2 ) , la probabilidad que el precio final de la mercancía
exceda el precio de ejercicio. El valor actual del costo esperado de ejercer la put condicionado a que el
vencimiento la put esté in-the-money es Se ( br )T N (d1 ) . Si ejercemos la put, debemos entregar el subyacente
como cumplimiento de nuestra obligación de manera que el valor actual del precio final esperado de la
mercancía condicional al ejercicio sea nuestro costo hoy.
EJEMPLO 11.8
Calcular el precio de una put europea de moneda extranjera a tres meses con un precio de ejercicio de 40. El
tipo de cambio spot actual es 40, la tasa de interés doméstica es 8 porciento anual, la tasa de interés
extranjera es 12 porciento anual, y la desviación estándar del retorno compuesto continuamente es 30
porciento sobre una base anualizada. Noten que la tasa de costo de acarreo, b , es por lo tanto, -4 porciento.
Los valores de N (d1 ) y N (d 2 ) son 0.5563 y 0.4967, respectivamente, de manera que el precio de la put
europea es
Noten que este resultado, junto con el resultado del ejercicio 11.7 verifica la relación de paridad put-call
(11.26), esto es,
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 183/358
ln( S / X ) (b 0.5 2 )T
respectivamente, donde d1 y d 2 d1 T
T
El precio de la opción depende de seis variables – S , X , b, r , , T . En esta sección, analizamos cómo los
precios de la call y put europea responden a los cambios en las variables que definen la opción. Se discutirá
cada una de las variables, comenzando por el precio de la mercancía. Las derivaciones de cada una de las
expresiones a continuación están contenidas en el Apéndice 11.4.
c
c e ( br )T N (d1 ) 0 (11.29a)
S
El delta de la call es positivo, lo que implica que un incremento en el precio de la mercancía causa que se
incremente el precio de la Call. El resultado es intuitivo dado que la call conlleva el derecho a comprar la
mercancía subyacente a un precio fijo y la mercancía subyacente recién se ha vuelto más valiosa. La Figura
11.4 muestra cómo cambia el valor de la call europea conforme cambia el precio de la mercancía subyacente.
La opción tiene tres meses hasta el vencimiento. Noten que cuando la call está out-of-the-money, su
pendiente es casi horizontal. Las calls out-of-the-money tienen de valores delta muy bajos; esto es, no
responden rápidamente a los cambios en el precio de la mercancía. Conforme el precio de la mercancía se
incrementa y la call se vuelve at-the-money y luego in-the-money, la pendiente se hace cada vez más
empinada. Cuando la opción está muy in-the-money, el valor delta es casi uno, y el precio de la call cambia
en una proporción uno-a-uno con el precio de la mercancía. La Figura 11.5 muestra el valor delta de la
opción como una función del precio de la mercancía. El delta de la put es
p
p e (br )T N (d1 ) 0 (11.29b)
S
Esta derivada es negativa porque un incremento en el precio de la mercancía hace que la put sea menos
valiosa. Una vez más, se puede demostrar que la sensibilidad del precio de la put ante los cambios en el
precio de la mercancía subyacente es en sí mismo sensible al “moneyness” de la opción. La Figura 11.6
muestra esta sensibilidad para una put europea con tres meses hasta el vencimiento. La opción in-the-money
tiene una pendiente más empinada que la opción at-the-money, la cual a su vez, tiene una pendiente más
empinada que la opción out-of-the-money. La Figura 11.7 muestra el valor delta de la put como una función
del precio de la mercancía subyacente.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 184/358
Figura 11.4 Precio de una call europea ( c )como una función del precio de la mercancía subyacente ( S ). El
rango del precio de la mercancía es de 50 a 150. La opción tiene un precio de ejercicio ( X ) de 100 y un
tiempo hasta el vencimiento ( T ) de tres meses. La tasa de costo de acarreo (b) es 8 porciento y la tasa de
interés libre de riesgo ( r ) es 8 porciento. La desviación estándar del logaritmo de los ratios del precio de la
mercancía ( ) es 30 porciento.
Call and Lower Bound
60
50
40
30
20
10
Commodity P rice s (S )
FIGURA 11.5 El delta ( c ) de la call europea como una función del precio de la mercancía subyacente ( S ).
El rango del precio de la mercancía está entre 50 y 150. La opción tiene un precio de ejercicio ( X ) de 100 y
un tiempo al vencimiento ( T ) de tres meses. La tasa del costo de acarreo ( b ) es 8 porciento y la tasa de
interés libre de riesgo ( r ) es 8 porciento. La desviación estándar del logaritmo de los ratios del precio de la
mercancía ( ) es 30 porciento.
Call Delta
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
50 70 90 110 130 150
Commodity P rice (S )
96
Se puede demostrar que la elasticidad del precio de la call con respecto al precio de la mercancía es mayor que uno al re-escribir
(11.30a) como c 1
rT
Xe N ( d 2 )
1
Se( b r )T N ( d1 )
El último término en el denominador es menor que uno porque el precio de la call europeo no puede ser menos que cero, por lo
tanto, el valor de c , debe ser mayor que uno.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 185/358
S S
c e ( br )T N (d1 ) 1 (11.30a)
c c
y la elasticidad del precio de la put con respecto al precio de la mercancía es menos que uno,97 esto es
S S
p e ( br )T N (d1 ) 1 (11.30b)
p p
FIGURA 11.6 El precio de la put europea ( p ) como una función del precio de la mercancía subyacente ( S ).
El rango del precio de la mercancía está entre 50 y 150. La opción tiene un precio de ejercicio ( X ) de 100 y
un tiempo al vencimiento ( T ) de tres meses. La tasa del costo de acarreo ( b ) es 8 porciento y la tasa de
interés libre de riesgo ( r ) es 8 porciento. La desviación estándar del logaritmo de los ratios del precio de la
mercancía ( ) es 30 porciento.
50
45
40
35
30
25
20
15
10
5
0
0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-0.8
-0.9
-1
Commodity P rice (S )
97
La evidencia que p 1 se deduce según las mismas líneas que la evidencia que c 1 en la nota a pie de página anterior.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 186/358
La Tabla 11.1 contiene los precios de las opciones, los valores delta y las elasticidades para precios
alternativos de la mercancía subyacente. Es interesante notar que (i) las elasticidades tiene magnitudes
grandes y (ii) las magnitudes son mayores para las opciones más out-of-the-money. Si alguien está casi
seguro que el precio de una mercancía individual se va a elevar, una inversión en una call proporcionará una
tasa de retorno mayor que una inversión directamente en la mercancía y, más aún, una inversión en una call
out-of-the-money proporcionará una mayor tasa de retorno que una call in-the-money.
Tabla 11.1 Simulación de Precios de Calls y Puts de Índice de Acciones, deltas, etas y gammas de
los precios de las opciones. Parámetros: X = 100, b = 0.08, r = 0.08, T = 0.25, y = 0.30.
Precio de
Precio de Delta Eta Gamma Precio de Delta Eta Gamma
la
la Call c c c la Put p p p
Mercancía p
c
S
80 0.537 0.100 14.952 0.014 18.557 -0.899 -3.878 0.014
90 2.494 0.310 11.207 0.026 10.514 -0.689 -5.900 0.026
100 6.961 0.582 8.367 0.026 4.981 -0.417 -8.380 0.026
110 13.954 0.800 6.310 0.016 1.974 -0.199 -11.109 0.016
120 22.645 0.922 4.889 0.008 0.665 -0.077 -13.924 0.008
Sin embargo, estas tasas de retornos mayores no vienen sin un correspondiente incremente en el riesgo. De
hecho, así como las tasas de retorno sobre las opciones están proporcionalmente relacionadas con la tasa de
retorno sobre la mercancía, el riesgo o “beta” de una opción está proporcionalmente relacionado con el beta
de la mercancía, esto es, c c S y p p S . El incremento en la tasa de retorno esperada como
resultado de mantener una call es exactamente lo que se justifica sobre la base del modelo de precio del
activo de capital.
c e (br )T n(d1 )
c 0 (11.31a)
S S T
p e (br )T n(d1 )
p c 0 (11.31b)
S S T
donde n(d1 ) es la densidad en d1 . En resumen, este valor nos dice cuán rápidamente cambia delta conforme
cambia el precio de la mercancía. Dado que el gamma de una opción es mayor cuando las opciones están
aproximadamente at-the-money, estas opciones son las más difíciles de proteger. Además, si usted piensa
que el precio de la mercancía está por moverse en una u otra dirección (recuerden que la motivación para
colocar un margen de volatilidad), el margen at-the-money maximizará la respuesta en dólares de la cartera a
los movimientos del precio de la mercancía subyacente. La Figura 11.8 muestra el gamma de la opción como
una función del precio de la mercancía subyacente.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 187/358
y
p
e rT N (d 2 ) 0 , (11.32b)
X
respectivamente. Noten que si el precio de ejercicio de las opciones se incrementa el valor de la call
disminuye y el valor de la put se incrementa. Esto se deduce del hecho que la call se pondría más out-of-the-
money y la put más in-the-money.
Las derivadas parciales de los precios de la opción, con respecto al precio de ejercicio, son de muy
poco valor práctico en el sentido que una vez creada la opción, el precio de ejercicio no cambia. Aquí sólo
son expresadas para completar la idea.
FIGURA 11.8 El gamma ( c p ) de la opción europea como una función del precio de la mercancía
subyacente ( S ). El rango del precio de la mercancía está entre 50 y 150. La opción tiene un precio de
ejercicio ( X ) de 100 y un tiempo al vencimiento ( T ) de tres meses. La tasa del costo de acarreo ( b ) es 8
porciento y la tasa de interés libre de riesgo ( r ) es 8 porciento. La desviación estándar del logaritmo de los
ratios del precio de la mercancía ( ) es 30 porciento.
0.03
0.025
0.02
0.015
0.01
0.005
0
50 70 90 110 130 150
Commodity P rice (S )
Conforme se incrementa la tasa del costo de acarreo, el valor de la call se incrementa, manteniendo constante
el precio spot y otras variables. Mientras mayor sea el costo de acarrear la mercancía subyacente, mayor será
la tasa de apreciación en el precio de la mercancía y por lo tanto mayor será el valor de la call. Sin embargo,
la magnitud de la derivada es pequeña. Para la call de moneda extranjera valuada en el Ejemplo 11.7, la
derivada parcial con respecto a la tasa del costo de acarreo es 4.884. En otras palabras, si la tasa del costo de
acarreo sobre la mercancía subyacente se incrementa en 100 puntos base, el precio de la call se incrementará
en aproximadamente cinco centavos. La derivada parcial del precio de la put con respecto a la tasa del costo
de acarreo es
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 188/358
p
TSe ( br )T N (d1 ) 0 . (11.33b)
b
Conforme se incrementa la tasa del costo de acarreo, la tasa de apreciación esperada en el precio de la
mercancía se incrementa y, por lo tanto, disminuye el valor de la put. El valor numérico de esta derivada
parcial para la put en el Ejemplo 11.8 es –4.8200.
c
TXe rT N (d 2 ) 0 . (11.34a)
r
El precio de la call se incrementa con un incremento en la tasa de interés porque el valor actual del precio de
ejercicio disminuye. El valor de esta derivada es 4.3489 para la call en el Ejemplo 11.7. La derivada parcial
del precio de la put con respecto a la tasa de interés libre de riesgo es
p
TXe rT N (d 2 ) 0 . (11.34b)
r
Aquí el signo es negativo porque, conforme se incrementa la tasa de interés libre de riesgo, disminuye el
valor actual del precio de ejercicio recibido al ejercer la opción. El valor de la derivada para la put en el
Ejemplo 11.8 es –5.4531, lo que implica que un incremento de 100 puntos base en la tasa de interés reduce el
valor de la opción en casi cinco centavos.
El signo de la derivada es positivo, indicando que conforme se incrementa la volatilidad del retorno de la
mercancía subyacente, se incrementa el valor de la call. La intuición para este resultado es que un incremento
en la tasa de volatilidad incrementa la probabilidad de grandes movimientos ascendentes en el precio de la
mercancía subyacente. La probabilidad de grandes movimientos descendentes en el precio de la mercancía se
incrementa; sin embargo, no es de consecuencia ya que el tenedor de la opción tiene responsabilidad
limitada.
El valor numérico del vega de la call implica que el precio de las opciones es más sensible a la
volatilidad que a la tasa del costo de acarreo o la tasa de interés. La opción en el Ejemplo 11.7 tiene un vega
de 7.7428. Un incremento en la volatilidad de 100 puntos incrementa el precio de la call en casi ocho
centavos.
El vega de la put es el mismo que el de la Call, esto es
p
Vega p Se (br )T n(d1 ) T Vegac 0 . (11.35b)
98
Hasta este punto, hemos utilizado el término “desviación estándar” para describir la dispersión de los retornos de la mercancía,
. En la industria, a normalmente se le conoce como la volatilidad o la tasa de volatilidad de los retornos de la mercancía
subyacente, y nosotros adoptamos esa idea en el resto del capítulo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 189/358
El valor de la put también se incrementa con un incremento en la volatilidad dado que se incrementa la
probabilidad de una gran disminución en el precio de la mercancía. El valor numérico del vega para la put en
el Ejemplo 11.8, por lo tanto, también es 7.7428.
La Figura 11.9 muestra el vega de la opción como una función del precio de la mercancía. Noten que
el vega tiene su valor más alto cuando la opción está aproximadamente at-the-money.
La expresión muestra que la sensibilidad del precio de la call ante cambios en el tiempo hasta el vencimiento
de la opción es la suma de tres componentes. El primer término al lado derecho es positivo y refleja el
incremento en el precio de la opción debido al hecho que un incremento en el tiempo hasta el vencimiento
incrementa la probabilidad de movimiento ascendentes en el precio de la mercancía y, por lo tanto,
incrementa el valor de la opción. El segundo término puede ser positivo o negativo dependiendo de si la tasa
de costo de acarreo, b , es mayor o menor que la tasa de interés, r . Si b r , el término es positivo dado que
conforme se incrementa el tiempo hasta el vencimiento, el valor actual del precio final esperado de la
mercancía aumenta (recuerden que el precio de la mercancía subyacente aumenta a la tasa b mientras que la
tasa de descuento del valor final de la opción es r ). Finalmente, el tercer término es positivo. Conforme se
incrementa el tiempo hasta el vencimiento, el valor actual del precio de ejercicio se hace más pequeño. Noten
que el único caso donde el valor global de theta es ambiguamente positivo es cuando b r .
Para la call en el Ejemplo 11.7, b r , de manera que sabemos que el theta no necesita ser positivo.
El valor de theta es, no obstante, positivo en 3.6927. En otras palabras, el precio de la opción se incrementa
conforme se incrementa el tiempo hasta el vencimiento. Para ver el origen de este resultado, examinamos los
valores de cada uno de los tres términos en la derivada: 4.6457, -2.3446 y 1.3916. El mayor componente del
theta de la call en esta ilustración, 4.6457, viene de una mayor probabilidad de grandes movimientos en el
precio. Dado que b r , el segundo término es negativo. Conforme aumenta el tiempo hasta el vencimiento,
el valor de la call cae debido a que se espera que el precio de la mercancía se incremente a una tasa menor
que la tasa de interés libre de riesgo. El valor de este componente es –2.3466. Finalmente, el valor del tercer
término es 1.3916, indicando que el valor de la call se incrementa porque el valor actual del precio del
ejercicio se reduce conforme aumenta el tiempo hasta el vencimiento.
Theta proporciona información sobre el deterioro del valor de la opción conforme pasa el tiempo
hasta el vencimiento. El theta de la call en el Ejemplo 11.7 es 3.6927. Esto implica que la disminución de
tiempo en esta opción es 3.6927 x 1/365 o ligeramente algo más que un centavo por encima del día siguiente
y 3.6927 x 7/365 o casi siete centavos por encima de la próxima semana, manteniendo los otros factores
constantes.
El theta de la put europea es
p
c Se( br )T n(d1 ) (b r ) Se ( br )T N (d1 ) rXe rT N (d 2 ) 0 . (11.36b)
T 2 T
Existe un paralelo entre la interpretación de los términos en la expresión del theta de la put y aquellos de la
call. El primer término es el incremento en el valor de la put resultante de la perspectiva de grandes
movimientos en el precio de la mercancía cuando el tiempo hasta el vencimiento es más largo. El segundo
término es negativo si b r . En el caso de la put, el valor de la opción se incrementa cuando la tasa del costo
de acarreo está por debajo de la tasa de interés. El tercer término es positivo. Refleja el hecho que un
incremento en el tiempo hasta el vencimiento demora la recepción del precio de ejercicio y por lo tanto
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 190/358
reduce el valor de la put. El valor del theta de la put en el Ejercicio 11.8 es 5.2143, siendo los componentes
individuales de la suma 4.6457, 2.3136 y –1.7450.
Figura 11.9 El vega de la opción europea como una función del precio de la mercancía subyacente ( S ). El
rango del precio de la mercancía está entre 50 y 150. La opción tiene un precio de ejercicio ( X ) de 100 y un
tiempo al vencimiento ( T ) de tres meses. La tasa del costo de acarreo ( b ) es 8 porciento y la tasa de interés
libre de riesgo ( r ) es 8 porciento. La desviación estándar del logaritmo de los ratios del precio de la
mercancía ( ) es 30 porciento.
35
30
25
20
15
10
5
0
50 70 90 110 130 150
Commodity Price (S)
d 2 d1 T (11.37b)
1 2 2 1, 2 1 2
2 2 2
(11.37c)
Las variables con subíndice 1 se aplican a la mercancía 1 y las variables con subíndice 2 se aplican a la
mercancía 2. El significado de cada variable se describió anteriormente en este capítulo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 191/358
Una observación importante con relación a (11.37) es que fórmula de call descrita anteriormente en
este capítulo es simplemente un caso de esta ecuación de valuación. Supongamos que permitimos que la
mercancía 2 sea un activo libre de riesgo. Por lo tanto, el precio actual de la mercancía 2, S 2 , es Xe rT , la
tasa del costo de acarreo, b2 , la tasa de interés libre de riesgo, r y la desviación estándar del retorno, 2 , es
igual a cero. Con estas sustituciones, la ecuación (11.37) se convierte en la fórmula de la call europea
(11.25).
Otra observación importante con relación a (11.39) es que el valor de una call para “comprar” la
mercancía 1 con la mercancía 2, c( S1 , T ; S 2 ) , es igual al valor de una put para “vender” la mercancía 2 por la
mercancía 1, p ( S 2 , T ; S1 ) . En el caso de una call, la opción se ejerce al vencimiento si los ingresos
S1,T S 2,T 0 , esto es, si el precio final para la mercancía 1 excede el precio final de la mercancía 2; de otra
manera, no se ejerce. En el caso de un put, la opción se ejerce al vencimiento si los ingresos S1,T S 2,T 0 ;
esto es, si la mercancía 2 es más barata que la mercancía 1; de otra manera, no se ejerce. Pero las estructuras
de estos dos problemas de valuación son idénticas, entonces
c( S1 , T ; S 2 ) p ( S 2 , T ; S1 ) . (11.38)
Regresando a la especificación del contrato de futuros de bonos-T, recuerden que al final del Capítulo 8
dijimos que el precio de los futuros es igual al precio del bono a entregarse más barato menos el valor de la
opción de calidad. Si el contrato de futuros de bonos-T tiene sólo dos emisiones de bonos-T elegibles para
entregarse, la fórmula de valuación (11.38) puede utilizarse para valorizar la opción de calidad. Con más
emisiones elegibles, el modelo se debe generalizar.99
EJEMPLO 11.9
Supongamos que existen dos bonos elegibles para entregarse sobre un contrato de futuros de bonos-T. El
tiempo al vencimiento de los futuros es de tres meses. Actualmente el Bono 1 es el más barato de entregar.
Su precio es de 99 y su cupón es 6 porciento. El Bono 2 tiene un precio de 102 y su cupón es 9 porciento. La
desviación estándar del retorno compuesto continuamente es 15 porciento para el Bono 1 y 12 porciento para
el Bono 2. La correlación entre sus tasas de retorno es 0.9. La tasa de interés libre de riesgo es 7 porciento.
Calcular el valor de la opción de intercambio. Por simplicidad, asuman que ambos bonos tiene factores de
conversión iguales a uno y que el interés del cupón es pagado continuamente a lo largo de la vida restante del
contrato de futuros
p( B2 , T ; B1 ) 99e ( 0.070.060.07 ) 0.25 N (d1 ) 102e ( 0.070.090.07 ) 0.25 N (d 2 )
99
La fórmula de la opción de intercambio para el caso de dos activos donde ambos activos tiene una tasa de costo de acarreo igual
a la tasa de interés libre de riesgo fue derivada por Margrabe (1978). La fórmula aquí presentada generaliza el resultado Margrabe
para permitir que los activos tengan diferentes tasas de acarreo. La opción de intercambio de n activos fue desarrollada más
tarde por Margrabe (1982).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 192/358
C ( S , T ; X ) c( S , T ; X ) C ( S , T ; X ) , (10.9a)
P( S , T ; X ) p( S , T ; X ) P ( S , T ; X ) . (1 0.9b)
El valor del privilegio de ejercicio antes del vencimiento, C ( S , T ; X ) , depende de la relación entre la tasa de
costo de acarreo, b , y la tasa libre de riesgo, r . En el caso de una call, el privilegio de ejercicio temprano o
antes del vencimiento tiene valor sólo si b r . En este caso, el costo de acarreo de la mercancía subyacente
es menor que el costo de los fondos atados en la mercancía. Como resultado, puede ser beneficioso ejercer la
call y tomar posesión de la mercancía porque las ganancias sobre la mercancía exceden el costo de los
fondos atados en la mercancía. Por ejemplo, puede ser deseable ejercer la call sobre moneda extranjera para
obtener un interés sobre la moneda extranjera si la tasa de la moneda extranjera excede la tasa
estadounidense. Si b r , el ejercicio temprano de una call no es óptimo porque existe el costo de mantener
la mercancía.
Al continuar manteniendo la call, también se pueden lograr todas las ganancias potenciales que se
podrían obtener por mantener la mercancía , y se evita el costo de mantener la mercancía. Por ejemplo, nunca
es óptimo ejercer antes del vencimiento una opción sobre una acción que no paga dividendos (es decir,
b r ).
En el caso de la put, el ejercicio temprano siempre es una posibilidad. Intuitivamente, el ejercicio
temprano es deseable si la utilidad de la put es lo suficientemente grande como para que el interés que se
pudiera obtener al invertir la utilidad exceda la posibilidad de una utilidad aún mayor si se continúa
manteniendo el put.
Se desconocen las soluciones analíticas explícitas para el precio de las opciones americanas. Si la
opción americana no se ejerce temprano, se mantienen las fórmulas para las opciones europeas. Pero si es
deseable un ejercicio temprano, el valor de la opción americana excede el valor europea por una cantidad
(con frecuencia bastante pequeña) que sólo puede ser aproximada por técnicas numéricas. En la práctica
comúnmente se aplican dos técnicas de aproximación que son el método del binomio del Cox, Ross y
Rubinstein (1979), y el método de aproximación cuadrática de Barone-Adesi y Whaley (1987). El método
del binomio se utiliza en el Capítulo 13 para valorizar una put sobre una acción que paga dividendos y en el
Capítulo 14 se utiliza la aproximación cuadrática para valorizar opciones sobre el índice de acciones y
futuros del índice de acciones.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 193/358
parámetro sobre el cual más comúnmente están en desacuerdo los inversionistas) es la desviación estándar de
la tasa de retorno de la mercancía subyacente. En general, se utilizan dos métodos –el cálculo de la
volatilidad histórica y el cálculo de la volatilidad implícita.
donde T es el número de observaciones del retorno de las series de tiempo en el cálculo,101 Rt es la tasa de
retorno compuesta continuamente sobre la mercancía en el mes t [es decir, ln(St / St 1 ) ], y ̂ es el cálculo de
la tasa de retorno media,
1 T
̂ t 1 Rt (11.40)
T
Se puede obtener un estimado de la desviación estándar del retorno sobre la mercancía sacando la raíz
cuadrada de la varianza estimada, esto es, ˆ h ˆ h2 .
Las tasas de retorno utilizadas en las ecuaciones (11.39) y (11.40) pueden durar cualquier período –
un día, una semana o un mes. En general, es mejor mientras más corta sea la distancia entre las
observaciones de precios ya que se utiliza más información para el cálculo. De manera que los retornos
semanales son efectivamente mejores que los retornos mensuales para el cálculo de la volatilidad,
manteniendo constante la duración global del período de cálculo.
Siguiendo la misma lógica, parecería que los retornos diarios son una mejor fuente de información
que los retornos semanales. Sin embargo, generalmente este no es el caso. Por ejemplo, los retornos de las
acciones demuestran estacionalidad por día de la semana.102 En general, los retornos de las acciones de los
viernes son significativamente mayores que el promedio, y los de los lunes son significativamente menores
que el promedio. Otras mercancías también tienen estacionalidad por día de la semana, pero la estacionalidad
tiene una estructura diferente. Es más, independiente de la mercancía subyacente, utilizando datos diarios, el
investigador decide cómo se deben tomar en cuenta los retornos del fin de semana. ¿La tasa de retorno del
cierre del viernes al cierre del lunes debe tomarse en cuenta como una tasa de retorno de tres días (días
calendario) o una tasa de retorno de 1 día (días comerciales)? Dadas las anomalías empíricas asociadas con
los retornos diarios y dado que el problema del manejo de los retornos del fin de semana no se han resuelto
satisfactoriamente, probablemente los retornos semanales son los mejores cuando se trata de calcular la
volatilidad histórica de los retornos de las mercancías. Otro punto que surge cuando se utiliza un factor
histórico se relaciona con cuántas observaciones de retornos utilizar en el cálculo de la volatilidad. Por un
lado, mientras más información se utilice en el proceso de cálculo, más preciso es el cálculo. Por otro lado,
mientras más largo sea el período sobre el cual se calcula la volatilidad, mayor será la probabilidad que se
viole la premisa estacionaria, en cuyo caso, la fórmula (11.39) ya no es un cálculo no parcializado de la
101
Noten que se necesitan T + 1 observaciones de precio pata generar las tasas de retorno T .
102
Ver, por ejemplo, French (1980) y Gibbons y Hess (1981).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 194/358
varianza de la tasa de retorno de la mercancía. En la ausencia de la información que indica que se ha violado
la premisa estacionaria, veintiséis semanas de observaciones de los retornos probablemente son suficientes
para asegurar un cálculo razonablemente exacto de la volatilidad.
El cálculo de la volatilidad realizado utilizando la ecuación (11.39) calcula la varianza de la tasa de
retorno para el intervalo de tiempo entre las observaciones de precios utilizadas para calcular las tasas de
retorno. Por lo tanto, si se utilizan retornos semanales, el cálculo de la varianza de (11.39) es la varianza de
la tasa de retorno sobre una semana. Para anualizar este valor, tenemos que multiplicar la varianza por el
número de semanas en el año, esto es, ha 2
52 hw
2
, donde los subíndice w y a denotan la semana y el año,
respectivamente. Por lo tanto, la transformación para la anualización de la desviación estándar semanal es
ha 52 hw .103
Vale la pena notar un último punto con relación al cálculo de la volatilidad histórica. Los factores de
cálculo arriba mostrados generalmente utilizar información de precio cierre-a-cierre al generar las tasas de
retorno. Esto ha sido aceptado como una práctica común ya que tradicionalmente la historia registrada de los
precios de la mercancía son precios reportados para la última transacción del día. Con el surgimiento de la
tecnología de cómputo y base de datos sofisticada, ahora se ha vuelto más fácil registrar y mantener
conjuntos de información más grandes, con la mayoría de las bolsas de valores y mercancías manteniendo
archivos de los precios de las transacciones. Esta información más refinada permite un cálculo más preciso
de la volatilidad. Por ejemplo, Parkinson (1980) y Garman y Klass (1980) desarrollaron factores de cálculo
alternativos para la varianza que utilizan los precios alto, bajo y de cierre de la mercancía y muestran que
estos factores de cálculo son ocho veces “mejores” que el factor de cálculo tradicional (11.39).
V j Vˆ j ( j ) (11.41)
donde V es el precio observado de la opción, V es el precio modelo de la opción y hay que despejar . Se
puede derivar una expresión analítica para el parámetro de variación; sin embargo, la aproximación exacta es
posible mediante “prueba-y-error”, casi de la misma manera en que se despeja el rendimiento al vencimiento
de un bono. Las volatilidades calculadas de esta manera son llamadas “volatilidades implícitas” o
“desviaciones estándar implícitas”. Pueden ser interpretadas como la volatilidad consenso del mercado en el
sentido que se utiliza el precio de mercado de la opción para imputar el cálculo de volatilidad. Si alguien
considera todas las opciones escritas sobre una mercancía dada, parecería razonable pensar que todas
rendirán el mismo estimado de volatilidad sobre la mercancía subyacente. Sin embargo, este no es el caso.
Existe una variedad de razones que causan que los estimados sean diferentes.
103
Para transformar el cálculo de volatilidad estimado utilizando retornos diarios a una volatilidad anual, el cálculo diario
usualmente se multiplica por la raíz cuadrada del número de días comerciales en el año (típicamente, 253 ), en lugar que el
número de días calendario en el año ( 365 ). El motivo para este ajuste es que los estudios de los retornos diarios de las acciones
indican que la volatilidad del retorno del cierre del viernes al cierre del lunes (tres días) es casi igual que la volatilidad de cierre-a-
cierre durante cualquier otro par de días comerciales adyacentes (un día). Ver, por ejemplo, Stoll y Whaley (1990a). Por lo tanto, el
tratar los fines de semana como un solo día de negociación proporciona el ajuste más apropiado para las volatilidades de los
retornos diarios de las acciones. Sin embargo, la evidencia empírica relacionada con la volatilidad de fin de semana de los
mercados son acciones es escasa, de manera que la generalidad de este resultado para otras mercancías todavía es desconocida.
Para los mercados sin acciones, un procedimiento más seguro podría ser utilizar los retornos semanales, como vimos anteriormente
en esta sección.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 195/358
No-Simultaneidad de los Precios. El cálculo de volatilidad implícita utilizando (11.41) asume que el
precio de la opción y el precio de la mercancía son observados en el mismo instante en el tiempo. Con
frecuencia, se da el caso que la única información disponible es el precio de la opción y la mercancía en
momentos en que fueron negociados por última vez. Es poco probable que estas negociaciones, una en el
mercado de opciones y una en el mercado de mercancías, ocurrieran en el mismo instante, hasta el punto en
que no son contemporáneas, habría un error en el cálculo de la volatilidad.
Precios de Oferta y Demanda. Inclusive si las observaciones en el precio de la opción y la mercancía
utilizadas en (11.41) son simultáneas, existe un problema sobre qué representan los precios. Si los mercados
son perfectamente líquidos y libres de fricción, las negociaciones son liquidadas al precio de equilibrio del
título valor. Sin embargo, ninguno de los descriptores es verdadero. Los elaboradores de mercado
proporcionan liquidez manteniéndose listos para inmediatamente compra o vender título valores. Dado que
los elaboradores de mercados tienen capital (tanto de inversión como humano) ocupado en sus operaciones,
ellos exigen una tasa de retorno sobre su capital, la cual extraen estableciendo un precio de oferta del título
valor por debajo del precio solicitado. Por lo tanto, cuando se ejecutan las órdenes del mercado están en el
precio ofertado o solicitado, dependiendo de si la persona que ingresa al mercado quiere comprar o vender.
Dado que no hay manera de discernir la motivación del negociador que estuvo involucrado en la última
transacción observada, los cálculos de la volatilidad implícita tienen error cuando el precio de oferta de la
opción es igual al precio solicitado por la mercancía y viceversa.
Los esquemas de ponderación utilizados en al literatura han sido muchos y variados. Schmalensee y Trippi
(1978) y Patell y Wolfson (1979), por ejemplo, utilizan igual promedio ponderado, j 1 / n, j 1, , n .
Su motivación para hacerlo es que cada cálculo de volatilidad es igualmente valiosos en la
determinación de la volatilidad global para la mercancía. Latane y Rendleman (1976), por otro lado,
ponderan de acuerdo a la derivada parcial del precio call con respecto a la desviación estándar del retorno de
la mercancía, esto es, V j / ˆ j , j 1,, n . Al hacerlo, los cálculos de la desviación estándar de las opciones
que son teóricamente más sensibles al valor de son sopesados con mayor peso que aquellos que no lo son.
Chiras y Manaster (1978) siguen una lógica similar al utilizar la elasticidad del precio de la call con respecto
a la desviación estándar, V jˆ j / ˆ jV j , j 1,, n . Desafortunadamente este esquema tiene serias
deficiencias. Utilizar la elasticidad como la base del esquema de ponderación implica que los cálculos de la
volatilidad para las opciones out-of-the-money reciben el peso más alto. Las opciones out-of-the-money
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 196/358
generalmente no producen estimados de volatilidad muy exactos porque los mercados para estas opciones
son relativamente ilíquidos (induciendo a problemas serios de no-simultaneidad) y las opciones mismas
tienen márgenes altos de oferta-demanda (induciendo a errores oferta-demanda). Finalmente, Whaley (1982)
utiliza una regresión no lineal para calcular un valor de utilizando simultáneamente toda la información de
precios de opciones disponible, esto es
V j Vˆ j ( ) j (11.43)
Las propiedades del estimador de probabilidad máxima de (11.43) son, tal vez, las alternativas que mejor se
entienden.
Sin embargo, sin importar el esquema de ponderación, parece existir un fuerte respaldo empírico a
favor de la medida de la volatilidad implícita. Latane y Rendleman y Chiras y Manaster correlacionan las
medidas históricas e implícitas sobre la desviación estándar real del retorno y concluyen que el cálculo
implícito es un pronosticador notablemente superior. Aparentemente el mercado utiliza más información que
un simple estimado histórico al evaluar la volatilidad esperada de la mercancía.
11.9 RESUMEN
En este capítulo se han derivado en detalle las ecuaciones de precio de las opciones europeas sobre diferentes
tipos de mercancías subyacentes. El capítulo comienza con una discusión intuitiva del enfoque de valuación
neutral al riesgo utilizado para derivar las fórmulas de precio. Luego se describieron las distribuciones de
retorno y precio asumidas. En la sección 3, la valuación neutral al riesgo de la opción europea es llevado a
cabo en detalle y se muestran varias ecuaciones de valuación básica para diversos tipos de mercancías
subyacentes. Luego utilizando la paridad put-call, se derivaron las ecuaciones de valuación.
El precio de una opción sobre una mercancía depende del precio spot de la mercancía, el precio de
ejercicio de la opción, el costo de acarreo de la mercancía, la tasa libre de riesgo, la desviación estándar del
retorno de la mercancía y el tiempo hasta el vencimiento. En la sección 5, se analizó el efecto de los cambios
en cada una de estas variables sobre el precio de la opción.
La sección 6 presenta la ecuación de valuación que permite al tenedor intercambiar una mercancía
riesgosa por otra. Esta opción, llamada opción de intercambio, es introducida en muchos tipos de contratos
de futuros y se introdujo en capítulos anteriores. Se utiliza como ejemplo la valuación de una opción de
entrega introducida en un contrato de futuros de bonos del Tesoro (bono-T).
Los factores subyacentes al valor de una opción americana son los mismos subyacentes a una opción
europea excepto que la opción americana tiene el beneficio adicional del ejercicio antes del vencimiento. La
sección 7 nombra dos enfoques populares para la valuación de opciones americanas. Los métodos se
describen en detalle en los Capítulos 13 y 14.
En la práctica, la variable más importante que afecta el precio de una opción es la volatilidad de la
mercancía subyacente. La sección 8 explica los dos enfoques para calcular la volatilidad histórica y la
volatilidad implícita.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 197/358
APPENDIX 11.1
~
Proof that E ( ST ) S 0 eT , 2 / 2 where and are the mean and the variance of the normally
distributed continuously compounded rate of return.
Begin by rewriting the expected terminal price as the expected price relative,
~ ~
E ( ST / S 0 ) eT E (e x ) , (Al.1)
where ~x is the normally distributed, continuously compounded rate of return from 0 through T . ~
x can be re-
expressed in terms of , , and the unit normally distributed variable z . Using (11.10),
~ ~ T ~z
ST / S 0 e x e T
T~ T ~z
eT E (e T z
) e T E (e ) (Al.2)
T ~z
The term E (e ) in (Al.2) may be simplified as follows:
T z z2 / 2 T zz2 / 2 2
T / 2 2T / 2 T z z 2 / 2
T ~z e e e e
E (e ) dz dz dz
2
2
2
2
T / 2 T z z 2 / 2 T z )2 / 2
2T / 2 e 2T / 2 e (
2
e dz e dz e T /2
(Al.3)
2
2
Substituting (Al.3) into (Al.2), taking the logarithm of both sides, and then factoring T gives
2 / 2 (Al.4)
The probability that a drawing from a unit normal distribution will produce a value less than the constant d
is
z2 / 2
d e
~
Pr( z d ) N (d ) dz
2
Below are two polynomials that provide reasonably accurate approximations for the above integral.
Approximation 1
2
N ( d ) a0 e d /2
(a1t a2t 2 a3t 3 )
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 198/358
With this approximation method, the value of d must be greater than or equal to 0. The maximum absolute
error of this approximation method is 0.00001.
Approximation 2
2
N ( d ) a0 e d /2
(a1t a2t 2 a3t 3 a4t 4 a5t 5 )
With this approximation method, the value of d must be greater than or equal to 0. The maximum absolute
error of this approximation method is 0.000000075.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 199/358
The probability that a drawing from a unit normal distribution will produce a value less than the constant d is
z2 / 2
d e
~
Pr( z d ) N (d ) dz
2
Range of d : -4.99 d -2.50
d -0.00 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 -0-08 -0.09
-4.90 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.80 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.70 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.60 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.50 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.40 0.00001 0.00001 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
-4.30 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001
-4.20 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001
-4.10 0.00002 0.00002 0.00002 0.00002 0.00002 0.00002 0.00002 0.00002 0.00001 0.00001
-4.00 0.00003 0.00003 0.00003 0.00003 0.00003 0.00003 0.00002 0.00002 0.00002 0.00002
-3.90 0.00005 0.00005 0.00004 0.00004 0.00004 0.00004 0.00004 0.00004 0.00003 0.00003
-3.80 0.00007 0.00007 0.00007 0.00006 0.00006 0.00006 0.00006 0.00005 0.00005 0.00005
-3.70 0.00011 0.00010 0.00010 0.00010 0.00009 0.00009 0.00008 0.00008 0.00008 0.00008
-3.60 0.00016 0,00015 0.00015 0.00014 0.00014 0.00013 0.00013 0.00012 0.00012 0.00011
-3.50 0.00023 0.00022 0.00022 0.00021 0.00020 0.00019 0.00019 0.00018 0.00017 0.00017
-3.40 0.00034 0.00032 0.00031 0.00030 0.00029 0.00028 0.00027 0.00026 0,00025 0.00024
-3.30 0.00048 0.00047 0.00045 0.00043 0.00042 0.00040 0.00039 0.00038 0.00036 0.00035
-3.20 0.00069 0.00066 0.00064 0.00062 0.00060 0.00058 0.00056 0.00054 0.00052 0.00050
-3.10 0.00097 0.00094 0.00090 0.00087 0.00084 0.00082 0.00079 0.00076 0.00074 0.00071
-3.00 0.00135 0.00131 0.00126 0.00122 0.00118 0.00114 0.00111 0.00107 0.00104 0.00100
-2.90 0.00187 0.00181 0.00175 0.00169 0.00164 0.00159 0.00154 0.00149 0.00144 0.00139
-2.80 0.00256 0.00248 0.00240 0.00233 0.00226 0.00219 0.00212 0.00205 0.00199 0.00193
-2.70 0.00347 0.00336 0.00326 0.00317 0.00307 0.00298 0.00289 0.00280 0.00272 0.00264
-2.60 0.00466 0.00453 0.00440 0.00427 0.00415 0.00402 0.00391 0.00379 0.00368 0.00357
-2.50 0.00621 0.00604 0.00587 0.00570 0.00554 0.00539 0.00523 0.00508 0.00494 0.00480
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 200/358
The probability that a drawing from a unit normal distribution will produce a value less than the constant d is
z2 / 2
d e
~
Pr( z d ) N (d ) dz
2
Range of d : -2.49 d 0.00
d -0.00 -0.01 -0.02 -0-03 -0.04 -0.05 -0.06 -0.07 -0-08 -0.09
-2.40 0.00820 0.00798 0.00776 0.00755 0.00734 0.00714 0.00695 0.00676 0.00657 0.00639
-2.30 0.01072 0.01044 0.01017 0.00990 0.00964 0.00939 0.00914 0.00889 0.00866 0.00842
-2.20 0.01390 0.01355 0.01321 0.01287 0.01255 0.01222 0.01191 0.01160 0.01130 0.01101
-2.10 0.01786 0.01743 0.01700 0.01659 0.01618 0.01578 0.01539 0.01500 0.01463 0.01426
-2.00 0.02275 0.02222 0.02169 0.02118 0.02068 0.02018 0.01970 0.01923 0.01876 0.01831
-1.90 0.02872 0.02807 0.02743 0.02680 0.02619 0.02559 0.02500 0.02442 0.02385 0.02330
-1-80 0.03593 0.03515 0.03438 0.03362 0.03288 0.03216 0.03144 0.03074 0.03005 0.02938
-1.70 0.04457 0.04363 0.04272 0.04182 0.04093 0.04006 0.03920 0.03836 0.03754 0.03673
-1-60 0.05480 0.05370 0.05262 0.05155 0.05050 0.04947 0.04846 0.04746 0.04648 0.04551
-1-50 0.06681 0.06552 0.06426 0.06301 0.06178 0.06057 0.05938 0.05821 0.05705 0.05592
-1.40 0.08076 0.07927 0.07780 0.07636 0.07493 0.07353 0.07215 0.07078 0.06944 0.06811
-1-30 0.09680 0.09510 0.09342 0.09176 0.09012 0.08851 0.08691 0.08534 0.08379 0.08226
-1.20 0.11507 0.11314 0.11123 0.10935 0.10749 0.10565 0.10383 0.10204 0.10027 0.09853
-1.10 0.13567 0.13350 0.13136 0.12924 0.12714 0.12507 0.12302 0.12100 0.11900 0.11702
-1.00 0.15866 0.15625 0.15386 0.15150 0.14917 0.14686 0.14457 0.14231 0.14007 0.13786
-0.90 0.18406 0.18141 0.17879 0.17619 0.17361 0.17106 0.16853 0.16602 0.16354 0.16109
-0.80 0.21186 0.20897 0.20611 0.20327 0.20045 0.19766 0.19489 0.19215 0.18943 0.18673
-0.70 0.24196 0.23885 0.23576 0.23270 0.22965 0.22663 0.22363 0.22065 0.21770 0.21476
-0.60 0.27425 0.27093 0.26763 0.26435 0.26109 0.25785 0.25463 0.25143 0.24825 0.24510
-0.50 0.30854 0.30503 0.30153 0.29806 0.29460 0.29116 0.28774 0.28434 0.28096 0.27760
-0.40 0.34458 0.34090 0.33724 0.33360 0.32997 0.32636 0.32276 0.31918 0.31561 0.31207
-0-30 0.38209 0.37828 0.37448 0.37070 0.36693 0.36317 0.35942 0.35569 0.35197 0.34827
-0.20 0.42074 0.41683 0.41294 0.40905 0.40517 0.40129 0.39743 0.39358 0.38974 0.38591
-0.10 0.46017 0.45620 0.45224 0.44828 0.44433 0.44038 0.43644 0.43251 0.42858 0.42465
0.00 0.50000 0.49601 0.49202 0.48803 0.48405 0.48006 0.47608 0.47210 0.46812 0.46414
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 201/358
The probability that a drawing from a unit normal distribution will produce a value less than the constant d is
z2 / 2
d e
~
Pr( z d ) N (d ) dz 0
2
Range of d : 0.00 d 2.49
d 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0.00 0.50000 0.50399 0.50798 0.51197 0.51595 0.51994 0.52392 0.52790 0.53188 0.53586
0.10 0.53983 0.54380 0.54776 0.55172 0.55567 0.55962 0.56356 0.56749 0.57142 0.57535
0.20 0.57926 0.58317 0.58706 0.59095 0.59483 0.59871 0.60257 0.60642 0.61026 0.61409
0.30 0.61791 0.62172 0.62552 0.62930 0.63307 0.63683 0.64058 0.64431 0.64803 0.65173
0.40 0.65542 0.65910 0.66276 0.66640 0.67003 0.67364 0.67724 0.68082 0.68439 0.68793
0.50 0.69146 0.69497 0.69847 0.70194 0.70540 0.70884 0.71226 0.71566 0.71904 0.72240
0.60 0.72575 0.72907 0.73237 0.73565 0.73891 0.74215 0.74537 0.74857 0.75175 0.75490
0.70 0.75804 0.76115 0.76424 0.76730 0.77035 0.77337 0.77637 0.77935 0.78230 0.78524
0.80 0.78814 0.79103 0.79389 0.79673 0.79955 0.80234 0.80511 0.80785 0.81057 0.81327
0.90 0.81594 0.81859 0.82121 0.82381 0.82639 0.82894 0.83147 0.83398 0.83046 0.83891
1.00 0.84134 0.84375 0.84614 0.84850 0.85083 0.85314 0.85543 0.85769 0.85993 0.86214
1.10 0.86433 0.86650 0.86864 0.87076 0.87286 0.87493 0.87698 0.87900 0.88100 0.88298
1.20 0.88493 0.88686 0.88877 0.89065 0.89251 0.89435 0.89617 0.89796 0.89973 0.90147
1.30 0.90320 0.90490 0.90658 0.90824 0.90988 0.91149 0.91309 0.91466 0.91621 0.91774
1.40 0.91924 0.92073 0.92220 0.92364 0.92507 0.92647 0.92785 0.92922 0.93056 0.93189
1.50 0.93319 0.93448 0.93574 0.93699 0.93822 0.93943 0.94062 0.94179 0.94295 0.94408
1.60 0.94520 0.94630 0.94738 0.94845 0.94950 0.95053 0.95154 0.95254 0.95352 0.95449
1.70 0.95543 0.95637 0.95728 0.95818 0.95907 0.95994 0.96080 0.96164 0.96246 0.96327
1.80 0.96407 0.96485 0.96562 0.96637 0.96712 0.96784 0.96856 0.96926 0.96995 0.97062
1.90 0.97128 0.97193 0.97257 0.97320 0.97381 0.97441 0.97500 0.97558 0.97615 0.97670
2.00 0.97725 0.97778 0.97831 0.97882 0.97932 0.97982 0.98030 0.98077 0.98124 0.98169
2.10 0.98214 0.98257 0.98300 0.98341 0.98382 0.98422 0.98461 0.98500 0.98537 0.98574
2.20 0.98610 0.98645 0.98679 0.98713 0.98745 0.98778 0.98809 0.98840 0.98870 0.98899
2.30 0.98928 0.98956 0.98983 0.99010 0.99036 0.99061 0.99086 0.99111 0.99134 0.99158
2.40 0.99180 0.99202 0.99224 0.99245 0.99266 0.99286 0.99305 0.99324 0.99343 0.99361
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 202/358
The probability that a drawing from a unit normal distribution will produce a value less than the constant d is
z2 / 2
d e
~
Pr( z d ) N (d ) dz 0
2
Range of d : 2.50 d 4.99
d 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
2.50 0.99379 0.99396 0.99413 0.99430 0.99446 0.99461 0.99477 0.99492 0.99506 0.99520
2.60 0.99534 0.99547 0.99560 0.99573 0.99585 0.99598 0.99609 0.99621 0.99632 0.99643
2.70 0.99653 0.99664 0.99674 0.99683 0.99693 0.99702 0.99711 0.99720 0.99728 0.99736
2.80 0.99744 0.99752 0.99760 0.99767 0.99774 0.99781 0.99788 0.99795 0.99801 0.99807
2.90 0.99813 0.99819 0.99825 0.99831 0.99836 0.99841 0.99846 0.99851 0.99856 0.99861
3.00 0.99865 0.99869 0.99874 0.99878 0.99882 0.99886 0.99889 0.99893 0.99897 0.99900
3.10 0.99903 0.99906 0.99910 0.99913 0.99916 0.99918 0.99921 0.99924 0.99926 0.99929
3.20 0.99931 0.99934 0.99936 0.99938 0.99940 0.99942 0.99944 0.99946 0.99948 0.99950
3.30 0.99952 0.99953 0.99955 0.99957 0.99958 0.99960 0.99961 0.99962 0.99964 0.99965
3.40 0.99966 0.99968 0.99969 0.99970 0.99971 0.99972 0.99973 0.99974 0.99975 0.99976
3.50 0.99977 0.99978 0.99978 0.99979 0.99980 0.99981 0.99981 0.99982 0.99983 0.99983
3.60 0.99984 0.99985 0.99985 0.99986 0.99986 0.99987 0.99987 0.99988 0.99988 0.99989
3.70 0.99989 0.99990 0.99990 0.99990 0.99991 0.99991 0.99992 0.99992 0.99992 0.99992
3.80 0.99993 0.99993 0.99993 0.99994 0.99994 0.99994 0.99994 0.99995 0.99995 0.99995
3.90 0.99995 0.99995 0.99996 0.99996 0.99996 0.99996 0.99996 0.99996 0.99997 0.99997
4.00 0.99997 0.99997 0.99997 0.99997 0.99997 0.99997 0.99998 0.99998 0.99998 0.99998
4.10 0.99998 0.99998 0.99998 0.99998 0.99998 0.99998 0.99998 0.99998 0.99999 0.99999
4.20 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999
4.30 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999 0.99999
4.40 0.99999 0.99999 1.00000 1.00000 1,00000 1.00000 1.00000 1.00000 1.00000 1.00000
4.50 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
4.60 1,00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
4.70 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
4.80 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
4.90 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 203/358
The valuation equations for the European call and put options are
and
respectively, where
ln( S / X ) (b 0.5 2 )T
d1 and d 2 d1 T
T
d 2 d1 T (A4.3)
d 2 d 1 2d1 T 2T
2 2
(A4.4)
d 2 d 1 2[ln(S / X ) bT 0.5 2T ] 2T d 12 2[ln(Se bT / X )]
2 2
d 22 d 12 2[ln(Se bT / X ) (A4.5)
d 22 /2
e
n( d 2 ) (A4.6)
2
d 12 / 2[ln( SebT / X ) d 12 [ln( SebT / X )
e e e
n( d 2 ) n(d1 ) Se bT / X (A4.7)
2 2
n(d1 ) n(d 2 ) X / Se bT (A4.8)
c N (d1 ) N (d 2 )
c e (br )T N (d1 ) Se ( br )T Xe rT
S S S
c N ( d ) d N (d 2 ) d 2
c e ( br )T N (d1 ) Se (br )T 1 1
Xe rT
S d1 S d 2 S
c d d
c e ( br )T N (d1 ) Se ( br )T n(d1 ) 1 Xe rT n(d 2 ) 2
S S S
c d d
c e ( br )T N (d1 ) Se ( br )T n(d1 ) 1 Xe rT 1 n(d1 ) Se bT / X
S S S
c
c e ( br )T N (d1 ) 0 (A4.9a)
S
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 204/358
p d d
p e ( br )T N (d1 ) [ Xe rT n(d1 ) Se bT / X ] 1 Se ( br )T n(d1 ) 1
S S S
p
p e ( br )T N (d1 ) 0 (A4.9b)
S
c / c S S
c c e ( br )T N (d1 ) 1 (A4.10a)
S / S c c
p / p S S
p p e ( br )T N (d1 ) 1 (A4.10b)
S / S p p
c e ( br )T N (d1 ) d e ( b r ) T n ( d1 )
c e ( b r ) T n ( d1 ) 1 0 (A4.11a)
S S S S T
e ( br )T N ( d1 ) d e ( b r ) T n ( d1 )
p p e ( b r ) T n ( d1 ) 1 c 0 (A4.11b)
S S S S T
c
e rT N (d 2 ) 0 (A4.12a)
X
p
e rT N (d 2 ) 0 (A4.12b)
X
c N (d1 ) N (d 2 )
TSe (br )T N (d1 ) Se ( br )T Xe rT TSe (br )T N (d1 ) 0 (A4.13a)
b b b
p N ( d 2 ) N (d1 )
Xe rT TSe ( br )T N (d1 ) Se ( br )T TSe ( br )T N (d1 ) 0 (A4.13b)
b b b
c N ( d1) N (d 2 )
TSe ( br )T N (d1 ) TSe ( br )T N (d1 ) Se ( br )T TXe rT N (d 2 ) Xe rT N (d 2 )
r r r
rT
TXe N (d 2 ) 0 (A4.14a)
p N ( d 2 ) N (d1 )
TXe rT N (d 2 ) Xe rT N (d 2 ) TSe ( br )T N ( d1 ) TSe ( br )T N (d1 ) Se (br )T
r r r
rT
TXe N (d 2 ) 0 (A4.14b)
c N (d1 ) N (d 2 ) d d
Vegac Se ( br )T Xe rT Se (br )T n(d1 ) 1 Xe rT n(d 2 ) 2
c d d d d
Vegac Se ( br )T n(d1 ) 1 [ Xe rT n(d1 ) Se bT / X ] 2 Se (br )T n(d1 ) 1 2
( b r )T
Se n(d1 ) T 0 (A4.15a)
p N (d 2 ) N (d1 ) d d
Vega p Xe rT Se ( br )T Xe rT n(d 2 ) 2 Se ( br )T n(d1 ) 1
p d d
Vega p Se ( br )T n(d1 ) 1 2 Se ( br )T n(d1 ) T Vegac 0 (A4.15b)
c N(d1) N(d2 )
c (b r)Se(br )T N(d1) Se(br )T rXerT N(d2 ) XerT
T
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 205/358
c d d
c Se (br )T n(d1 ) 1 2 (b r ) Se ( br )T N (d1 ) rXe rT N (d 2 )
T T T
c
c Se (br )T n(d1 ) (b r ) Se (br )T N (d1 ) rXe rT N (d 2 ) 0 (A4.16a)
T 2 T
p N (d 2 ) N (d1 )
p rXe rT N (d 2 ) Xe rT (b r ) Se ( br )T N (d1 ) Se (br )T
T T T
p d d
p rXe rT N (d 2 ) (b r ) Se (br )T N (d1 ) Se ( br )T n(d1 ) 1 2
T T T
p
p Se ( br )T n(d1 ) (b r ) Se ( br )T N (d1 ) rXe rT N (d 2 ) 0 (A4.16b)
T 2 T
d1 d 2 ln(S / X ) b ln(S / X ) b
T T 2T 3 / 2 2 T 4 T 2T 3 / 2 2 T 4 T 2 T
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 206/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 207/358
12.5 RESUMEN
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 208/358
Premisas
La única premisa nueva en esta sección es que todas las posiciones, incluyendo cualquier posición en el
subyacente, se mantiene hasta el vencimiento de las opciones, T . Como en los capítulos anteriores, el costo
de acarreo del subyacente ocurre a la tasa b . Una posición larga (corta) implica que el tenedor paga (recibe)
el costo de acarreo. También, el costo de acarreo incluye, además del interés, los cargos (recibos) adicionales
por mantener la mercancía. El costo de acarreo de un contrato de opciones es sólo la tasa de interés libre de
riesgo, r . El precio inicial de la mercancía se denota como S y el precio final se denota como ST . Los
precios iniciales de la call y la put se denotan como c y p , respectivamente. Si la opción es comprada, el
precio de compra se financia a la tasa r , y si la opción es vendida, los ingresos de la venta son invertidos a la
tasa r hasta el vencimiento de la opción. A no ser que se especifique de otra manera, se asume que la call y
la put tienen el mismo precio de ejercicio. Una ganancia ocurre cuando una posición gana más que el costo
del interés de los fondos atados en la posición.
El análisis de cada estrategia es en dos pasos. Primero, presentamos la función de ganancia al
vencimiento, T , y luego mostramos la función de ganancia en un diagrama. Luego describimos los precios
finales de punto de equilibrio de la mercancía (cuando la estrategia tiene una ganancia cero al vencimiento),
la pérdida máxima y la ganancia máxima. Los diagramas trazan la posición de ganancia cero como una línea
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 209/358
sólida horizontal y trazan la función de ganancia como otra línea sólida. La intersección de las líneas sólidas
representa la posición de ganancia cero.
Seis Posiciones Básicas: Las funciones de ganancia final de las seis posiciones básicas opción/mercancía
son:
60
40
20
Payoff
0
-20 0 50 100 150 200
-40
-60
Terminal Commodity Pirce
2. Posición Corta en Futuros de Mercancías (o Spot): La mercancía es vendida y los ingresos de la venta
obtienen una tasa b hasta el vencimiento en T .
T SebT ST (12.1b)
Figura 12.1b: Posición Corta en Futuros de Mercancías
Short Commodity Futures
60
40
20
Payoff
0
-20 0 50 100 150 200
-40
-60
Terminal Commodity Futures
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 210/358
50
40
30
Profit
20
10
0
-10 50 70 90 110 130 150
Punto de equilibrio: ST X ce rT .
Pérdida Máxima: cerT si ST X .
Ganancia Máxima: Ilimitada, si ST se eleva sin límite.
4. Call Corta: La call es vendida, y los ingresos de la venta ganan r hasta el vencimiento en T .
ST X cerT si ST X
T rT . (12.2b)
ce si ST X
10
0
-10 50 70 90 110 130 150
Profit
-20
-30
-40
-50
Terminal Commodity Price
Punto de equilibrio: ST X ce rT .
Ganancia Máxima: cerT si ST X .
Pérdida Máxima: Ilimitada, si ST se eleva sin límite.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 211/358
50
40
Profit 30
20
10
0
-10 50 70 90 110 130 150
6. Put Corta: La put es vendida, y los ingresos de la venta ganan r hasta el vencimiento de la opción en T .
perT si ST X
T (12.3b)
X ST pe si ST X
rT
10
0
-10 50 70 90 110 130 150
Profit
-20
-30
-40
-50
Terminal Commodity Price
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 212/358
En las posiciones de las opciones descritas arriba y más adelante en el capítulo, la prima pagada al
comprador o recibida por el vendedor depende del precio de la mercancía en el momento que el contrato fue
escrito. Al trazar los diagramas de ganancia, hacemos supuestos razonables sobre las primas iniciales de la
put o la call, pero la forma de los diagramas de ganancia no se ve afectada por ninguna prima inicial pagada.
Para posiciones que involucran un solo precio de ejercicio, nosotros generalmente asumimos que estamos at-
the-money cuando comenzamos a negociar. Para posiciones que involucran precios de ejercicio diferentes,
generalmente asumimos que una opción está at-the-money y la otra no.
Conversion Profit
15
10
5
Porfit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
La función de ganancia (12.4) muestra que la ganancia de la cartera al vencimiento de la opción está
asegurada porque el precio final de la mercancía, ST , no aparece en la expresión. Esto también se puede ver
en la Figura 12.4, si la ganancia de la cartera es descrita por la línea horizontal superior. Dado que esta línea
nunca cruza la línea de ganancia cero debajo de ella, la ganancia está asegurada.
El significado económico de este resultado de ganancia segura es importante. Recuerden que todas
las posiciones en la mercancía/opción son completamente autofinanciadas. Por lo tanto, se puede asumir que
si las opciones y la mercancía tienen un precio apropiado en el mercado, la ganancia final de la estrategia es
cero. Y, si la ganancia final es cero, entonces la relación de paridad put-call
(c p) Se (br )T Xe rT (10.22)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 213/358
que fue derivada en el Capítulo 10, se mantiene. Cuando la paridad put-call se mantiene, las dos líneas
sólidas en la Figura 12.4 coinciden.
El arbitraje de conversión entra en juego cuando hay errores temporales en los precios. Por ejemplo,
un negociador institucional podría comprar un gran número de calls del índice en reacción a información
nueva sobre el mercado de valores. Como resultado del exceso de la presión de compra sobre los call, el
precio podría incrementar en más de lo que está garantizado con relación a los precios de la put y el índice de
valores subyacente. Conforme el elaborador del mercado vende los calls al negociador institucional,
simultáneamente compra los puts y la cartera del índice para deshacerse del riesgo del precio de la mercancía
de la posición call corta, capturando así la violación temporal de la paridad put-call. Esta actividad de
negociación asegura una cierta ganancia positiva.
Posiciones Sintéticas
Las conversiones explotan las oportunidades de arbitraje al crear una posición sintética en cualquier activo –
a partir de posiciones juiciosamente seleccionadas en otros dos activos. En una conversión, por ejemplo, una
posición larga en una mercancía, combinada con una posición put larga, es equivalente a una posición call
larga. Cuando esta posición call larga se combina con una posición call corta, se tiene como resultado una
cartera libre de riesgo. Para demostrar esta idea con más detalle, ahora mostramos cómo se pueden crear
posiciones sintéticas largas y cortas en el subyacente utilizando calls y put.
15
10
5
Profit
0
-5 80 90 100 110 120
-10
-15
Terminal Commodity Price
La función de ganancia de la posición call larga/put corta es idéntica a la posición larga de la mercancía en la
Figura 12.1a. La única diferencia entre las funciones de ganancia (12.1a) y (12.5a) es que (12.1a) tiene el
término SebT , mientras que(12.5a) tiene el término X (c p)e rT . Pero esto es esperado. Si la paridad put-
call (10.22) se mantiene, estos dos valores son iguales. Si por cualquier razón, SebT X (c p)e rT , es más
barato crear sintéticamente una posición larga en la mercancía que comprarla.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 214/358
15
10
5
Profit 0
-5 80 90 100 110 120
-10
-15
Terminal Commodity Price
La Figura 12.6 muestra los diagramas de ganancia para la compra de una, dos y tres calls. Noten que el punto
de equilibrio es independiente del número de calls comprados. El cambio importante es que se pierde más de
la prima de la opción si la opción vence out-of-the-money, y la tasa de ganancia por dólar del precio de la
mercancía se incrementa si la opción vence in-the-money. Este concepto es utilizado más adelante en el
capítulo cuando se discuten varias estrategias de margen de ratios y estrategias para escribir ratios.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 215/358
130
80
Profit 30
Márgenes de Volatilidad. Los márgenes de volatilidad involucran la compra de una put y una call o la venta
de una put y una call.
15. Un “straddle” largo o un margen de volatilidad largo: Comprar call y comprar put.
S X (c p)erT si ST X
T T (12.7)
X ST (c p)e si ST X
rT
Como lo demuestran la función de ganancia (12.7) y la Figura 12.7, la estrategia produce ganancias positivas
cuando el precio del subyacente sube o baja en una cantidad suficiente. Por esta razón, con frecuencia el
comprar un “straddle” se le llama comprar volatilidad. La estrategia pierde dinero cuando el precio final de
la mercancía, ST , está dentro de la banda X (c p)e rT al vencimiento de las opciones.
16. Straddle corto o margen de volatilidad corto: Vender call y vender put.
Esta estrategia es inversa al straddle largo. Los inversionistas realizan straddles cortos o venden volatilidad
cuando piensan que el precio de la mercancía no se moverá mucho antes del vencimiento de las opciones.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 216/358
25
15
Profit
5
-5 70 90 110 130
-15
Terminal Commodity Price
17. Strangle largo: Comprar call y comprar put, con el precio de ejercicio de la put, X p , menor que el precio
del ejercicio de la call X c , (es decir, X p X c )
ST X c (c p )e rT si ST X c
T (c p)erT si X p ST X c . (12.8)
X p ST (c p )e si ST X p
rT
10
5
Profit
0
70 90 110 130
-5
-10
Terminal Commodity Price
Un strangle y un straddle largo tienen el mismo objetivo. La diferencia es que el strangle requiere
menor inversión porque la call o la put están más out-of-the-money que con el straddle. Sin embargo, con un
costo de inversión menor viene una región más amplia en la cual la estrategia no es rentable. El precio de la
mercancía debe moverse más que con el straddle para que el strangle sea rentable al vencimiento. Una
variación en el strangle largo es que el precio de ejercicio de la call está por debajo del precio de ejercicio de
la put. Como tanto la put como la call están in-the-money, el costo y la rentabilidad son mayores.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 217/358
Combinación de Opciones con Precios de Ejercicio Diferentes. Las calls o puts con precios de ejercicio
diferentes pueden combinarse para crea márgenes bull o bear, márgenes de ratio o ratio inverso y márgenes
“butterfly” largos o cortos.
19. Margen Bull con Calls. Comprar una call con un precio de ejercicio menor, X l , y vender una call
idéntica excepto con un precio de ejercicio mayor, X h (es decir, X l X h ).
X h X l (cl ch )erT si ST X h
T ST X l (cl ch )erT si X l ST X h . (12.9a)
(cl ch )e si ST X l
rT
20
15
10
Profit
5
0
-5 70 90 110
-10
Terminal Commodity Price
El término “bull” se debe a que esta estrategia obtiene ganancia cuando se incrementa el precio del
subyacente. La estrategia es bastante conservadora en el sentido que si los inversionistas están confiados en
que se va a incrementar el precio de la mercancía, una posición call larga sería más rentable. El beneficio de
comprar el margen bull es que el vender la call out-of-the-money proporciona un ingreso que compensa el
costo de comprar otra call. El costo es la pérdida de la ganancia potencial si es que el precio de la mercancía
se eleva drásticamente.
Antes del vencimiento, un margen bull toma ventaja del hecho que el delta el mayor para la call in-
the-money que para la call out-of-the-money. Conforme se incrementa el precio de la mercancía, la ganancia
sobre la Call Larga in-the-money sobrepasa la pérdida sobre la call corto out-of-the-money. Conforme se
eleva más el precio de la mercancía y ambas calls se vuelven más in-the-money o conforme se acerca el
vencimiento, la ganancia y la pérdida sobre las dos posiciones se compensan completamente.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 218/358
20. Margen Bear con Calls: Vender una call con un precio de ejercicio menor, X l , y comprar una call
idéntica excepto con un precio de ejercicio mayor, X h (es decir, X l X h ).
X l X h (cl ch )e rT si ST X h
T X l ST (cl ch )erT si X l ST X h . (12.9b)
(cl ch )e si ST X l
rT
20
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
-20
Terminal Commodity Price
21. Margen Bull con Puts: Comprar una put con un precio de ejercicio menor, X l , y vender una put idéntica
excepto con un precio de ejercicio mayor, X h (es decir, X l X h ).
( ph pl )e rT si ST X h
T ST X h ( ph pl )erT si X l ST X h (12.10a)
X l X h ( ph pl )e si ST X l
rT
Punto de equilibrio: ST X h ( ph pl )e rT
Pérdida máxima: X l X h ( ph pl )e rT , si ST X l
Ganancia máxima: ( ph pl )e rT , si ST X h
El margen “bull” con puts es equivalente al margen bull utilizando calls, aunque las fuentes de los resultados
son diferentes. Para el margen bull utilizando put, la ganancia, si se incrementa el precio de la mercancía,
viene de la prima sobre la put vendida a un precio de ejercicio mayor. Para el margen bull utilizando call, la
ganancia, si se incrementa el precio de la mercancía, viene del hecho que la ganancia sobre la Call Larga a
un precio de ejercicio menor excede la pérdida sobre la call corto a un precio de ejercicio mayor.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 219/358
20
15
10
5
Profit
0
-5 50 70 90 110 130 150
-10
-15
-20
Terminal Commodity Price
22. Margen “Bear” con Puts: Vender la put a un precio de ejercicio menor, X l , y comprar una put idéntica
excepto con un precio de ejercicio mayor, X h , (es decir, X l X h ).
( ph pl )e rT si ST X h
T X h ST ( ph pl )erT si X l ST X h (12.10b)
X h X l ( ph pl )e si ST X l
rT
20
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
-20
Terminal Commodity Price
Punto de equilibrio: ST X h ( ph pl )e rT
Pérdida máxima: X l X h ( ph pl )e rT , si ST X l
Ganancia máxima: ( ph pl )e rT , si ST X h
23. Margen de Ratio con Calls: Vender la call a un precio de ejercicio menor, X l , y vender n calls idénticas
excepto con un precio de ejercicio mayor, X h , (es decir, X l X h ).
nX h X l (n 1) ST (nch cl )e rT si ST X h
T ST X l (nch cl )erT si X l ST X h (12.11a)
(nch cl )e si ST X l
rT
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 220/358
30
20
10
Profit
0
70 90 110 130
-10
-20
Terminal Commodity Price
nX h X l ( nc h cl )e rT
Puntos de equilibrio: (a) S T . (b) ST X l (nch cl )e rT si nch cl .
( n 1)
Pérdida máxima: (a) ilimitada, si ST se eleva sin límite. (b) (nch cl )e rT , si nch cl y ST X l .
Ganancia máxima: X h X l (nch cl )e rT , si ST X h .
Un margen de ratio call es como un margen bull utilizando calls excepto que varias opciones idénticas son
vendidas al precio de ejercicio mayor.
El “ratio” del margen de ratio se define en términos de la cantidad de call vendidos, n , por call
comprado, esto es, un margen de ratio n : 1 . El margen de ratio descrito en la Figura 12.1 es un margen de
ratio 2:1, el cual involucra la venta de dos calls con un precio de ejercicio mayor contra la compra de una call
con un precio de ejercicio menor.
El margen de ratio es más rentable cuando el precio de la mercancía no se mueve mucho. Si el precio
de la mercancía cae, el riesgo de bajada es fijo. Si el precio de la mercancía cae por debajo del precio de
ejercicio, ambas opciones vencen sin valer nada. La inversión inicial puede ser un débito neto o un crédito
neto, dependiendo de las magnitudes relativas sobre las primas y el ratio del margen. La Figura 12.11a indica
que el margen de ratio 2:1 mostrado tiene un crédito neto inicial. Si el precio de la mercancía sube sin límite,
la estrategia pierde dinero sin límite.
24. Margen de ratio inverso con Calls: Vender la call con un precio de ejercicio menor, X l , comprar n calls
idénticas excepto que tienen un precio de ejercicio mayor, X h (es decir, X l X h ).
(n 1) ST - nX h X l (nch cl )erT si ST X h
T X l ST (nch cl )erT si X l ST X h . (12.11b)
(nch cl )e si ST X l
rT
nX h X l (nch cl )e rT
Puntos de equilibrio: (a) ST (b) ST X l (nch cl )e rT , si nch cl
(n 1)
Ganancia máxima: (a) ilimitada, si ST se eleva sin límite (b) (nch cl )e rT , si nch cl y ST X l
Pérdida máxima: X h X l (nch cl )e rT , si ST X h
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 221/358
40
30
20
10
Profit
0
-10 50 70 90 110 130 150
-20
-30
-40
Terminal Commodity Value
25. Margen de Ratio con Puts: Comprar la put con un precio de ejercicio mayor, X h , vender n puts idénticas
excepto con un precio de ejercicio menor, X l (es decir, X l X h ).
(npl ph )erT si ST X h
T X h ST (npl ph )erT si X l ST X h . (12.12a)
X h nX l (n 1) ST (npl ph )e si ST X l
rT
20
15
10
5
Profit
0
-5 60 80 100 120
-10
-15
-20
Terminal Commodity Price
nX l X h (npl ph )e rT
Puntos de equilibrio: (a) ST (b) ST X h (npl ph )e rT , si npl ph
(n 1)
Pérdida máxima: (a) X h nX l (npl ph )e rT , si ST cae hasta cero (b) (npl ph )e rT , si npl ph y ST X h .
Ganancia máxima: X h X l (npl ph )e rT , si ST X l
26. Margen de Ratio Inverso con Puts: Vender la put con un precio de ejercicio mayor, X h , vender n puts
idénticos excepto con un precio de ejercicio menor, X l (es decir, X l X h ).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 222/358
(npl ph )erT si ST X h
n T ST X h (npl ph )e rT si X l ST X h . (12.12b)
nX l X h (n 1) ST (npl ph )e si ST X l
rT
20
15
10
5
Profits
0
-5 60 80 100 120
-10
-15
-20
Terminal Commodity Price
nX l X h (npl ph )e rT
Puntos de equilibrio: (a) ST (b) ST X h (npl ph )e rT , si npl ph
(n 1)
Pérdida máxima: (a) X h nX l (npl ph )e rT , si ST cae hasta cero (b) (npl ph )e rT si npl ph , y ST X h .
Ganancia máxima: X h X l (npl ph )e rT si ST X l
27. Margen “Butterfly” Largo con Calls: Vender la call con un precio de ejercicio menor, X l , comprar dos
calls con el precio de ejercicio medio, X m , y vender la call con un precio de ejercicio mayor, X h .
X l 2 X m X h (cl 2cm ch )e rT si ST X h
S (2 X m X l ) (cl 2cm ch )e si X m ST X h
rT
T T (12.13)
X l ST (cl 2cm ch )e si X l ST X m
rT
(c 2c c )e rT si ST X l
l m h
Un margen butterfly largo combina un margen bear y un margen bull. Las ganancias son similares a las de
un margen bull si el precio de la mercancía se incrementa y a aquellas de un margen bear si el precio de la
mercancía cae. El inversionista pierde dinero cuando el precio de la mercancía se mantiene neutral. El
diagrama de ganancia resultante (Figura 12.13) se asemeja a una mariposa (butterfly) –de allí el nombre.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 223/358
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
28. Margen “Butterfly” Corto con Calls: Comprar la call con un precio de ejercicio menor, X l , vender dos
calls con el precio de ejercicio medio, X m , y vender la call con un precio de ejercicio mayor, X h .
29. Margen “Butterfly” Largo con Puts: Vender la put con un precio de ejercicio menor, X l , comprar dos
puts con el precio de ejercicio medio, X m , y vender la put con un precio de ejercicio mayor, X h .
30. Margen “Butterfly” Corto con Puts: Comprar la put con un precio de ejercicio menor, X l , vender dos
puts con el precio de ejercicio medio, X m , y comprar la put con un precio de ejercicio mayor, X h .
Márgenes calendario: Un margen calendario requiere la compra de una call o put con un vencimiento y la
venta de una opción idéntica con un vencimiento diferente. Presentar una función de ganancia para un
margen calendario es problemático porque se requiere una ecuación de precio para mostrar el valor de una
opción distante en la fecha de vencimiento más cercana.104 Por estar razón, vamos inmediatamente al
diagrama de ganancia.
31. Margen Calendario Largo: Comprar una call con un vencimiento distante, y vender una call idéntica con
un vencimiento cercano.
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
104
Las ecuaciones de la call y put europea (11.25) y (11.28) se utilizan para poner precio a la opción distante en la fecha de
vencimiento más cercana.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 224/358
Las opciones at-the-money son utilizadas para generar el margen calendario de la Figura 12.14, y los
resultados son trazados al vencimiento de una call cercana. Este margen es neutral dado que se obtienen
ganancias positivas siempre y cuando el precio de la mercancía no se mueva mucho debido al vencimiento
de la opción cercana.105 Dado que la opción de plazo más largo tiene un precio mayor, esta estrategia tiene
una posición de débito neto (es decir, pagamos la diferencia entre los precios de la opción cuando se forma la
posición). Sin embargo, la pérdida máxima se limita a la cantidad de débito neto. La ganancia máxima ocurre
cuando el precio de la mercancía es igual al precio de ejercicio de la opción cercana, pero la cantidad no está
clara dado que depende de la vida restante de la opción distante y la volatilidad del retorno de la mercancía.
La amplitud del rango de ganancia y los puntos de equilibrio también son funciones de la volatilidad y el
tiempo hasta el vencimiento.
Manteniendo los otros factores constantes, la rentabilidad del margen calendario se debe a la
disminución de tiempo de las primas de las opciones. Como demostraremos más adelante en el capítulo, la
tasa de disminución de tiempo (es decir, el theta de la opción) es mayor mientras menor sea el tiempo hasta
el vencimiento. En un margen calendario largo, se establece una posición corta en una opción cercana para
poder capturar su disminución de tiempo a costa de la disminución de tiempo en la opción distante.
32. Margen Calendario Corto con Calls: Vender la call con un vencimiento distante, y comprar una call
idéntica con un vencimiento cercano.
33. Margen Calendario Largo con Puts: Comprar la put con un vencimiento distante, y comprar una put
idéntica con un vencimiento cercano.
34. Margen Calendario Corto con Puts: Vender la put con un vencimiento distante, y comprar una put
idéntica con un vencimiento cercano.
35. Margen Calendario de Ratio Largo con Calls: Comprar una call con un vencimiento distante, y vender
más de una call idéntica con un vencimiento cercano.
30
20
10
Profit
0
-10 50 70 90 110 130 150
-20
-30
Terminal Commodity Price
Al escribir más de una call con vencimiento cercano, el que realiza un margen calendario usualmente recibe
un crédito inicial (él recibe dinero cuando se forma una posición). El crédito neto incrementa la ganancia si
el precio de la mercancía cae por debajo del precio de ejercicio antes del vencimiento de la opción cercana.
Los incrementos en el precio de la mercancía más allá del precio de ejercicio de la opción cercana reducen la
cantidad de la ganancia dado que, cuando ambas opciones están muy in-the-money, el que realiza el margen
está sintéticamente corto en la mercancía. En general, la posición tiende a la baja. La Figura 12.15 muestra el
diagrama de ganancia de un margen calendario con ratio 2:1.
105
Si se utilizan calls out-of-the-money para formar el margen calendario, la posición tiende ligeramente al alza.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 225/358
36. Margen Calendario de Ratio Corto con Calls: Vender una call con un vencimiento distante, y comprar
más de una call idéntica con un vencimiento cercano.
37. Margen Calendario de Ratio Largo con Puts: Comprar una put con un vencimiento distante, y vender más
de una put idéntica con un vencimiento cercano.
38. Margen Calendario de Ratio Corto con Puts: Vender una put con un vencimiento distante, y comprar más
de una put idéntica con un vencimiento cercano.
Márgenes Diagonales. En general, los márgenes diagonales son cualquier posición de margen conformada
con precios de ejercicio diferentes y vencimientos diferentes. Una margen diagonal largo requiere que la
opción distante sea comprada y la opción cercana sea vendida (comprada). Si el ratio de margen es 1:1, se
tiene como resultado un margen bull (bear) diagonal, dependiendo de si la opción distante tiene un precio de
ejercicio menor (mayor). Los márgenes diagonales largos y cortos que utilizan otros ratios producen una
amplia variedad de posiciones bull y bear. A continuación se describe un margen diagonal posible.
39. Margen bull diagonal: Comprar una call con un precio de ejercicio menor y vencimiento distante, y
vender una call idéntica con un precio de ejercicio mayor y un vencimiento cercano.
Como lo muestra la Figura 12.6, un margen bull diagonal es muy similar al margen bull descrito
anteriormente en esta sección. La pérdida máxima se limita a la diferencia entre los precios de opciones
distantes y cercanas (es decir, la cantidad de débito neto). La ganancia máxima se da cuando el precio de la
mercancía es igual al precio de ejercicio de la opción cercana al vencimiento de la opción cercana. Más allá
de ese nivel, los incrementos en el precio de la mercancía reducen el nivel de ganancia a la diferencia entre
los precios de ejercicio y la cantidad de débito neto.
15
10
5
Profit
0
-5 75 95 115 135
-10
-15
Terminal Commoidty Price
40. Escribir call cubierta: Vender una call contra una posición larga en el subyacente.
X Se ce si ST X
bT rT
T (12.14)
ST Se ce si ST X
bT rT
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 226/358
15
10
5
Profit
0
-5 80 90 100 110 120
-10
-15
Terminal Commodity Price
La Figura 12.17 muestra que el que escribe una call cubierta recibe la prima de la opción a cambio de un
mayor potencial de una posición larga en la mercancía. La posición es equivalente a vender una put sin
subyacente. Esta estrategia sólo tiene sentido si un inversionista piensa que el precio de la mercancía no se
moverá mucho durante la vida de la opción. No se beneficia si el precio de la mercancía se eleva, y la prima
de la opción es un consuelo pequeño si el precio de la mercancía cae drásticamente.
Los grandes fondos de acciones con frecuencia se involucran en una forma especial en escribir calls
cubiertas llamada “overwriting” de opciones. En el caso usual, el fondo tiene administradores de cartera de
acciones y opciones separados. El gerente de la cartera de acciones maneja la inversión en acciones y
aconseja al que realiza el overwrite de la opción sobre la composición actual de la cartera de acciones. El que
realiza el overwrite de la opción luego escribe calls contra las acciones. En el caso que una call se ejerza
contra el que realiza el overwrite de la opción, el que realiza el overwrite de la opción debe comprar las
acciones para entregarlas sobre la opción, ya que él no tiene autoridad para entregar una posición de acciones
existentes. Sin embargo, el propietario del fondo, debe esperar que se tengan que liquidar algunas de sus
acciones, ya que el escribir opciones contra las acciones es una estrategia para reducir el riesgo.
41. Combinación escribir call cubierta: Vender calls in-the-money a la mitad de la posición de la mercancía,
y vender calls out-of-the-money contra la otra mitad.
30
20
10
Profit
0
-10 70 90 110 130
-20
-30
Terminal Commodity Price
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 227/358
Esta estrategia es, generalmente, la misma que la estrategia de escribir call cubierta. La Figura 12.18 muestra
que la ganancia es sólo ligeramente diferente. A lo largo del rango de precio de la mercancía entre los
precios de ejercicio, el que escribe la opción comparte la mitad de cualquier ganancia realizada en el precio
de la acción. Sin embargo, como en la estrategia de la call cubierta anterior, el mayor potencial de una
posición larga en la mercancía se niega completamente una vez que el precio de la mercancía excede cierto
nivel, en este caso el precio de ejercicio de la opción out-of-the-money.
42. Escribir ratio de call: Vender más de una call contra una posición larga en el subyacente.
nX (n 1) ST nce rT si ST X
T . (12.15)
ST Se nce si ST X
bT rT
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
Como los márgenes de ratio, los ratios escritos se expresan en términos del número opciones c vendidas, n ,
por unidad del subyacente. Por lo tanto, un ratio de 2:1 se refiere a escribir dos calls contra una unidad de la
mercancía. En un ratio 2:1, la mitad de los calls están cubiertos mientras que la otra mitad no. Un ratio de 2:1
como el que se muestra en la Figura 12.19 crea un diagrama de retribución que se ve exactamente como si
hubiésemos escrito un straddle. La ganancia máxima ocurre cuando el precio de la mercancía es igual al
precio de ejercicio al vencimiento de la opción. Sin embargo, grandes cambios en cualquier dirección en el
precio de la mercancía, producen pérdidas.
Usualmente los que piensan que el precio de la mercancía no va a moverse durante la vida de la
opción escriben un ratio cuando se trata de lograr un ingreso adicional. Los calls son escritos al precio de
ejercicio más cercano al precio actual de la mercancía. Las ganancia se obtienen si el precio de la mercancía
se mantiene relativamente sin cambio. Sin embargo, las pérdidas pueden ser significativas si el precio del
subyacente cambia significativamente.
43. Escribir ratio variable: Vender calls in-the-money y calls out-of-the-money contra una posición larga en
una mercancía, de manera que la posición larga en la mercancía sea menos que suficiente para cubrir la
entrega si se ejercen las opciones.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 228/358
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
La estrategia de escribir un rateo variable 2:1 se muestra en la Figura 12.20. Como lo muestra la ilustración,
el escribir un ratio variable puede producir un diagrama de ganancia que se ve exactamente como una
posición strangle corta. La ganancia máxima se obtiene cuando el precio de la mercancía cae entre los dos
precios de ejercicio al vencimiento de las opciones. Los grandes movimientos de precios en cualquier
dirección producen pérdidas.
44. Venta corta protegida: Comprar una call contra una posición corta en el subyacente.
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
Ocasionalmente, un inversionista está corto en la mercancía y quiere asegurarse a sí mismo contra grandes
incrementos en el precio del subyacente. Comprar una call proporciona esta seguridad. Como lo muestra la
Figura 12.21, comprar una call contra una posición corta en la mercancía produce una estructura de ganancia
de cartera que se ve exactamente como una posición put larga. La posición también es lo opuesto a una call
cubierta. La ganancia máxima es igual a SebT ce rT si el precio de la mercancía cae hasta cero. La pérdida
máxima es SebT X cerT que debe ser aproximadamente igual al valor de una put con un precio de
ejercicio X y un tiempo hasta el vencimiento T .
45. Cobertura inversa o straddle simulado: Comprar más de una call contra una posición corta en el
subyacente.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 229/358
Reverse Hedge
20
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
-20
Terminal Commodity Price
Comprar dos calls contra una posición corta en el subyacente crea un diagrama de ganancia que se ve
exactamente como un straddle largo. Por esta razón, esta estrategia algunas veces se conoce como un
straddle simulado. La posición es también opuesta a la posición de escribir una ratio de call descrita
anteriormente. La pérdida máxima se da cuando el precio de la mercancía es igual al precio de ejercicio de
las opciones al vencimiento. La ganancia en el lado de movimientos ascendentes es ilimitada, si el precio de
la mercancía se eleva sin límite. Los movimientos descendentes en el precio de la mercancía también son
beneficiosos, dado que las opciones vencen sin valor y el inversionista tiene una posición corta en la
mercancía.
46. Posición protegida en la mercancía: Comprar una put contra una posición larga en la mercancía.
ST SebT pe rT si ST X
T . (12.16)
X Se pe si ST X
bT rT
20
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
-20
Terminal Commodity Price
Comprar put protectoras es una forma muy común de asegurar la mercancía. Como lo indica la Figura 12.23,
un inversionista con una posición larga en la mercancía está bien protegido en el caso que el precio de la
mercancía caiga drásticamente. El costo de este seguro es la prima de la opción put. La posición resultante es
la misma que comprar una call.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 230/358
47. Escribir una call cubierta protegida: Comprar una put contra una call cubierta.
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
Terminal Commodity Price
Escribir una call cubierta significa que el inversionista mantiene una posición larga en la mercancía/corta en
la call. En el caso que haya pasado algún tiempo desde que se formó la call cubierta y el precio de la
mercancía no se haya movido, el inversionista puede querer asegurar su ganancia proveniente de la
disminución de tiempo de la call comprando una put. Cuando lo hace, en efecto, crea un arbitraje de
conversión. Luego, independientemente de la dirección en que se mueve el precio de la mercancía, la
ganancia de la cartera no cambia. La Figura 12.24 lo demuestra claramente.
X SebT (c p)erT si ST X
T . (12.17)
2 ST Se X (c p)e si ST X
bT rT
20
15
10
5
Profit
0
-5 70 90 110 130
-10
-15
-20
Terminal Commodity Price
SebT X (c p)e rT
Punto de equilibrio: ST .
2
Pérdida Máxima: SebT X (c p)e rT , si ST cae hasta cero.
Ganancia Máxima: X SebT (c p)e rT si ST X .
En este caso, el inversionista ha escrito una call y una put contra una posición en el subyacente. Él ha
cobrado dos primas, lo que es igual a la cantidad de la ganancia de cartera si el precio de la mercancía está
por encima del precio de ejercicio al vencimiento de la opción. Sin embargo, si el precio de la mercancía cae,
la ganancia de la cartera cae en el doble de la cantidad, dado que el inversionista no pierde solamente sobre
la posición larga en la mercancía sino también en la posición corta de la put.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 231/358
49. Comprar call contra la mercancía: Comprar una call contra una posición larga en la mercancía.
20
10
Profit
0
70 90 110 130
-10
-20
Terminal Commodity Price
50. Vender put contra la mercancía: Vender una put contra una posición larga en la mercancía.
20
10
Profit
0
70 90 110 130
-10
-20
Terminal Commodity Price
El comprar una call contra una posición larga y vender una put contra una posición larga en la mercancía
sirve como apalancamiento para la tasa de las ganancias de la cartera. La posición larga en la mercancía/call
larga de la Figura 12.26a, por ejemplo, muestra que por debajo de cierto nivel en el precio de la mercancía, la
ganancia de la cartera es menor que la posición larga en la mercancía, dado que se tiene que comprar la call.
Sin embargo, por encima de cierto nivel en el precio de la mercancía, la ganancia de la cartera se incrementa
al doble de la tasa que el precio de la mercancía por sí misma. Por lo tanto, hemos incrementado el
apalancamiento de la estrategia.
El escribir una put contra una posición larga en la mercancía tiene un efecto similar. Los ingresos de
la venta de la opción aumentan la ganancia de la cartera en el lado superior. En el lado inferior, si el precio
de la mercancía cae, el inversionista pierde tanto sobre la posición larga en la mercancía como en la posición
corta del put.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 232/358
presentados en el Capítulo 11. Dado que el concepto de ganancia esperada se basa en el cálculo de
probabilidades, primero veremos el cálculo de probabilidades.106
EJEMPLO 12.1
Asuman que el precio actual de la mercancía es $50 y que los precios de las opciones a tres meses at-the-
money son $3.35 para la call y $2.90 para la put. Calculen la probabilidad que una posición straddle larga
utilizando estas opciones sea rentable al final de los tres meses. Asuman que la tasa de costo de acarreo para
el subyacente es 4 porciento, la tasa de volatilidad d el subyacente es 32 porciento y la tasa de interés libre de
riesgo es 6 porciento.
El segundo paso es transformar los puntos de equilibrio del precio de la mercancía a puntos de equilibrio en
términos de la distribución normal unitaria, esto es,
ln(50 / 43.656) (0.04 0.50.32 2 )0.25
da 0.8305 y
0.32 0.25
ln(50 / 56.344) (0.04 0.50.32 2 )0.25
db 0.7641 .
0.32 0.25
106
En esta sección asumimos que estamos en un mundo neutral al riesgo en el que los precios futuros de las mercancías son
descontados o capitalizados a una tasa libre de riesgo.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 233/358
Reordenando, las expresiones para el máximo y el mínimo del rango del precio de la mercancía son
2
S min Se (b0.5 )T 4 T (12.1 8a)
y
2
S max Se ( b0.5 )T 4 T (12.18b)
Una segunda consideración tiene que ver con el cálculo de la ganancia para una probabilidad dada. Incluso
con un rango pre-especificado del precio final de la mercancía, ST , existe un número infinito de precios de la
mercancía y, por lo tanto, un número infinito de ganancias y probabilidades de cartera. El cálculo es práctico
si aproximamos la distribución continua de los precios finales de la mercancía con una distribución discreta.
Para hacerlo, dividimos la distribución del precio final de la mercancía en n incrementos iguales de Sinc ,
donde
S S min
Sinc max (12.19)
n 1
Luego comenzamos por el precio más bajo de la mercancía y asumimos que, durante el primer
intervalo de S min 0.5Sinc a Smin 0.5Sinc , el precio de la mercancía es S min . Más generalmente, el precio de
la mercancía se asume como Si durante el intervalo i ésimo , el cual tiene un rango de Si ,T 0.5Sinc ,
donde
Si ,T S min (i 1) Sinc (12.20)
La probabilidad que el precio final de la mercancía caiga en este rango
~
Pr[ Si ,T 0.5Sinc ST Si ,T 0.5Sinc ] N (d l ,i ) N (d u ,i ) , (12.21)
EJEMPLO 12.2
Calcular el precio esperado de la mercancía en tres meses, asumiendo que el precio actual de la mercancía es
$50, la tasa de costo de acarreo es 4 porciento y la tasa de volatilidad es 32 porciento.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 234/358
El siguiente paso es dividir el rango de los precios de la mercancía en intervalos de igual espacio. Eligiendo
n 11 , el tamaño de cada intervalo es
94.5589 26.2909
Sinc 6.8268
11 1
El punto medio en cada intervalo se asume como Si ,T 26.2909 6.8268(i 1) , cuyos valores son reportados
en la segunda columna de la Tabla 12.1.
Los puntos finales de cada intervalo entonces son definidos como Sl ,i Si 0.5Sinc y Su ,i Si 0.5Sinc para
i 1,, n . Basados en los valores de punto final del intervalo, los límites de la integral normal unitaria son
calculados y reportados como la tercera y cuarta columna de la Tabla 12.1. Basados en estos límites, la
probabilidad que el precio final de la mercancía caiga en el intervalo i ésimo se calcula utilizando (12.21) y
se reportan en la quinta columna.
La última columna contiene el producto del precio final de la mercancía y su probabilidad respectiva.
Sumando los valores reportados en la última columna, encontramos que el precio final de la mercancía es
~
E ( ST ) i 1[ N (d l ,i ) N (d u ,i )]Si ,T $50.5013
11
Noten que este valor corresponde muy estrechamente con el verdadero precio final de la mercancía, por el
Capítulo 11 sabemos que S $50e 0.04 ( 0.25) 50.5025
La pequeña discrepancia se debe a que el método numérico para calcular el precio final esperado de
la mercancía es sólo una aproximación, aunque una muy exacta en esta ilustración. Se puede obtener una
mayor exactitud estableciendo n en un valor mayor o expandiendo el rango posible de los precios finales de
la mercancía considerados.
Extendiendo este enfoque para calcular la ganancia final de una cartera de opciones es un método
directo: simplemente reemplazando el precio final de la mercancía Si ,T en (12.22) con una ganancia de
cartera de opciones, dado un precio de la mercancía Si ,T , esto es,
E (~T ) i1[ N (d l ,i ) N (d u ,i )] ( Si ,T ) $50.5013 (12.23)
11
En la última sección se presentaron las funciones de ganancia () para una amplia variedad de estrategias.
EJEMPLO 12.3
Calcular la ganancia final esperada de una call at-the-money, asumiendo que el precio actual de la mercancía
es $50, la tasa de costo de acarreo es 4 porciento y la tasa de volatilidad es 32 porciento. La tasa libre de
riesgo se asume como 6 porciento y el precio actual de la call es $3.410.
Todos los pasos en este ejemplo son los mismos que los del Ejemplo 12.2, excepto que en lugar de
multiplicar la probabilidad por el precio final de la mercancía en el intervalo, multiplicamos la probabilidad
por la ganancia de la call condicionada al precio final de la mercancía, como se muestra en la Tabla 12.2. La
ganancia es la diferencia entre el valor de ejercicio de la call y el precio inicial de la call ajustado por el
interés. Noten que la ganancia final esperada de la cartera de la call es aproximadamente $0.1937, lo cual
parece indicar un error en el precio.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 235/358
El valor teórico de esta call utilizando la valuación (11.25) es $3.410, lo mismo que el valor de la put,
lo que significa que no hay error. La ganancia positiva surge de la aproximación implícita en la Tabla 12.2 y
el hecho que la ganancia de la call es una función no lineal del precio final de la mercancía. Para rectificar
este problema, debemos tener cuidado en establecer n mayor. Con un número mayor de pasos, la
discrepancia se reducirá. Por ejemplo si n 500 la ganancia final esperada es 0.0010 –un error de
aproximación de un décimo de centavo.
Tabla 12.1 Cálculo del precio final esperado de la mercancía, utilizando el enfoque de la
distribución discreta del precio de la mercancía con espacios iguales: S 50 b 0.04 T 0.25 , y
0.32 .
(1) Límite Integral Límite Integral
(2) (1)
Intervalo Precio de la Inferior Superior
N (d l ,i ) N (du ,i )
108 Por
No. Mercancía d l ,i d u ,i
107 (2)
1 26.2909 4.8692 3.2371 0.0006 0.0159
2 33.1177 3.2371 1.9441 0.0253 0.8391
3 39.9445 1.9441 0.8733 0.1653 6.6028
4 46.7713 0.8733 .-0.0405 0.3249 15.1975
5 53.5981 -0.0406 -0.8377 0.2827 15.1536
6 60.4249 -0.8377 -1.5446 0.1399 8.4521
7 67.2517 -1.5446 -2.1796 0.0466 3.1326
8 74.0785 -2.1796 -2.7559 0.0117 0.8681
9 80.9053 -2.7559 -3.2836 0.0024 0.1953
10 87.7321 -3.2836 -3.7702 0.0004 0.0378
11 94.5589 -3.7702 -4.2216 0.0001 0.0066
~
E ( S T ) 50.5013
107
Si ,T es el precio final de la mercancía en el punto medio del intervalo i ésimo .
108
N (dl ,i ) N (d u ,i ) es la probabilidad que el precio final de la mercancía caiga en el intervalo i ésimo .
109
El rebalanceo dinámico de una cartera que consiste de la mercancía y pagarés del tesoro es otra manera de imitar o duplicar una
opción a largo plazo. Discutimos esta posibilidad en el Capítulo 14 bajo el título “Seguro de una Cartera Dinámica”.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 236/358
finales de todas las opciones m a corto plazo, VSTj ( Si ,t ) , que se asumen como disponibles, j 1,, m . Se
utiliza los valores de la opción a largo plazo como la variable dependiente y los valores de la opción a corto
plazo como las variables independientes, y se realiza una regresión que minimiza la suma de los errores
2
cuadrados, Mini1 pi [VLT ( Si ,t ) b0 j 1 b j [VST , j ( Si ,t )]
n m
Los coeficientes de regresión estimados, bˆ j , j 1,, m , son las cantidades de las inversiones en las opciones
a corto plazo. El término de intersección estimado, b0 es la cantidad invertida en el activo libre de riesgo. Se
puede chequear lo bien que está funcionando la técnica comparando el valor actual de la cartera de opciones
a corto plazo con el valor teórico de la opción a largo plazo.110
Tabla 12.2 Cálculo de la ganancia final esperada de una posición call larga, cuyo precio es
c 3.410 . Los parámetros de precio son: S 50 , X 50 , b 0.04 , T 0.25 , r 0.06 , y
0.32 .
(2)
Intervalo Precio de la (1)
112 Ganancia (1) Por (2)
No. Mercancía111 Si ,T N (d l ,i ) N (du ,i )
113 i,T
EJEMPLO 12.4
Asuman un inversionista que posee una cartera de mercancía y quiere comprar una put europea con un precio
de ejercicio de 100 y un vencimiento de un año. El precio actual de la mercancía es 100, la tasa de costo de
acarreo es 4 porciento y la volatilidad es 32 porciento. La tasa de interés libre de riesgo es 6 porciento. El
precio teórico de esta opción es $10.3887 sobre la base de la ecuación de valuación de opciones europeas
(11.28). Sin embargo, no existe este tipo de opción a largo plazo.
El inversionista está considerando comprar una cartera de opciones put a tres meses que se pueda
utilizar para duplicar la opción a un año durante los próximos tres meses. En tres meses, se puede establecer
110
Choie y Novomestsky (1989) señalan que si el valor final de la cartera de la opción a corto plazo corresponde al valor de la
opción a largo plazo para todos los niveles del precio de la mercancía en el momento t, entonces en la ausencia de oportunidades
de arbitraje libre de costo en el mercado, el valor actual de la cartera de la opción a corto plazo es igual al valor actual de la opción
a largo plazo.
111
Si ,T es el precio final de la mercancía en el punto medio del intervalo i esim o .
112
N ( d l ,i ) N ( d u ,i ) es la probabilidad que el precio final de la mercancía caiga en el intervalo i esim o .
113
i ,T max(0, Si ,T X ) ce rT
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 237/358
una nueva posición a corto plazo para imitar la entonces opción de nueve meses.114 Están disponibles siete
opciones a tres meses, la composición de la cartera es
Con la excepción de recortar la put de 110, todas las otras puts son compradas. La suma de los pesos de la
cartera por el precio de los activos, 10.3887 es igual al precio de la put a largo plazo. (El precio del pagaré
del Tesoro se asume como e 0.06( 0.25) 0.9851 . En la Figura 12.27 podemos ver una comparación del valor de
la put a largo plazo real con el valor simulado de la put. Noten lo cercanos que son los valores hasta que el
precio de la mercancía se vuelve muy alto.
Este procedimiento puede ser refinado para tomar en cuenta las restricciones de no-negatividad,
liquidez del mercado y los errores en precios observados de las opciones a corto plazo. También son posibles
cambios en la volatilidad a lo largo de la vida de la opción a largo plazo.115 Nuestro enfoque asume que las
compras y las ventas de opciones a corto plazo se permiten libremente en cualquier cantidad demandada.
También asumimos que los precios de la opción están de acuerdo con las ecuaciones de valuación de
opciones europeas (11.25) y (11.28) y que la tasa de volatilidad es constante durante la vida de la opción a
largo plazo.
Los administradores de cartera interesados en las opciones con fechas a largo plazo pueden crearlas
como se indicó anteriormente o, como se da con más frecuencia, comprarlas a bancos de inversión en una
114
En la práctica, el uso de opciones a corto plazo con más de una fecha de vencimiento (por ejemplo, opciones a tres meses y seis
meses) y/o el refinanciamiento de posiciones de opciones a corto plazo antes de su vencimiento puede proporcionar una
duplicación más efectiva de la posición de la opción a largo plazo.
115
Ver, por ejemplo, Jamshidian y Zhu (1990).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 238/358
transacción fuera de bolsa. El banco de inversión vende la opción por su cuenta y cubre su posición tomando
una posición de compensación en la cartera de imitación de opciones a corto plazo.
Precio de la Opción
Mercancía S
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 239/358
100
In-the-money call
80
60
Out
Expected Return
In-the-money put 40 At
20
0 Commodity
-15 -10 -5 -20 0 5 10 15
At
-40
-60 Riskless Asset
Out
-80
Beta
Enfocándose primero en la medida de riesgo beta, podemos encontrar los atributos riesgo/retorno-esperado
de las opciones al encontrar primero sus betas respectivos, y luego encontrar los retornos esperados en
equilibrio basados en sus betas. Por ejemplo, el beta para la call in-the-money, que tiene una elasticidad de
c 5.289 , se puede calcular como c c S = 5.289(l.20) = 6.35.
Asumiendo que el mercado de capital está en equilibrio, el retorno esperado sobre la mercancía es
E ( RS ) r [ E ( RM ) r ] S .
Sustituyendo E ( RS ) , el retorno esperado de la mercancía, y por r , la tasa de interés libre de riesgo,
encontramos que [ E ( RM ) r ] S es igual a 0.10. Para encontrar el retorno esperado para la call in-the-
money, utilizamos una vez más la línea del mercado de valores del modelo de precio del activo de capital:
E ( Rc ) r [ E ( RM ) r ] c r [ E ( RM ) r ]c S 0.06 0.10c 0.06 0.10(5.289) 58.89%
Utilizando un procedimiento similar para las opciones restantes, encontramos los siguientes retornos
esperados y betas:
La Figura 12.28a ilustra estos resultados. La relación beta/retorno-esperado descrita en la Figura 12.28a es
asombrosa. Los betas y retornos esperados de las opciones son drásticamente diferentes del retorno-esperado
y beta del subyacente. Las posiciones de la call larga, por ejemplo, tienen retornos esperados y betas muy
altos –de hecho, varias veces mayores que los del subyacente. La ilustración también muestra que el retorno-
esperado y el beta de la call largo se incrementan conforme la call se hace cada vez más out-of-the-money.
Por otro lado, vemos que las puts generalmente tienen retornos esperados negativos y betas negativos, y
mientras más out-of-the-money esté la put, menor (más negativo) es su retorno y su beta esperado.
Los administradores de cartera también están interesados en conocer el nivel de la volatilidad del
retorno. Como vimos anteriormente, la volatilidad del retorno de la opción es simplemente la elasticidad del
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 240/358
precio de la opción con respecto al precio de la mercancía por la volatilidad del retorno del subyacente. En la
ilustración, la volatilidad del retorno de la mercancía es 40 porciento. Por lo tanto las volatilidades de las
opciones son
Opción Elasticidad Volatilidad del Retorno
Call 45 5.289 211.56
Call 50 6.636 265.44
Call 55 8.069 322.76
Put 45 -7.559 302.36
Put 50 -5.920 236.80
Put 55 -4.645 185.80
Mercancía 1 40.00
Activo Libre de Riesgo 0 0
La Figura 11.28b ilustra estos resultados. El grado de riesgo extremo de las opciones se confirma aún más
con estos valores. Donde la volatilidad del retorno del subyacente es 40 porciento, las volatilidades del
retorno de la opción exceden, en algunos casos, varios cientos porciento.
Sin embargo, combinando opciones con una posición en el subyacente puede ayudar a reducir el
riesgo. Por ejemplo, una estrategia de venta de call cubierta (es decir, escribir una call contra una mercancía
subyacente) o una estrategia put protectora (es decir, comprar una put contra el subyacente) reduce el riesgo
de la posición global. El retorno esperado, el beta y la volatilidad del retorno de una cartera que consiste de
una opción y una mercancía subyacente pueden calcularse utilizando las siguientes ecuaciones.
E ( R p ) X S E ( RS ) (1 X S ) E ( Ro ) , (12.25)
p X S S (1 X S ) o , (12.26)
y
p X 2 2 (1 X S ) 2 2 2d o X (1 X S ) o
S S o S S
(12.27)
dondel subíndice o indica opción y la variable indicadora d o 1 es para calls y d o 1 es para puts (es
decir, los retornos de la call [put] están perfecta y positivamente [negativamente] correlacionados con los
retornos de la mercancía). El peso de X S es la proporción de dólares S invertidos directamente en la
mercancía, esto es,
S noOo
XS (12.28)
S
donde no es el número de opciones compradas (es decir, un valor positivo de no indica que las opciones son
compradas y un valor negativo indica que las opciones son vendidas) y Oo es el valor de mercado de cada
opción. Noten que cuando no es igual a cero, todo el valor de la cartera es invertido en la mercancía. El valor
1 X S es la proporción de la inversión original en las opciones.
Para reforzar estos procedimientos, reconsideren la ilustración anterior y asuman que se crea una call
cubierta al vender la call in-the-money contra una posición larga en la mercancía. Los ingresos de la venta de
la call son invertidos en la mercancía de manera que la inversión total en la mercancía es
50-(-1)(7.061)=57.061.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 241/358
100
80 Out
Commodity
60
Expected Return
At
40
Riskless Asset In-the-money call
20
0
-20 0 100 200 300 400
-40 Out
-60 In-the-money put
At
-80
Volatility
En otras palabras, el escribir la call in-the-money contra el subyacente reduce el retorno esperado y el riesgo
de la cartera subyacente. De hecho, todos y cada una de las ventas de calls cubiertos y todas y cada una de
las compras de put protectores compartirán estos atributos. Para nuestra ilustración, las características de la
mercancía y las seis cartera mercancía /opción son
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 242/358
Expected Return
14
12
10
8 5
6 3
4 1
2 Riskless Asset
0
0 0.5 1
Beta
14
12
10 5
8
6 3
4 1
2
Riskless Asset
0
0 10 20 30 40
Volatility
Estas cantidades muestran claramente que las ventas de la call cubierto y las compras de put protector sirven
para reducir el retorno esperado y el riesgo de la cartera.116 Para los calls cubiertos, la reducción
retorno/riesgo se hace mayor mientras más in-the-money esté la opción. Esto es simplemente porque el que
vende la opción está dispuesto a aceptar más efectivo (es decir, prima de la opción) a cambio del potencial de
movimientos hacia arriba en el precio de la mercancía. Esta actividad es completamente análoga a retirar la
inversión de la mercancía y e invertir en el activo libre de riesgo. De hecho, dado que la call 45 tiene un delta
de .747, la cartera de la call cubierto 45 tiene un valor delta neto de 1.141-0.747=0.394. Si creamos una
cartera que consiste de 0.394 en la mercancía y 0.606 en el activo libre de riesgo, el retorno esperado, el beta
y la volatilidad del retorno son:
116
Recuerden que el retorno esperado y el beta de las posiciones de la opción se mantienen para la siguiente instancia en la cual las
variables subyacentes no cambian mucho.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 243/358
50+4(7.061)= 78.244.
40
30 1
Expected Return
20 3 2
5 6 6
10
4
0
0 20 40 60 80 100
-10
-20
Volatility
En la Figura, vemos la mercancía con un retorno esperado de 16 porciento y una volatilidad de 40 porciento.
El punto llamado “venta call 1:1” muestra que el retorno esperado y la volatilidad de la cartera se reducen
cuando se escribe una sola call 45 contra la mercancía. Conforme se escriben o venden más calls, el retorno
esperado y la volatilidad continúan disminuyendo hasta que finalmente, cuando se escriben 1.65 calls contra
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 244/358
la mercancía, la cartera está libre de riesgo.117 Más allá de este número, si se escriben más calls, el retorno
esperado continúa disminuyendo y la volatilidad se incrementa.
Eventualmente, el retorno esperado de la cartera se vuelve negativo, y, si el número de calls escritos
continúa aumentando, la volatilidad comienza a exceder el 40 porciento. La venta de la call 4:1 se mantiene
aislada. La cartera, por ejemplo, tiene una mayor volatilidad que el número crítico de la mercancía de los
calls escritos contra la mercancía para generar una volatilidad de 40 porciento y es 3.30, exactamente el
doble del número de calls para generar la cobertura libre de riesgo.118
V i1 ni Oi nS S B
N
(12.29)
La derivada parcial del valor de cartera con respecto a un cambio en los determinantes de la opción, k , es
V O S B
i1 ni i nS
N
(12.30)
k k k k
117
Antes ilustramos que una posición call cubierta era análoga a una cartera que consiste de alguna fortuna en la mercancía y
alguna fortuna en el activo libre de riesgo. La cantidad invertida en la mercancía es ( S nc c nc S c ) . Si establecemos este valor
igual a cero (es decir, toda el valor se invierte en el activo libre de riesgo) y despejamos para nc , obtenemos nc 1/( c c / S ) .
Sustituyendo los valores del ejemplo, nc 1/(0.747 7.061/50) 1.65 .
118
En el último pie de página, utilizamos el hecho que un cal cubierto es como una cartera que consiste de alguna fortuna en la
mercancía y alguna fortuna en el activo libre de riesgo para poder deducir la composición de la cobertura libre de riesgo. La
cantidad invertida en la mercancía es ( S nc c nc S c ) . Si establecemos este valor igual a S (es decir, una venta corta de la
mercancía subyacente) y despejamos pata nc , obtenemos nc 2 /( c c / S ) . Sustituyendo en los valores del ejemplo,
nc 2/(0.747 - 7.061/50) 3.30 .
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 245/358
Ahora, regresemos al problema de manejar los cambios en el precio de la mercancía. El cambio en el valor
de la cartera con respecto a un cambio en el precio de la mercancía (es decir, el delta de la cartera) es
V i 1 ni o nS (12.31)
N
Noten que por premisa, el valor del activo de riesgo no cambio con los cambios en el precio de la mercancía.
Para inmunizar esta cartera contra los cambios en el precio de la mercancía, simplemente calculamos V y
luego tomamos una posición en opciones o el subyacente que forma el valor cero del delta de la cartera.
EJEMPLO 12.5
Asuman un creador de mercado de opciones de futuros tiene posiciones largas de 150 calls con un precio de
ejercicio de 45 y un tiempo hasta el vencimiento de dos meses, 200 puts con un precio de ejercicio de 50 y
tiempo hasta el vencimiento de tres meses, y 225 calls con un precio de ejercicio de 55 y un tiempo hasta
vencimiento de tres meses. En lugar de enfrentar el riesgo que el precio de los futuros subyacentes pueda
moverse significativamente de la noche a la mañana, él decide cubrir la posición utilizando futuros o calls
con un precio de ejercicio de 50 y un tiempo hasta el vencimiento de tres meses. Comparen la efectividad de
utilizar futuros y la call 50 al crear una cobertura neutral al riesgo. Asuman que el precio actual de los futuros
es $50, la call 50 tiene un precio de $2.455 y tiene un delta de 0.5171, la tasa de interés libre de riesgo es 6
porciento y la volatilidad es 25 porciento.
El delta de la cartera de 87.75 nos dice que mantener esta cartera es como mantener una posición larga en
87.75 contratos de futuros. Para crear una cartera neutral al delta, podemos (a) vender 87.75 contratos de
futuros o (b) vender 87.75/0.5171=169.70 calls. La Figura 12.31 muestra la efectividad de cada cobertura.
La Figura 12.31 demuestra que tanto la cobertura de futuros neutral al delta como la cobertura de
opción neutral al delta reducen el rango de los valores de cartera posibles. La cartera sin cobertura tiene un
rango de valor entre 1400 y 3600 sobre el rango de los precios de la mercancía –la cobertura de futuros tiene
entre 1475 y 3000 y la cobertura de opciones entre 1475 y 2500. Claramente, la cobertura de opciones es la
más efectiva.
La razón por la que la cobertura de opciones resulta ser la más efectiva es que tiene que ver con
gamma –el cambio en delta conforme cambia el precio de la mercancía. Conforme cambia el precio de la
mercancía, el valor del delta de la cartera de opciones cambia. De hecho, el gamma de la cartera de la opción
es 30.08, como se muestra a continuación. El contrato de futuros tiene un gamma igual a cero, de manera que
la cobertura de futuros neutral al delta aún tiene un gamma de 30.08. Por otro lado, las 169.70 calls que
vendimos tiene un gamma de –169.70(0.062758)=-10.65, de manera que el gamma de la cobertura de la
opción neutral al delta es 19.43.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 246/358
No obstante, la reducción del gamma con la cobertura de opciones es incidental en este caso, de
manera que ahora ilustraremos cómo tomar en cuenta tanto el delta como el gamma en la cobertura de una
cartera contra los movimientos en el precio.
EJEMPLO 12.26
Una vez más, estamos considerando al creador de mercado descrito en el Ejemplo 12.5. Su posición de
cartera es
Tipo de Precio de Tiempo Hasta el Precio de la
Cantidad Delta Gamma
Opción Ejercicio Vencimiento Opción
150 Call 45 0.16667 5.325 0.852 0.04304
200 Put 50 0.25 2.455 -0-468 0.06276
225 Call 55 0.25 0.828 0.238 0.04922
Como en el ejemplo anterior, el valor de la cartera es 1476.05 y el delta de la cartera es 87.75. El gamma de
la cartera es 87.75. El gamma de la cartera es
V 150(0.04304) 200(0.06276) 225(0.04922) 30.08 .
Para cubrir tanto el riesgo del delta como del gamma se requieren por lo menos dos opciones (es decir, los
futuros tienen un gamma igual a cero, de manera que no son efectivos para ajustar el riesgo de gamma de la
cartera). Además de la call 50, la cual está disponible en el Ejemplo 12.5, también asumimos que está
disponible una put 55 a tres meses. Su precio es $7.754, su delta es -0.7468 y su gamma es –0.04922. La call
50 tiene un gamma de 0.06276.
Para calcular la cobertura óptima neutral al delta/gamma de estas dos opciones, resolvemos el
siguiente sistema de ecuaciones. Queremos que la cartera sea neutral al delta, de manera que
Portfolio Value
Unhedged Portfolio
Commodity Price
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 247/358
Para despejar, podemos aislar nc de la primera ecuación, sustituir en la segunda y despejar n p . El valor de n p
es –138.93. Luego sustituimos para n p en la primera ecuación y encontramos que nc es –370.34.
El valor de la cartera neutral al delta y al gamma se traza en la Figura 12.32, junto con el valor de la
cartera no cubierta. Claramente, la cobertura es efectiva. Donde la cartera no cubierta varía en más de 2000
sobre el rango del precio de la mercancía mostrado, la cartera protegida parece variar en menos de 200.
Noten que en el Ejemplo 12.6, se asumen que sólo están disponibles dos opciones para establecer la
cobertura neutral al delta/neutral al gama. En la práctica, están disponibles muchas opciones con las cuales
establecer la cobertura. Se necesitan por lo menos dos opciones para ejecutar una cobertura, pero se pueden
utilizar más opciones. Algunas veces se utiliza la programación lineal para encontrar el conjunto de opciones
de costo mínimo que eliminará el riesgo del delta y gamma.
Figura 12.32 Valor de Cartera como una Función del Precio de la Mercancía
Portfolio Value
Unhedged Portfolio
Commodity Price
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 248/358
EJEMPLO 12.7
Una vez más, estamos considerando al creador de mercado en los contratos de opciones de futuros descrito
en el Ejemplo 12.5. Sin embargo, la cartera de este creador de mercado es diferente. Él está corto 180 calls
50 a tres meses y 200 puts 50 a tres meses.
Noten que las opciones de futuros at-the-money tienen el mismo precio. La cartera es claramente neutral al
delta, como se muestra a continuación:
V -180(0.517) - 200(-0.468) 0.54 .
Portfolio Value
Unhedged Portfolio
Vega-hedged Portfolio
Commodity Volatility( %)
Esto significa que si la volatilidad se incrementa de su nivel actual de 25 porciento a, digamos, 26 porciento,
el valor de la cartera disminuirá en 37.26.
Para cubrir este riesgo, asuman que está disponible la put 55 a tres meses del Ejemplo 12.6. La put 55
tiene un vega de 7.69. Para poder eliminar el riesgo de vega de la cartera debemos comprar
3,726.28
np 484.56 puts. La Figura 12.33 muestra la efectividad de este procedimiento.
7.69
Claramente, la cobertura del vega se muestra como efectiva para eliminar los efectos de los cambios en la
volatilidad. Noten, sin embargo, que la compra de puts 55 afecta drásticamente el delta de la cartera. Ahora
está en un nivel de 484.56(-0.7468)+0.54-361.33. Este ejemplo está diseñado sólo para mostrar cómo se
maneja el riesgo del vega. Obviamente, considerar simultáneamente el delta, el gamma y el vega puede ser
razonable; esto se pude realizar con tres o más opciones disponibles.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 249/358
Option Price
Days to Expiration
EJEMPLO 12.8
Consideren la cartera del creador del mercado de una opción de futuros del Ejemplo 12.7. ¿Cómo puede esta
persona eliminar el deterioro del tiempo en su cartera utilizando la call 50?
lo cual significa que, durante el siguiente día, el valor de la cartera disminuirá en 3,137.90(1/365)=8.597.
Se puede crear una cobertura theta utilizando la call 50. Su theta es 4.756. Para eliminar el deterioro del
tiempo, debemos vender 3,137.90/4.756=659.78 contratos.
En esta sección, hemos mostrado cómo se pueden utilizar los valores delta, gamma, vega y theta de
una cartera para manejar de manera efectiva el valor de una cartera de opciones. Aunque las ilustraciones se
enfocan típicamente en una sola dimensión a la vez, está claro que es necesario un mantenimiento diario de
las posiciones de la cartera.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 250/358
12.5 RESUMEN
En este capítulo, hemos logrado cuatro cosas. Primero, hemos desarrollado y analizado más de cincuenta
estrategias de negociación de opciones. Cada estrategia se formó desde sus componentes básicos y se ilustró
con un diagrama de ganancia. Se proporcionaron los precios de las mercancías, las pérdidas máximas y
ganancias máximas. Segundo, mostramos que utilizando la premisa de distribución lognormal del precio de
la mercancía del Capítulo 11, podemos calcular las probabilidades de pérdidas y ganancias, así como la
ganancia esperada para cada estrategia de negociación. En la tercera sección, mostramos que se puede
utilizar un enfoque de regresión para crear opciones a largo plazo a partir de opciones a corto plazo.
Finalmente, hemos discutido el manejo del riesgo diario de las carteras de opciones.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 251/358
13.6 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 252/358
Common stock option contracts have been traded in the U.S. for many decades. Trading began in the over-
the-counter market more than fifty years ago. In April 1973, the Chicago Board Options Exchange (CBOE)
became the first organized secondary market in call options on sixteen NYSE common stocks. In 1977, the
CBOE introduced put options on stocks. Today both call and put options are traded on over five hundred
different stocks and on five exchanges. In addition to the CBOE, active secondary stock option markets exist
on the American Stock Exchange, the Pacific Coast Exchange, the Philadelphia Stock, Exchange and the
New York Stock Exchange.
This chapter focuses on stock options. In the first section, the stock option market is described. In the
second section, we adapt the arbitrage pricing, relations of Chapter 10 to stock option contracts. The
principles are modified somewhat to account for the fact that common stocks typically pay discrete dividends
during the option’s life. In section 3, we value European and American call options on dividend-paying,
stocks. Even though an American call option can, be exercised early, a valuation equation for this option
exists. For American put options on stocks, no valuation equation exists. In section 4, the binomial method is
used to approximate the value of American put options on dividend-paying, stocks. Although the application
is specific, the binomial method can be applied to valuation of virtually any type of option. This method is
applied again in chapter 15, for example, to value interest rate options. Finally in section 5, warrants used to
raise new capital are studied. Warrants are exercised, the company’s equity may be diluted, something that
must be incorporated into warrant valuation.
119
For practical purposes, assuming the option expires Friday seems prudent since both the stock market and the option market are
closed on Saturday.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 253/358
120
The exercise price is, however, adjusted for stock splits and stock dividends. Where the stock split/dividend produces a
fractional result, the exercise price is rounded to the nearest eighth.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 254/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 255/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 256/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 257/358
Non-Dividend-Paying Stocks
To derive the lower price bounds and the put-call parity relations for options on non-dividend-paying stocks,
simply set the cost-of-carry rate, b , equal to the riskless rate of interest, r , in the relations presented in
Chapter 10. The only cost of carrying the stock is interest. The lower price bounds for the European call and
put options are
c( S , T ; X ) max[0, S Xe rT ] (13.1a)
and
p( S , T ; X ) max[0, Xe rT S ] , (13.1b)
respectively, and the lower price bounds for the American call and put options are
C ( S , T ; X ) max[0, S X ] (13.2a)
and
P( S , T ; X ) max[0, Xe rT S ] (13.2b)
respectively. The put-call parity relation for non-dividend-paying European stock options121 is
c( S , T ; X ) p( S , T ; X ) S Xe rT , (13.3a)
and the put-call parity relation for American options on non-dividend-paying stocks is
S X C ( S , T ; X ) P( S , T ; X ) S Xe rT (13.3 b)
For non-dividend-paying stock options, the American call option will not rationally be exercised early, while
the American put option may be.122
121
The original formulation of put-call parity, for European stock options is contained in Stoll (1969).
122
For proofs of any of the relations (13.1a) through (13.3b), see Chapter 10.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 258/358
Dividend-Paying Stocks
If dividends are paid during the option’s life, the above relations must reflect the stock’s drop in value when
the dividends are paid. To manage this modification, we assume that the underlying stock pays a single
dividend during the option’s life at a time that is known with certainty. The dividend amount is D and the
time to ex-dividend is t .123
If the amount and the timing of the dividend payment is known, the lower price bound for the
European call option on a stock is
c( S , T ; X ) max[0, S De rt Xe rT ] . (13.4a)
In this relation, the current stock price is reduced by the present value of the promised dividend. Because a
European-style option cannot be exercised before maturity, the call option holder has no opportunity to
exercise the option while the stock is selling cum dividend. In other words, to the call option holder, the
current value of the underlying stock is its observed market price less the amount that the promised dividend
contributes to the current stock value, that is, S De rt . To prove this pricing relation, we use the same
arbitrage transactions as in Chapter 10, except we use the reduced stock price S De rt in place of S .
The lower price bound for the European put option on a stock is
p( S , T ; X ) max[0, Xe rT S De rt ] . (1 3.4b)
Again, the stock price is reduced by the present value of the promised dividend. Unlike the call option case,
however, this serves to increase the lower price bound of the European put option. Because the put option is
the right to sell the underlying stock at a fixed price, a discrete drop in the stock price such as that induced by
the payment of a dividend serves to increase the value of the option. An arbitrage proof of this relation is
straightforward when the stock price, net of the present value of the dividend is used in place of the
commodity price.
The lower price bounds for American stock options are slightly more complex. In the case of the
American call option, for example, it may be optimal to exercise just prior to the dividend payment because
the stock price falls by an amount D when the dividend is paid. The lower price bound of an American call
option expiring at the ex-dividend instant would be 0 or S Xe rt , whichever is greater. On the other hand, it
may be optimal to wait until the call option’s expiration to exercise. The lower price bound for a call option
expiring normally is (13.4a).124 Combining the two results, we get
The last two terms on the right-hand side of (13.5a) provide important guidance in deciding whether to
exercise the American call option early, just prior to the ex-dividend instant. The second term in the squared
brackets is the present value of the early exercise proceeds of the call. If the amount is less than the lower
price bound of the call that expires normally, that is, if
the American call option will not be exercised just prior to the ex-dividend instant.
123
Generalizations of the results to cases where there are more than one known dividend are derived in the same manner as the
single dividend results shown here.
124
Recall that in Chapter 10 we showed that an American call is never optimally exercised early if b r . Between dividends, the
cost-of-carry rate is r , so exercise is not optimal. At the ex-dividend instant, however, the call option holder may wish to exercise
to capture the dividend.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 259/358
D X [1 e r (T t ) ] (13.7)
In other words, the American call will not be exercised early if the dividend captured by exercising prior to
the ex-dividend date is less than the interest implicitly earned by deferring exercise until expiration.
Figure 13.1 depicts a case in which early exercise could occur at the ex-dividend instant, t . Just prior
to ex-dividend, the call option may be exercised yielding proceeds St D X , where St is the ex-dividend
stock price. An instant later, the option is left unexercised with value c( St , T t; X ) , where c() is the
European call option formula. Thus, if the ex-dividend stock price, St , is above the critical ex-dividend stock
price where the two functions intersect, S *t , the option holder will choose to exercise her option early just
prior to the ex-dividend instant. On the other hand, if S t S *t , the option holder will choose to leave her
position open until the option’s expiration.
c(St* ,T t;X)
Ex - D stock Price S t
X/e r (T t ) X D S t*
Figure 13.2 depicts a case in which early exercise will not occur at the ex-dividend instant, t . Early exercise
will not occur if the functions, St D X and c(St , T - t; X) do not intersect, as is depicted in Figure 13.2.
r(T - t)
In this case, the lower boundary condition of the European call, S t Xe , lies above the early exercise
proceeds, St D X , and hence the call option will not be exercised early. Stated explicitly, early exercise
is not rational if
St D X St Xe r(T - t)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 260/358
Exercise value of
c(S,T t;X)
America Call Option, S t D X
Ex - D stock Price S t
X/e r (T t ) X D
This condition for no early exercise is the same as (13.6), where St is the ex-dividend stock price and where
the investor is standing at the ex-dividend instant, t . The condition can also be written as
D X [1 e r(T - t) ] (13.7)
In words, if the ex-dividend stock price decline-the dividend-is less than the present value of the interest
income that would be earned by deferring exercise until expiration, early exercise will not occur. When
condition (I 3.7) is met, the value of the American call is simply the value of the corresponding European
call.
The lower price bound of an American put option is somewhat different. In the absence of a dividend,
an American put may be exercised early. In the presence of a dividend payment, however, there is a period
just prior to the ex-dividend date when early exercise is suboptimal. In that period, the interest earned on the
exercise proceeds of the option is less than the drop in the stock price from the payment of the dividend. If tn
represents a time prior to the dividend payment at time t , early exercise is suboptimal, where ( St X )e r(t - t n )
is less than ( X S D ). Rearranging, early exercise will not occur between tn and t if125
ln(1 XD S )
tn t (13.8)
r
125
It is possible that the dividend payment is so large that early exercise prior to the dividend payment is completely precluded.
For example, consider the case where X 50, S 40, D 1, t 0.25 and r 0.10 . Early exercise is precluded if
tn 0.25 ln[1 1/(50 40)]/ 0.10 0.7031 . Because the value is negative, the implication is that there is no time during the current
dividend period (i.e., from 0 to t ) where it will not pay the American put option holder to wait until the dividend is paid to exercise
his option.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 261/358
Early exercise will become a possibility again immediately after the dividend is paid. Overall, the lower
price bound of the American put option is
P( S , T ; X ) max[0, X ( S De rt )] . (13.5b)
Put-call parity for European options on dividend-paying stocks also reflects the fact that the current stock
price is deflated by the present value of the promised dividend, that is,
c( S , T ; X ) p( S , T ; X ) S De rt Xe rT . (13.9)
That the presence of the dividend reduces the value of the call and increases the value of the put is again
reflected here by the fact that the term on the right-hand side of (13.9) is smaller than it would be if the stock
paid no dividend.
Put-call parity for American options on dividend-paying stocks is represented by a pair of
inequalities, that is,
S De rt X C ( S , T ; X ) P( S , T ; X ) S De rt Xe rT . (13.10)
To prove the put-call parity relation (13.10), we consider each inequality in turn.
The left-hand side condition of (13.10) can be derived by considering the values of a portfolio that consists
of buying a call, selling a put, selling the stock, and lending X De rt risklessly. Table 13.2 contains these
portfolio values.
In Table 13.2, it can be seen that, if all of the security positions stay open until expiration, the
terminal value of the portfolio will be positive, independent of whether the terminal stock price is above or
below the exercise price of the options. If the terminal stock price is above the exercise price, the call option
is exercised, and the stock acquired at exercise price X is used to deliver, in part, against the short stock
position. If the terminal stock price is below the exercise price, the put is exercised. The stock received in the
exercise of the put is used to cover the short stock position established at the outset. In the event the put is
exercised early at time , the investment in the riskless bonds is more than sufficient to cover the payment
of the exercise price to the put option holder, and the stock received from the exercise of the put is used to
cover the stock sold when the portfolio was formed. In addition, an open call option position that may still
have value remains. In other words, by forming the portfolio of securities in the proportions noted above, we
have formed a portfolio that will never have a negative future value. If the future value is certain to be
nonnegative, the initial value must be nonpositive, or the left-hand inequality of (13.10) holds.
The right-hand side of (13.10) may be derived by considering the portfolio used to prove European
put-call parity. Table 13.3 contains the arbitrage portfolio transactions. In this case, the terminal value of the
portfolio is certain to equal zero, should the option positions stay open until that time. In the event the
American call option holder decides to exercise the call option early, the portfolio holder uses his long stock
position to cover his stock obligation on the exercised call and uses the exercise proceeds to retire his
outstanding debt. After these actions are taken, the portfolio holder still has. an open long put position and
cash in the amount of X [1 e r (T t ) ] . Since the portfolio is certain to have nonnegative outcomes, the initial
value must be nonpositive or the right-hand inequality of (13.10) must hold.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 262/358
TABLE 13.2 Arbitrage transactions for establishing put-call parity for American stock options.
S De rt X C ( S , T ; X ) P( S , T ; X )
Ex-Dividend Put Exercised Put Exercised
Day ( t ) Early ( ) Normally ( T )
Intermediate Terminal Value
Position Initial Value ~ ~
Value ST X ST X
~ ~
Buy American Call C C 0 ST X
~ ~
Sell American Put P ( X S ) ( X ST ) 0
~ ~ ~
Sell Stock S D S ST ST
Lend De rt De rt D
Lend X X Xe r XerT Xe rT
~
Net Portfolio Value C P S De rt X 0 C X [e r 1] X [e rT 1] X [e rT 1]
ln(S / X (r 0.5 2 )T
where d1 , and d 2 d1 T
T
TABLE 13.3 Arbitrage transactions for establishing put-call parity for American stock options.
C ( S , T ; X ) P( S , T ; X ) S De rt Xe rT
Ex-
Put Exercised Put Exercised
Dividend
Early ( ) Normally ( T )
Day ( t )
Intermediate Terminal Value
Position Initial Value ~ ~
Value ST X ST X
~ ~
Sell American Call C ( S X ) 0 ( ST X )
~ ~
Buy American Put P P X ST 0
~ ~ ~
Buy Stock S D S ST ST
Borrow De rt De rt D
Borrow Xe rT Xe rT Xe r (T ) X X
~
Net Portfolio Value C P S De rt Xe rT 0 P X [1 e r (T ) ] 0 0
126
This equation is often referred to as simply the Black-Scholes formula, given the important impact that the Black-Scholes
(1973) paper has had on the theory of option pricing and, more generally, the practice of finance.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 263/358
S x S De rt (13.12)
is substituted for the stock price parameter in the European call option pricing formula (13.11). The valuation
equation of a European call option on a dividend-paying stock is
c( S , T ; X ) S x N (d1 ) Xe rT N (d 2 ) , (13.13)
ln(S x / X (r 0.5 2 )T
where: d1 , and d 2 d1 T
T
Naturally, the value of the call decreases as a result of the cash disbursement on the stock.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 264/358
Assuming that the future stock price net of the present value of the promised dividend is lognonnally
distributed, the expected values on the right-hand side of (13.14) become the valuation equation of an
American call option on a dividend-paying stock:
C ( S , T ; X ) e rt [ S x e rT N1 (b1 ) ( X D) N1 (b2 )]
e rT [ S x e rT N 2 (a1 ,b1; t / T ) XN 2 (a2 ,b2 ; t / T )] (13.15a)
where
ln(S x / X ) (r 0.5 2 )T ln(S x / St* ) (r 0.5 2 )t
a1 , and a2 a1 T ; b1 , and b2 b1 t
T t
N1 (b) is the cumulative univariate normal density function with upper integral limit b 127 and N 2 (a, b; ) is
the cumulative bivariate normal density function with upper integral limits, a and b , and correlation
coefficient, .128 St* is the ex-dividend stock price for which
EXAMPLE 13.1
Compute the value of an American-style call option whose exercise price is $50 and whose time to
expiration is 90 days. Assume the riskless rate of interest is 10 percent annually, the underlying stock price is
$50, the standard deviation of the rate of return of the stock is 30 percent, and the stock pays a dividend of $2
in exactly 60 days.
127
Recall that the function N (b) was used for the first time in Chapter 11. The subscript 1 is used here only to contrast the
univariate integral from the bivariate integral.
128
More details about the bivariate normal probability, as well as a method of computation and a numerical example, are contained
in Appendix 13.1.
129
Roll (1977) provides a framework for analytically valuing the American call option. The valuation formula (13.15b) is from
Whaley (1981).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 265/358
We compute the value of this call first using the European formula (13.13) and then using the
American formula (13.15b). In this way, we can identify the value of the early exercise premium on the call.
The current stock price net of the present value of the promised dividend is
showing that early exercise is not precluded and that formula (13.15b) should be used.
The value of the American call is now computed as
The values b1 and b2 depend on the critical ex-dividend stock price St* , which is determined by
c( St* ,30 / 365;50) St* 2 50
and, in this example, equals 49.824. The bivariate normal probabilities are N 2 (a1 ,b1; ) 0.1135 and
N 2 (a2 ,b2 ; ) 0.1056 , and the univariate normal probabilities are N1 (b1 ) 0.4581 and N1 (b2 ) 0.4103 .
The value of the American call is 2.931; hence, the early exercise premium on the American option is 2.931 -
2.513 = 0.418.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 266/358
TABLE 13.4 Simulated American and European call option values oil a stock with a known discrete
dividend. The call option has an exercise price of $50 and a time to expiration of 90 days. The
riskless rate of interest is 10 percent, and the standard deviation of the stock return is 30 percent.
The stock pays a dividend of $2 in 60 days.
Stock Price European Call Price American Call Price Early Exercise Premium
S c( S , T ; X ) C (S ,T ; X ) C
40 0.126 0.136 0.011
45 0.760 0.867 0.107
50 2.515 2.931 0.418
55 5.610 6.481 0.871
60 9.726 10.974 1.248
The size of the early exercise premium of an American call option on a dividend-paying stock becomes
larger as the option goes deeper in the money. In Table 13.4, we extend the results of Example 13.1 by
allowing the stock price to vary from $40 to $60. It is interesting to note that the dividend payment induces a
fairly large early exercise premium on the call option, particularly when the call is deep in-the-money. At a
$60 stock price, for example, the value of the early exercise premium is about $1.25, more than 11 percent of
the call’s overall value.
Dividend Spreads
In practice, not all call options are exercised when they should be. And, when they are not, there are ways to
profit risklessly. Consider, for example, two in-the-money call options written on a stock that is about to pay
a dividend. Assume the deeper in-the-money call is sold and the other is purchased. Now, on the day before
ex-dividend, exercise the purchased option and wait. If the holder of the deeper in-the-money call exercises
her option before ex-dividend, deliver the stock received from the exercise of the purchased option and pay
the net difference between the exercise prices of the options. On the other hand, suppose the holder of the
deeper in-the-money call option forgets to exercise her option. In this case, sell the stock the next morning
and buy back the remaining option. In the first case, profit equals zero, and, in the second, a profit in the
amount of the dividend would be received. A numerical example may serve to clarify this strategy.
Assume that a stock is currently priced at $60 and will pay a $2.00 dividend tomorrow. Call options
with exercise prices of 50 and 55 and time to expiration of 30 days are priced at $10.01 and $5.01,
respectively. (The riskless rate of interest is assumed to be 10 percent, and the standard deviation of stock
returns is 30 percent.) Now, assume the 50 call is sold, and the 55 call is purchased, yielding net proceeds at
the outset of $5.00. At the end of the day, the investor exercises the 55 call, receiving proceeds St 1 + 2.00 -
55. (Day t 1 is the day before ex-dividend, and the notation St 1 is the stock price net of the value of the
escrowed dividend.) If the 50 call option is exercised against the investor before ex-dividend (which will not
be known until the next day before market opening), the investor’s obligation is - ( St + 2.00 - 50), and the
net terminal value is St + 2.00 - 55 - St - 2.00 + 50 or - $5, exactly a wash, considering $5 was collected up
front. However, if the 50 call option is not exercised, the investor goes into the next morning with a long
position in the stock (acquired from exercising the 55 call) and a short position in the 50 call, which has a
value St 2.00 55 Ct . For all intents and purposes, this position is riskless because the uncertainty in the
value of the long stock position is offset by that of the short in-the-money call. The outstanding call, being
deep in-the-money, is selling for about its floor value of S, - 50, so the net value of the position is 2.00 - 55 +
50 or - $3. Net of the initial cash receipt of $5, the profit is $2.00, exactly the amount of the dividend. This
strategy is usually called a dividend capture or a dividend spread.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 267/358
p ( S , T ; X ) Xe rT N ( d 2 ) SN (d1 ) (13-16)
where
ln(S / X ) (r 0.5 2 )T
d1 , and d 2 d1 T
T
This result follows straightforwardly from substituting the riskless rate of interest, r , for the cost-of-carry
rate, b , in equation (11.28) from Chapter 11.
p ( S , T ; X ) Xe rT N ( d 2 ) S x N (d1 ) , (13.17)
ln(S x / X ) (r 0.5 2 )T
where d1 , and d 2 d1 T
T
Note that the put value increases as a result of the cash disbursement on the stock.
u e t (1 3.18a)
and d 1/ u (13.18b)
130
See Cox, Ross, and Rubinstein (1979).
131
The factors are consistent with the Black-Scholes model. See Cox, Ross, and Rubinstein (1979) for details.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 268/358
FIGURE 13.3 Possible paths that the stock price may follow under the binomial model.
0 1 2 3 4 n (even) n (odd)
u n S0 u n S0
u 4 S0
3
u S0
2 2
u S0 u S0
uS0 uS0 uS0
S0 S0 S0 S0
dS0 dS0 dS0
2 2
d S0 d S0
3
d S0
4
d S0
d n S0 d n S0
FIGURE 13.4 Possible paths that the stock price may follow under the binomial model, where the current
stock price is 50, the riskless rate of interest is 10 percent annually, and the standard deviation of stock
returns is 30 percent annually. The time to expiration of the option is 90 days, and the number of time steps
is also 90. The time increment t is, therefore, 1 day or 0.00274 years.
Figure 13.4 provides a numerical illustration of the binomial lattice for the stock price. The current stock
price, S0 , is 50, the riskless rate of interest, r , is 10 percent, and the standard deviation of stock returns, ,
is 30 percent. The time to expiration of the option, T , is 90 days, and the number of time steps, n , is 90. The
size of the time increment t is, therefore, one day or 0.00274 years.
0 1 2 3 4 90 days
205.46
53.24
52.41
51.60 51.60
50.79 50.79
50 50 50 50
49.22 49.22
48.45 48.45
47.70
46.96
12.17
u e 0.30 0.00274
1.01583 and d 1 / 1.01583 0.98442
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 269/358
0 where S n , j X
Pn , j ( S n , j ) (13.19)
X - S n , j where S n , j X
The option values one step back in time (in column n 1 ) are computed by taking the present value of the
expected future value of the option. At any point j in column n 1 , the stock price can move up with
probability p or down with probability 1 p . The value of the option at time n if the stock price moves up
is Pnj 1 , and if the stock price moves down is Pnj . The present value of the expected future value of the
option is, therefore,
( p ) Pn , j 1 (1 p ) Pn , j
Pn 1, j (13.20)
r*
where r * e rt . Using this present value formulation, all of the option values in column n 1 may be
identified.
Before proceeding back another time increment, t , in the valuation, it is necessary to see if any of
the computed option values are below their early exercise proceeds at the respective nodes, X S n 1, j . If the
exercise proceeds are greater than the computed option value, the computed value is replaced with the early
exercise proceeds. If they are not, the value is left undisturbed. If this step is not performed, the procedure
will produce the value of a European put option.
Once the checks are performed, we go to column n 2 , repeat the steps and so on back through time.
Eventually, we will work our way back to time 0, and the current value of the American put option (in
column 0) will be identified.
To complete the binomial method illustration, suppose that the stock price lattice shown in Figure
13.4 underlies a 90-day American put option with an exercise price of 50. Applying the binomial method, the
value of the American put is $2.475. The value of this put using the European formula (13.16) is $2.364,
which means that the early exercise premium of the American put is worth about 11.10. Note that the value
produced by the binomial method for the European put (by not checking for the early exercise constraints) is
$2.355, which is different from the $2.364 obtained using the European formula. This is error due to the fact
that the binomial method is only an approximation. In general, the accuracy of the binomial method
increases with the number of time steps, holding other factors constant.
S 0x S 0 i 1 Di e rt i (13.21)
n
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 270/358
where Di (ti ) is the amount of (time to) the i th dividend paid during the option’s life and S0 is the current
stock price. Next, we set up the binomial lattice, beginning with S0x rather than S0 . That is, if the current
stock price net of dividends is S0x , the stock price at the end of the first time interval is either uS0x , or dS0x .
The values of u, d and p are computed as (13.18a), (13.18b), and (13.18c).
With the stock price lattice (net of dividends) computed, the approximation method starts at the end
of the option’s life and works back to the present. At the end of the option’s life, the option value at each
stock price node is given by the intrinsic value of the put (13.19), where S x replaces S . The option values
one step back in time (at time n 1 ) are computed by taking the present value of the expected future value of
the option (13.20). Before stepping back another time increment, it is again necessary to see if any of the
option values are below their early exercise value. Here is where dividends may enter the picture again. If a
dividend is paid at time n 1 , then the early exercise value is simply the ex-dividends are paid at time n 1 ,
however, the early exercise proceeds equal the exercise price less the lattice stock price less the dividend. If
any of the computed option values are below the exercise proceeds, they are replaced with the value of the
exercise proceeds.
As we repeat the process and step back further in time, we must keep track of the sum of the present
values of the dividends paid during the option’s remaining life. At time n 1 , there was only one dividend
and it was paid at time n 1 , so the sum equals the value of the single dividend paid at time n 1 . If we are
at time n 2 and there is a dividend paid at time n 2 as well as a dividend paid at time n 1 , the sum of
the present values of the promised dividends that should be included in the early exercise boundary check at
time n 2 is
Dn 1
PVDn 2 Dn 2 . (13.22)
r*
In other words, the early exercise boundary at time n 2 is X S nx 2, j PVDn 2 . By the time the iterative
procedure is complete, the early exercise boundary used to check the option price corresponding to the time
0 stock index level node will include the present value of all promised dividends as in equation (13.21).
EXAMPLE 13.2
Compute the value of an American-style put option that has an exercise price of $50 and a time to expiration
of 90 days. Assume the riskless rate of interest is 10 percent annually, the stock price is $50, the standard
deviation of the rate of return of the stock is 30 percent per year, and the stock pays a dividend of $2 in
exactly 60 days.
We proceed in two distinct steps. First we compute the European put option value using the formula
(13.17), and then we compute the American put option value using the binomial method. In this way, we can
identify the value of the early
so the European put value can be computed as p 50e 0.10 ( 90 / 365) N1 ( d 2 ) 48.033N1 (d1 )
where
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 271/358
and d 2 0.029 0.149 0.178 . The probabilities N (0.029) and N (0.178) are 0.512 and 0.571, so the
European put value is
The value of the American put is computed using the binomial method. The number of time steps is set equal
to 90, so the time increment t is one day or 0.00274 years. The values of the factors u and d are
u e 0.30 0.00274 1.01583 , and d 1 / 1.01583 0.98442 , with probabilities
p (e 0.10 ( 0.00274 ) 0.98442) /(1.01583 0.98442) 0.5048 and 1 p 0.4952 , respectively. The possible
values of the stock price (net of dividends) at the option’s expiration range from 11.69 to 197.38. The value
of the American put is 3.393, hence the early exercise premium on the American option is 3.393 - 3.262 =
0.131.
Table 13.5 demonstrates how the value of the early exercise premium increases as the put option goes deeper
in-the-money. At a stock price of $40, for example, the early exercise premium is about 350, about 3 percent
of the overall option value.
Finally, it is worthwhile to note that the dividend-adjusted binomial procedure outlined above not
only handles an American put option on a dividend-paying stock but also handles American call options.
Where the stock pays only a single dividend during the option’s life, the American call option valuation
equation (13.15b) is the most computationally efficient. Where the stock pays multiple dividends, however,
the dividend-adjusted binomial method is much faster. We address this issue again when we value the
American-style S&P 100 index options in the next chapter.
132
The approach used here is based on Smith (1976).
133
Executive stock options, a popular form of management incentive compensation, can be considered warrants from a valuation
standpoint.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 272/358
TABLE 13.5 Simulated American and European put option values on a stock with a known discrete
dividend. The put option has an exercise price of $50 and a time to expiration of 90 days. The riskless
rate of interest is 10 percent, and the standard deviation of the stock return is 30 percent. The stock
pays a dividend of $2 in 60 days.
Stock Price European Put Price American Put Price Early Exercise Premium
S p ( S , T ; X ) P ( S , T ; X ) P
40 10.875 11.230 0.355
45 6.510 6.757 0.247
50 3.264 3.393 0.129
55 1.360 1.406 0.046
60 0.476 0.492 0.016
Let S be the aggregate market value of the shares of the common stock currently outstanding; W ,
the aggregate market value of all warrants; r , the riskless rate of interest; and, V , the total market value of
the firm. The firm is assumed to have only two sources of financing-stock and warrants-so the total market
value of the firm may be defined as V S W . The number of shares of stock outstanding is nS , and nW is
the number of shares of stock sold if warrants are exercised. One warrant is assumed to provide the right to
buy one share. The dilution factor possible due to the presence of the warrants is nW /(nS nW ) . The
stock is assumed to pay no dividends during the warrant’s life. The standard deviation of the overall rate of
return on the firm is . Finally, the warrant contract parameters are T , the time to expiration of warrants,
and X , the aggregate amount paid by warrant holders to acquire the stock (i.e., the aggregate exercise price).
The assumptions used in the development of the warrant valuation equation are the same as those underlying
the European call option except it is assumed the total market value of the firm is lognorrnally distributed at
~
the warrants’ expiration, not the firm’s share price (i.e., ln(VT / V ) ) is normally distributed with variance
2 ).
Using risk-neutral valuation, the value of the firm’s warrants today is the present value of the
expected future terminal value, that is,
~
W e rT E (WT ) . (13.23)
The terminal value of the warrants, in turn, is the proportion of the firm that the warrant holders will own if
the warrants are exercised, , times the terminal value of the firm after the warrant holders pay the cash
~
exercise amount, VT X , less the cash exercise amount, X , that is,
~
~ (V X ) X if (VT X ) X
WT T which can be rewritten
0 if (VT X ) X
~
~ (VT ) (1 ) X if (VT ) ( 1 γ)X
WT (13.25)
0 if (VT ) ( 1 γ)X
Note the similarity between the structure of the warrants payoffs in (13.25) and the European call option
payoffs discussed in Chapter 11, that is,
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 273/358
~
~ ST X if ST X
cT (13.26)
0 if ST X
In Chapter 11, we showed that the expected terminal value of the European call option for a non-dividend-
paying common stock is
E (c~T ) e rT SN (d1 ) XN (d 2 ) (13.27)
ln(S / X ) (r 0.5 2 )T
where d1 and d 2 d1 T
T
where ST is lognormally distributed. In the warrant valuation problem, (VT ) , corresponds to the terminal
stock price ST and is assumed to be lognormally distributed. The term (1 ) X is certain and corresponds to
the exercise price of the stock option. It therefore follows that
~
E (WT ) e rT ( )VN (d1 ) (1 ) XN (d 2 ) , (13.28)
ln(( )V /(1 ) X ) (r 0.5 )T 2
where d1 and d 2 d1 T
T
Substituting (13.28) into (13.23), we find-that the aggregate market value of the warrants of the firm is
S SV , (13.30)
where SV is the elasticity of the stock price with respect to the value of the firm, that is, the percentage
change in stock price for a given percentage change in the value of the firm. To estimate SV first recall that
by assumption the firm’s value is the sum of the market value of the stock and the market value of warrants.
Thus,
134
This idea was first suggested by Schulz and Trautmann (1991).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 274/358
S V W
The change in the market value of the stock for a given change in the value of the firm is, therefore,
S W
1
V V
W
( ) N (d1 ) so
V
S
1 ( ) N (d1 ) . (13.31)
V
Finally, multiplying this term by V / S , we find that the elasticity of the stock price with respect to the value
of the firm is
S / S V
SV [1 ( ) N (d1 )] (13.32)
V / V S
Hence, to value the warrant as a function of the stock’s volatility rate, we simply substitute the following
term for the in (13.29):
1 S
S S (13.33)
SV [1 ( ) N (d1 )]V
Another somewhat perplexing consideration in applying (13.29) to value warrants is that the warrant
value appears on both sides of the equation, directly on the left-hand side and indirectly through V (i.e., W
is embedded in V ) on the right-hand side. This poses no great difficulty. We simply find the value of W that
satisfies the equation through some sort of numerical search procedure, just as we do when finding the yield
to maturity of a coupon-bearing bond.
EXAMPLE 13.3
Compute the value of a one-year warrant whose exercise price is $50. The current stock price is $50, and the
stock pays no dividends. The firm has only two sources of financing: 2,000 shares of stock and 500 warrants.
One warrant entitles its holder to one share of common stock. Assume that the riskless rate of interest is 6
percent, and that the standard deviation of the rate of return on the firm is 30 percent.
The dilution factor posed by the warrants is
500
0.2
2,000 5000
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 275/358
V 2,000 50 W 100,000 W .
The solution to this problem is obtained iteratively. The values of d1 and d 2 on the final iteration are
0.4611 and 0.1611, respectively. The probabilities are N (0.4611) 0.6776 and N (0.1611) 0.5640 . The
aggregate market value of the warrants is $3,389.46, so the price per warrant is $6.78.
In the interest of completeness, it is worthwhile to note that the volatility rate of the stock in this exercise
equals
The volatility rate of the stock is lower than the volatility rate of the overall firm so the volatility rate of the
warrants must be higher. Since the returns of the stock and the warrant are perfectly correlated,
S W
S W
V V
100,000 103,389.46
Therefore, the warrant volatility rate is 0.30 0.2661 1.3002
103,389.46 3,389.46
13.6 SUMMARY
The focus of this chapter is stock options. Following a description of exchange-traded stock options in the
first section, we adapt the general pricing principles of Chapters IO and II to value call and put options on
non-dividend-paying stocks. The principles are modified somewhat to account for the fact that common
stocks typically pay discrete dividends during the option’s life. We assume that the amount and the timing of
the dividends paid during the option’s life are known with certainty.
In this chapter, we also introduce the use of the binomial method to price American options.
Although the specific application in this chapter is American-style options on stocks, the binomial method
can be applied to the valuation of virtually any type of option. We use it again in Chapter 15, for examples to
value interest rate options.
Finally, a special type of call option on common stock is considered. Specifically, firms often issue
fights or warrants to raise new capital. Like call options, these contracts provide the holder with the fight to
buy the common shares of the firm at a fixed price within a specified period of time. Unlike call options,
however, the firm sells (or gives away) the options, so, if the fights/warrants are exercised, the firm faces the
prospect of having the equity of the firm diluted. The prospect of dilution has an important effect on warrant
price.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 276/358
In Appendix 11.2, a cumulative univariate normal density function approximation was provided to help
compute the value of options on commodities with a constant, proportional cost-of-carry rate. In this chapter,
we found that if a common stock pays a discrete dividend during the option’s life, the American call option
valuation equation requires the evaluation of a cumulative bivariate normal density function.
While there are many available approximations for the cumulative bivariate normal distribution, the
approximation provided here relies on Gaussian quadratures.135 The approach is straightforward and
efficient, and its maximum absolute error is 0.00000055.
The probability that x is less than a and that y is less than b for a standardized bivariate normal
distribution is
1 a b x 2 2 xy y 2
Pr( x a and y b)
2 1 2
exp[
2(1 2 )
]dxdy N 2 (a, b; )
where
f ( xi , x j ) exp[a1 (2 xi a1 ) b1 (2 x j b1 ) 2( )( xi a1 )( x j b1 )] (2)
i, j w x
1 0.24840615 0.10024215
2 0.39233107 0.48281397
3 0.21141819 1.0609498
4 0.0033246660 1.7797294
5 0.00082485334 2.6697604
135
The Gaussian quadrature method for approximating the bivariate normal is from Drezner (1978), and the Gaussian quadratures
for the integral are from Steen, Byme, and Gelbard (1969). For a contingency table approach to this problem, see Wang (1987).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 277/358
where the values of N 2 () on the right-hand side are computed from the rules for ab 0 ,
a b N 2 (a, b; )
-1.00 -1.00 -0.50 0.003782
-1.00 1.00 -0.50 0.096141
1.00 -1.00 -0.50 0.096141
1.00 1.00 -0.50 0.686472
-1.00 -1.00 0.50 0.062514
-1.00 1.00 0.50 0.154873
1.00 -1.00 0.50 0.154873
1.00 1.00 0.50 0.745203
0.00 0.00 0.00 0.250000
0.00 0.00 -0.50 0.166667
0.00 0.00 0.50 0.333333
EXAMPLE 13.A
Compute the risk-neutral probability that IBM and GM will have stock prices above $120 and $60,
respectively, at the end of two months. The current price of IBM is $107 and the current price of GM is $48.
Assume the riskless rate of interest is 10 percent annually, IBM and GM returns are bivariate normally
distributed, IBM has a return volatility of 33 percent annually, GM has a return volatility of 36 percent
annually, and the correlation between the returns of the two stocks is 0.6. Neither stock pays a dividend
during the next two months.
The first step in finding this probability is to compute the upper integral limits for the standardized
normal bivariate density function. For IBM, the computation is
[ln(107 /120) (0.10 0.5(0.33) 2 )(2 /12)]
a 0.7948 and, for GM, the computation is
(0.33) (2 /12)
[ln(48 / 60) (0.10 0.5(0.36) 2 )(2 /12)]
b 1.4784
(0.36) (2 /12)
The next step is to apply an approximation method to compute the bivariate normal probability. Applying the
procedure described above, the probability is
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 278/358
for IBM and GM, respectively. Thus, if the returns of IBM and GM were independent (i.e., their return
correlation is 0), the probability that in two months IBM will have a stock price above 120 and that GM will
have a stock price above 60 is 0.2134 x 0.0697 or about 1.49 percent. The reason that this probability is less
than the 4.6 percent where the correlation is 0.6 is that, with high positive return correlation, an upward
movement in IBM’s stock price implies that GM’s stock price a will tend to move upward also. In the
extreme case where these two stocks have a perfect positive correlation (i.e., 1 ), the probability that in
two months IBM will have a stock price above 120 and that GM will have a stock price above 60 is the
lower of the two univariate probabilities, 6.97 percent.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 279/358
14.6 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 280/358
Tables 14.1 and 14.2 contain Wall Street Journal listings of the currently active index option and index
futures option contracts, respectively. - The index options in Table 14.1 are all cash settlement contracts.
When the option expires, usually at the close of trading on the third Friday of the contract month, the option
seller pays the option buyer an amount of cash equal to the difference between the closing index price and
the exercise price of the option, assuming the option is in-the-money. No delivery takes place. When the
index futures options in Table 14.2 expire, the option holder receives a position in the underlying futures
contract.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 281/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 282/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 283/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 284/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 285/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 286/358
The figures in Table 14.1 show that there is a wide disparity in the trading activity of different index option
contracts. The CBOE’s S&P 100 contract, for example, is by far the most active. The CBOE’s S&P 500
contract and the AMEX’s MMI contract are moderately active. The rest of the index option contracts are
fairly inactive. Given the substitutability of the broad-based index option contracts, it is not surprising that
the broad-based contracts introduced earliest have been the most successful. The lack of success of the
options on narrow indexes reflects the smaller market for these options.
The compositions of the stock portfolios that underlie each of the indexes with the most active option
contracts were discussed in Chapter 7, except for the S&P 100. The S&P 100 index has a fairly short history.
When the CBOE was considering an index option contract in the early 1980s, it decided on a value weighted
index of the one hundred largest stocks for which CBOE stock options existed. Originally, the index was
called the “CBOE 100.” Later, the CBOE reached an agreement to have Standard & Poors track the portfolio
composition, at which time the index was renamed the S&P 100.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 287/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 288/358
c( S , T ; X ) max[0, Se dT Xe rT ] (14.1a)
and
p( S , T ; X ) max[0, Xe rT Se dT ] (14.1b)
for the call and the put, respectively. The lower price bounds for the American call and put options on stock
indexes are
respectively. The put-call parity relation for European stock index options is
c( S , T ; X ) p( S , T ; X ) Se dT Xe rT (14.3)
and the put-call parity relation for American stock index options is
Se dT X C ( S , T ; X ) P ( S , T ; X ) S Xe rT . (14.4)
Note that in all of these relations, there is an implicit assumption that the dividend income received (or paid)
while the arbitrage portfolio is held is automatically reinvested in the stock index portfolio.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 289/358
for the call and the put, respectively. The lower price bounds for the American call and put options on stock
index futures are
C ( F , T ; X ) max[0, ( F X )] (14.6a)
and
P ( F , T ; X ) max[0, ( X F )] , (14.6b)
respectively. The put-call parity relation for European stock index futures options is
c( F , T ; X ) p( F , T ; X ) e rT ( F X ) , (14.7)
and the put-call parity relation for American stock index futures options is
Fe rT X C ( F , T ; X ) P ( F , T ; X ) F e rT X (14.8)
In all of these relations, there is an implicit assumption that the futures position is the rollover position
originally described in Chapter 3.
c( F , T ; X ) c( S , T ; X ) (14.9a)
and
p( S , T ; X ) p( F , T ; X ) (14.9b)
In this case, index options and index futures options are perfect substitutes for one another. If the options are
American, and if the dividend yield rate is below the riskless rate of interest, the price relations are
C ( F , T ; X ) C (S , T ; X ) (14.10a)
and
P( S , T ; X ) P( F , T ; X ) . (14.10b)
Violation of the conditions (14.9a) through (14.10b) implies that costless arbitrage profits may be earned.
It is also important to recognize that stock index options may be priced in relation to the futures contracts.
The put-call parity equations, (14.3) and (14.4), can be expressed in relation to the price of the index futures
contract (rather than bT the underlying index) by using the cost-of-carry relation, F Sebt , to substitute for
S . For European index options, (14.3) becomes
c( F , T ; X ) p( F , T ; X ) e rT ( F X ) , (14.11)
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 290/358
These relations are particularly important for index option market makers who use index futures as a way of
laying off the risk of index option portfolios they hold. It is considerably less expensive to buy and sell
futures contracts on the index than it is the index portfolio itself.
Likewise, index futures options may be priced in relation to the underlying index. For European index
futures options, the relation is
c( S , T ; X ) p( S , T ; X ) Se dT Xe rT (14.13)
and, for American index futures options, the relation is
Se dT X C ( S , T ; X ) P( S , T ; X ) Se( r d )T Xe rT . (14.14)
and the pricing equation for a European put option on a stock index is
c( S , T ; X ) A2 ( S / S ) 2 if S S
* q *
C (S , T ; X ) (14.17)
S X if S S *
where
136
This model is called the constant, proportional dividend yield model and first appeared in Merton (1973).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 291/358
S * is the critical index level above which the American call should be exercised immediately and is the
solution to
where c( S , T ; X ) is the European call option formula (14.15). For an American put option on a stock index,
the quadratic approximation is
p( S , T ; X ) A1 ( S / S ** ) q1 if S S **
P(S , T ; X ) (14.18)
X S if S S **
where
S ** is the critical index level below which the American put should be exercised immediately and is the
solution to
where p( S , T ; X ) is the European put option formula (14.16). All other notation is as defined for the
American call option.
EXAMPLE 14.1
Compute the price of a 90-day S&P 100 index put option with an exercise price of 350. The current S&P 100
index level is 355.00, the dividend yield on the index is 3.50 percent annually, and the volatility rate is 25
percent annually. The riskless rate of interest is 6 percent.
The S&P 100 index option is an American-style option, but, since the quadratic approximation
requires the value of the corresponding European option, we will compute it first. Applying
The values of N (d1 ) and N (d 2 ) are 0.4106 and 0.4594, respectively, so the European put option price is
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 292/358
Applying the quadratic approximation (14.18), we find that the critical index level, S ** , below which the
index put should be exercised immediately is 285.347. Since the index level is currently 355.00, we are not
in the early exercise region. The value of A, is 1.778; the value of q , is - 11.335. The approximate value of
the S&P 100 option is, therefore,
The value of the S&P 100 index put is 14.068, and the value of the right to early exercise is 0.15.
c( F , T ; X ) e rT [ FN (d1 ) XN (d 2 )] , (14.19)
and the value of a European put option on an index futures contract is
where
ln( F / X ) 0.5 2T
d1 , and d 2 d1 T
T
c( F , T ; X ) A2 ( F / F ) 2 if F F
* q *
C(F ,T ; X ) (14.21)
F X if F F *
F *{1 e dT N [d1 ( F * )]} ln( F / X ) 0.5 2T
where A2 , d1 ( F ) ,
q2 T
1 1 4k 2r
q2 , k 2
2 (1 e rT )
F * is the critical futures price above which the American call should be exercised immediately and is the
solution to
where c( F , T ; X ) is the European call option formula (14.19). For an American put option on a stock index
futures contract, the approximation is
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 293/358
p ( F , T ; X ) A1 ( F / F ** ) q1 if F F **
P( F , T ; X ) (14.22)
X F if F F **
where
F ** is the critical futures price below above which the American put should be exercised immediately and is
the solution to
where p ( F , T ; X ) is the European put option pricing formula (14.20). All other notation is as defined for the
American call option.
EXAMPLE 14.2
Compute the price of a 90-day call option on the S&P 500 futures contract. Assume the call has an exercise
price of 380 and the current futures price is 400.00. The riskless rate of interest is 6 percent annually and the
volatility rate of the futures contract is 25 percent annually.
The S&P 500 futures option contract is an American-style option, so we use the quadratic
approximation. As an intermediate step, however, we first compute the value of the corresponding European-
style futures option contract using equation (14.19).
Applying the quadratic approximation (14.21), we find that the critical index level, F ** , above which the call
should be exercised immediately is 491.458. Since the futures price is currently 400.00, we are not in the
immediate early exercise region. The value of A2 is 1.270, the value of q2 is 11.945. The approximate value
of the S&P 500 futures option is, therefore,
The value of the call option on the S&P 500 futures is 30.80. Hence the value of the right to early exercise is
about 12¢.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 294/358
Discrete Dividends
For many stock indexes, the assumption that dividends are paid at a constant, continuous rate is unrealistic.
To value S&P 100 index options precisely, for example, it is necessary to account for the fact that the
dividends on the index are discrete and seasonal in their nature. Finns tend to pay dividends in quarterly
cycles and generally use a December 31 fiscal year-end. For these reasons, monthly seasonality in index
dividend payments is expected.
Figure 14.1 summarizes the monthly dividend patterns of the S&P 100 index during the period March
1983 through December 1989.137 It shows that the most popular ex-dividend months are February, May,
August, and November. The average daily dividend payments are higher in these months, and the frequency
of zero-dividend days is lower. Of these four months, the average daily dividend during the month of
February is highest, 6.090, probably as a result of the fact that extra dividend payments are typically declared
in the last quarter of the fiscal year. Of the total number of February trading days in this sample, less than 32
percent have zero dividends. The least popular months are January, April, July, and October. During these
months, the average daily dividend is less than half of the average dividend across all days. In addition, the
frequency of zero-dividend days during these months exceeds 55 percent of the total trading days.
Figure 14.1 also shows that average dividends have generally increased over the seven-year sample
period. In August 1983, for example, the average daily payout was 4.450, and, in August 1989, the average
payout was 6.700. The growth in the payouts also occurred in non-peak months. In April 1983, the average
daily dividend was 0.430, and, in April 1989, the average dividend was 2.150.
Aside from the monthly seasonal pattern in S&P 100 dividends, there is also a daily seasonal pattern.
Figure 14.2 presents average dividends by day of the week. Monday has the largest average payment, 4.940.
In addition, of the total number of Mondays in the sample period, only 18 percent are zero-dividend days. At
the other extreme, Wednesday appears to be the least popular day to pay dividends. In more than 55 percent
of the total number of Wednesdays during the sample period, no dividends are paid. The average dividend
payment across all Wednesdays is 1.760. Moreover, Wednesday’s popularity as an ex-dividend day appears
to be declining over the sample period. Regarding the remaining days of the week, Friday is more popular
than Tuesday, which, in turn, is more popular than Thursday. For Tuesdays and Thursdays, there are more
zero-dividend days than non-zero-dividend days.
The above dividend descriptions are written in terms of average dividend amounts. The importance of
dividends in S&P 100 index option valuation has to do not only with the average level of dividends but also
the magnitudes of individual daily dividend payments. Many cash dividends on the S&P 100 index are quite
large. The largest cash dividend on the index during the March 11, 1983December 29, 1989, sample period
was 55.130 on Thursday, November 5, 1987.
On Friday, December 23, 1983, the cash dividend on the S&P 100 index was 46.830, and, on Thursday, May
7, 1987, it was 46.100. The holder of an S&P 100 index call, for example, may find it optimal to exercise her
option the day immediately before such a large dividend is to be paid, and the holder of an S&P 100 index
put may find it optimal to exercise his option just after.
137
This figure and the dividend information were obtained from Harvey and Whaley (1992b).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 295/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 296/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 297/358
mimics the payoffs of the put by dynamically rebalancing the composition of the portfolio between stocks
and short-term debt as the market goes up or down.
TABLE 14.4 Terminal value of stock portfolio with static portfolio insurance.
Terminal Value
Initial
Position ST X ST X
Value
Own Stock Portfolio S ST ST
Buy Index Put p X S T
These values are represented in Table 14.5. The current position is where the index level is 100. The
increments up (down) from 100 in the first column of the table are 5 percent of the preceding (succeeding)
index level. The second column is the value of the European put at the different index levels. Note that the
original price of the portfolio insurance is 3.79. Finally, the last column is the overall value of the insured
portfolio. Note that the overall portfolio value never goes below 96.08. This value is the present value of 100
over the six-month interval, that is, 96.08e0.08(0.5) 100 . Recall that the policy the manager purchased insured
a portfolio value of at least 100 at the end of six months.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 298/358
where w1 N (d1 ) is the number of units of the index and w2 N (d 2 ) is the number of units of T-bills.
Note that both w, and W2 Will change as the index level, S, and the time to expiration, T , change. In fact, in
theory, the weights must change as soon as we have an infinitesimal change in either variable.
TABLE 14.5 Insured stock portfolio value at alternative stock index levels, using static portfolio
insurance.138
Index Level Put Option Portfolio Value
S p Sp
59.87 36.21 96.0’8
63.02 33.06 96.08
66.34 29.75 96.09
69.83 26.29 96.13
73.51 22.70 96.21
77.38 19.03 96.41
81.45 15.38 96.83
85.74 11.87 97.61
90.25 8.67 98.92
95.00 5.94 100.94
To demonstrate the dynamic portfolio insurance scheme more clearly, consider Table 14.6. The first column
is the index level and the second column is the price of the short-term debt instrument. The third and fourth
columns are the portfolio weights w1 and w2 , and the final column is the portfolio value. Note that as -AV
the index level falls, funds are removed from stocks and placed in T-bills, and, as the index level rises, funds
are moved from T-bills and placed in stock. Note also that the portfolio values are identical to the values in
Table 14.5, showing that the two portfolio insurance schemes provide identical results in theory. The
problem with the dynamic approach is that, from a practical perspective, continuous rebalancing is not
possible.
Table 14.7 gives some perspective on what happens when the portfolio is not rebalanced. The
weights are established when the index level is 100 and are held at that level independent of the direction the
index moves. Note that when the index falls, we do not remove funds from stocks to place in T-bills, so the
portfolio value falls below the present value of the 100 we want to have on hand in six months. On the other
138
In the valuation of the European put, the index is assumed to pay no dividends, the riskless rate of interest is 8 percent, and the
standard deviation of the index return is 20 percent. The put option has an exercise price of 100 and a time to expiration of six
months.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 299/358
hand, if the index level rises, we do not profit by as much because we continue to hold a large proportion of
T-bills.
TABLE 14.6 Insured stock portfolio value at alternative stock index levels, using dynamic
portfolio insurance with no rebalancing with continuous rebalancing.139
Index Level T-Bill Price Stock Portfolio Weight T-Bill Weight
rT Portfolio Value
S Xe w1 w2
59.87 96.08 0.001 1.000 96.08
63.02 96.08 0.002 0.999 96.08
66.34 96.08 0.005 0.996 96.09
69.83 96.08 0.014 0.990 96.13
73.51 96.08 0.034 0.975 96.21
77.38 96.08 0.072 0.945 96.41
81.45 96-08 0.136 0.892 96.83
85.74 96.08 0.231 0.809 97.61
90.25 96.08 0.355 0.696 98.92
95.00 96-08 0.496 0.560 100.94
As a compromise between continuous rebalancing and no rebalancing, dynamic portfolio insurers usually
rebalance when the index level moves by a certain percent from the time the portfolio was last rebalanced. In
Table 14.8, we simulate this activity using a 5-percent trigger point. When the stock index is at a level of
100, we have the original portfolio weights seen in the previous tables. If the index moves down by 5
percent, for example, the new portfolio value will be 100.59 (i.e., 0.638 X 95.00 + 0.416 X 96.08), and the
portfolio will be rebalanced with the new weights, 0.495 and 0.558, for stocks and T-bills, respectively. Note
that this scheme does not completely insure downside protection. If the stock index level falls 15 percent, the
portfolio value falls below the 96.08 necessary to insure that 100 is on hand in six months. On the other side,
because we are slow to rebalance when the index level is going up, the upside potential is not as great as it
was in the continuous rebalancing case.
139
In the valuation of the European put, the index is assumed to pay no dividends, the riskless rate of interest is 8 percent, and the
standard deviation of the index return is 20 percent. The put option has an exercise price of 100 and a time to expiration of six
months.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 300/358
TABLE 14.7 Insured stock portfolio value at alternative stock index levels, using
dynamic portfolio insurance with no rebalancing.
Stock Portfolio T-Bill
Index Level T-Bill Price Weight Weight Portfolio Value
S Xe rT w1 w2
59.87 96-08 0.638 0.416 78.18
63.02 96.08 0.638 0.416 80.19
66.34 96.08 0.638 0.416 82.31
69.83 96.08 0.638 0.416 84.53
73.51 96.08 0.638 0.416 86.88
77.38 96.08 0.638 0.416 89.35
81.45 96-08 0.638 0.416 91.95
85.74 96.08 0.638 0.416 94.68
90.25 96-08 0.638 0.416 97.56
95.00 96.08 0.638 0.416 100.59
a. In the valuation of the European put, the index is assumed to pay no dividends, the riskless rate of interest
is 8 percent, and the standard deviation of the index return is 20 percent. The put option has an exercise price
of 100 and a time to expiration of six months.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 301/358
14.6 SUMMARY
This chapter focuses on stock index option contracts. Options on stock indexes and on stock index futures
are traded in the U.S. In sections 2 and 3, we discuss the pricing principles for these options, where the cash
index is assumed to pay dividends at a constant, continuous rate through time.
TABLE 14.8 Insured stock portfolio value at alternative stock index levels, using
dynamic portfolio insurance with discrete rebalancing.140
Stock Portfolio
Index Level T-Bill Price T-Bill Weight
rT
Weight Portfolio Value
S Xe w2
w1
59.87 96.08 0.001 0.985 94.65
63.02 96.08 0.002 0.984 94.66
66.34 96.08 0.005 0.982 94.67
69.83 96.08 0.014 0.976 94.72
73.51 96.08 0.034 0.961 94.85
77.38 96.08 0.071 0.933 95.12
81.45 96.08 0.135 0.882 95.67
85.74 96.08 0.229 0.802 96.65
90.25 96.08 0.353 0.691 98.24
95.00 96.08 0.495 0.558 100.59
Section 4 examines the effect of discrete cash dividend payments. The cash dividend payments on the S&P
100 index portfolio are shown to be discrete and seasonal. Some of the cash dividends are large enough to
induce early exercise. The wildcard option embedded in the S&P 100 option contract is also discussed.
The final section deals with portfolio insurance. Both static portfolio insurance using index put
options and dynamic portfolio insurance using portfolio rebalancing are discussed.
140
In the valuation of the European put, the index is assumed to pay no dividends, the riskless rate of interest is 8 percent, and the
standard deviation of the index return is 20 percent. The put option has an exercise price of 100 and a time to expiration of six
months. Rebalancing occurs if the index level moves by 5 percent.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 302/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 303/358
15.6 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 304/358
Table 15.1 contains a listing of interest rate instruments from the Wall Street Journal. Options on Treasury
bonds and notes are traded at the CBOE. These contracts expire on the Saturday after the third Friday of the
contract month, are American-style, and require the delivery of a specific Treasury bond or Treasury note.
Exchange-traded options on T-bonds and notes are not very active, however. In fact, no active contracts are
listed in Table 15. 1. The table reports only prices of relatively inactive CBOE options on short-term and
long-term bond indexes.
The most active interest rate options are those written on interest rate futures contracts. Of these, the
CBT’s Treasury bond futures option and the CME’s Eurodollar futures option contracts have the greatest
trading volume and open interest, as shown in Table 15.1. Upon exercising a T-bond futures option, a long or
short position in the nearby T-bond futures contract is assumed. These options are American-style and thus
may be exercised at any time up to and including the expiration day. The last day of trading is the Friday
preceding, by at least five business days, the first notice day for the corresponding T-bond futures contract.
In general, the first notice day of the futures is the first business day of the contract month.
The Eurodollar futures option is also American-style. The expiration day of the Eurodollar futures
option contract is the second London business day before the third Wednesday of the contract month, the
same as that of the underlying Euro-dollar future’s. Exercise of the Eurodollar option results in delivery of a
position in the Eurodollar futures contract of the same maturity. The Eurodollar futures contract, in turn,
fixes the price (or, equivalently, the yield) on a three-month Eurodollar deposit.
Table 15.2 on pages 370-371 shows the large number of Treasury issues outstanding on a given date.
Not all of these, nor even the majority of these, have exchange-traded options. If individuals want to buy or
sell options on particular bond issues, they usually go to OTC markets where bond option contracts can be
141
Recall that the assumption of a lognormal price distribution permits the price to rise without limit.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 305/358
tailored in any manner. It is commonplace to see both European- and American-style OTC bond options,
including ones with times to expiration of several years.
In Chapter 8, we discussed a number of conventions regarding T-bonds and T-bond price reporting.
For example, the decimal part of the reported price represents 32nds. Thus, the reported bid price (in Table
15.2) of 107:07 for the 81/2s of May 1997 is actually 1077/32. Two other conventions are that the face value
of a Treasury bond is $100,000 and the bond price is reported as a percentage of par. Thus, the bid price of
the 8/2s of May 1997 is actually 1077/32 X $1000 or $107,218.75. Finally, the reported bond price does not
include the accrued interest for the current coupon period. For the 81/2s of May 1997, this means that from
the
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 306/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 307/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 308/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 309/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 310/358
It is important to review these pricing conventions since similar conventions are used in bond option pricing.
For example, prices of T-bond options traded at the CBOE are also reported in 32nds and as a percentage of
par. Thus, a reported T-bond option price of 2-24 means that the cost of the option is 2 24/32 x $1000, or
$2,750. To provide a finer demarcation in option price, some OTC T-bond option bond futures dealers quote
prices in 64ths. The CBOT uses a similar practice for T bond futures options.142
The accrued interest convention also has an impact on T-bond option pricing mechanics. Assume, for
example, that a call option with an exercise price of 107 is written on the 81/2s of May 1997. If this option
were to be exercised on November 13, 1991, the bondholder would pay the exercise price and receive a T-
bond that she could immediately sell for $111,422.55. The exercise price, however, is not simply the stated
exercise price times $1,000. If the call is exercised, the option holder must not only pay the stated exercise
price but also the accrued interest as of the exercise date. Thus, the total exercise price on November 13,
1991, is (107 + 4.20380) x $1000, or $111,203.80.
The price of the Eurodollar futures option is expressed in decimal form, representing basis points.
Exercising the Eurodollar futures option implies that a futures position is assumed. The December 94.75 call
option contract implies that the option holder may buy a Eurodollar futures contract at an index level of
94.75.143 Each basis point of the price of the option is worth $25, so the price of the December is 12 x $25, or
$300. (See Table 15.1.) The $25 value comes from 94.75 call the value of 0.01 percent of $1 million for
three months (i.e., .0001 x $1,000,000 x 90/360 = $25).
142
Recall that the CBOT's T-bond futures contract is quoted in 32nds.
143
The translation of the index level to the yield on the $1,000,000 Eurodollar deposit is provided in Chapter 8.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 311/358
c( F , T ; X ) e rT E (cT ) (15.1)
The terminal value of the option is, in turn, a function of the Eurodollar futures index level, FT , that is,
F X if FT X
cT T (15.2)
0 if FT X
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 312/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 313/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 314/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 315/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 316/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 317/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 318/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 319/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 320/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 321/358
If we assume the terminal futures price is lognormally distributed, we would evaluate E (cT ) in the same
manner we did in Chapter 11, and substitute this into (15.1). The valuation equation would be (11.25) with
the cost-of-carry rate, b , set to zero because the underlying instrument is a futures contract.
The assumption of lognormally distributed yield requires a modification of the terminal value
function, (15.2), for the call. In Chapter 8, we discussed the fact that the Eurodollar futures price is an index
level computed by subtracting the yield on the Eurodollar deposit from 100. In other words, the futures price
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 322/358
is F 100 y . If we substitute this definition into (15.2) and rearrange, we find that the terminal call price
can be expressed as
But equation (15.3) looks surprisingly familiar. It is the terminal value function of a European put option,
where yT , has replaced ST and where ( 100 X ) has replaced X . Since yT is lognormally distributed, the
European put formula (11.28) of Chapter 11 can be applied directly. Using the fact that y (100 F ) , the
expected terminal call price is
where d1 y
, d 2 d1 y T
y T
and y is the standard deviation of the logarithm of the yield ratios, ln( yt / yt 1 ) . Substituting (15.4) into
(15.1), the price of a European call option on a Eurodollar futures contract144 is
By put-call parity for European futures options, the price of a European put option on a Eurodollar futures
contract is
EXAMPLE 15.1
Using the values reported in Tables 15.1 and 15.2, compute the implied yield volatility of the March 95.00
call option on the Eurodollar futures contract. According to the tables, the call price is .15, the underlying
futures price is 94.94, and the riskless rate of interest is about 4.6 percent. As of November 13, 1991t the
option has 124 days remaining to expiration.
Without showing the steps of the iterative search that is used to find the implied volatility, the solution is
y 10.26% .
Note that this volatility is upward biased since Eurodollar futures options are American style.
144
This approach to Eurodollar futures option valuation is described in detail in Emanuel (1985).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 323/358
y (100 F ) , (15.7)
dy dF (15.8)
dy dF F
(15.9)
y F 100 F
In other words, the rate of change in yield equals the rate of change in the index level scaled by the factor
F /(100 F ) . The yield volatility, y , therefore, equals the return volatility, F , times the factor
F /(100 F ) , that is,
F
y F . (15.10)
100 F
EXAMPLE 15.2
In Example 15.1, the implied yield volatility rate from the March 95 Eurodollar call is shown to be 10.26
percent. Compute the implied return volatility based on this estimate.
94.94
The implied futures price volatility is the solution to 0.1026 F which implies that
100 94.94
F 0.55% .
145
For short-term options of a year or less, this assumption is plausible, particularly if the underlying bond has a long time to
maturity. For long-term options and/or for short-term T-bonds and T-notes, this assumption is less tenable.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 324/358
accrued interest must be added to the reported bond price. We also noted that the exercise price of a bond
option is increased by the accrued interest. Since accrued interest is added to the reported bond price and to
the exercise price, one can simply ignore it and use the reported bond price and the stated exercise price in
the bond option pricing formula.
The option pricing formulas require the cost of carry for a bond, which is the short-term riskless rate
of interest less the coupon yield, that is, b r y . To compute the annualized coupon yield of a bond with
price B for use in the option pricing equations, recognize that a coupon payment, C , is received each half
year,
BC C
e y (0.5) or y 2 ln 1 (15.11)
B B
It is worth noting that bonds with high coupon yields tend to depreciate in price (since they initially sell
above par) and that bonds with low coupon yields tend to appreciate in price (since they initially sell below
par). The value of an option, in turn, depends on the price appreciation or depreciation. The higher (lower)
the rate of price appreciation on the bond, the higher (lower) the call price and the lower (higher) the put
price. We now apply the commodity option pricing results of Chapters 10 and 11 using this cost-of-carry
parameter.
EXAMPLE 15.3
Assume there exists a European-style call option on the 8½ s of May 1997 that we discussed earlier in the
chapter. The call has an exercise price of 107, a time to expiration of 100 days, and a current market price of
$12 1/32. Assume that the risk-less rate of interest is 4.6 percent. Compute the implied volatility of this
option based upon the average of the bid and ask bond prices.
The average of the bid and ask prices for the 8½s of May 1997 is 107 8/32 or 107.25. The coupon
yield on this bond is
4.25
y 2 ln 1 7.77%
107.25
Substituting into valuation equation (11.25), we get
1.656 107.25e(0.0460.07770.046)(100 / 365) N (d1 ) 107e 0.046(100 / 365) N (d 2 )]
ln(107.25 /107) [0.046 0.0777 0.5 y2 ](100 / 365)
where d1 , d 2 d1 y (100 / 365)
y (100 / 365)
Without showing the steps of the iterative search that is used to find the implied volatility, the solution is
8.97%
By far the most active long-term interest rate option market is the CBT’s Treasury bond futures options.
These options are American-style, and expire on the first Friday preceding, by at least five business days, the
first notice day for the corresponding T-bond futures contract. Also, the T-bond futures option has the
decimal part of its price reported in 64ths. Under the assumption that the futures price at the option’s
expiration is lognormally distributed, the valuation of these options is possible using the quadratic
approximation described in Chapter 14.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 325/358
EXAMPLE 15.4
Compare the theoretical price of a March 1992 T-bond futures put option, with a strike price of 100 to the
quoted price in Table 15.1, 2 17/64. The price of the March 1992 futures is 99 7/32, and, given that the
option expires on February 21, 1992, the time to expiration is 100 days. Assume the riskless rate of interest is
4.6 percent and the volatility rate is 9.00 percent.
The T-bond futures option contract is an American-style option, so we use the quadratic
approximation. We begin by computing the value of the corresponding European-style futures put contract
using equation (14.20).
To compare this theoretical value to the observed price, we need to transform the decimal price 2.33 to 64ths,
that is, 2 + (0.33 x 64)/64 = 2 21/64. In other words, the put option appears 4/64 underpriced.
1
and d (15.12b)
u
1 d
p (15.12c)
ud
and 1 p , respectively.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 326/358
Once the yield lattice is computed, it is necessary to compute the bond price at each node. Bond
valuation equations were presented in Chapter 8. Keep in mind that as you move forward through the lattice
computing bond prices at each yield node, the time to maturity of the bond decreases.
Figure 15.2 shows a binomial lattice with both yield and price computed at each node. The
underlying bond is assumed to be an 8-percent coupon bond with 20 years to maturity. It currently sells at
par, so the current yield to maturity, y0 , is 8 percent. The riskless rate of interest, r , is 6 percent, and the
yield volatility, y , is 50 percent. The time to expiration of the option, T , is 90 days, and the number of time
steps, n , is 90. The size of the time increment, t , is, therefore, one day or 0.00274 years.
FIGURE 15.1 Possible Paths that the Yield may follow under the
Binomial Model
Yield at the end of Interval:
0 1 2 3 4 n (even) n (odd)
u n y0 u n y0
u 4 y0
3
u y0
2 2
u y0 u y0
uy0 uy0 uy0
y0 y0 y0 y0
dy0 dy0 dy0
2 2
d y0 d y0
3
d y0
4
d y0
n n
d y0 d y0
0 if Bn , j X
Pn , j ( Bn , j ) (15.13)
X Bn , j if Bn , j X
The option values one step, t , back in time (in column n 1 ) are computed by taking the present value of
the expected future value of the option. At any point j in column n 1 , the yield can move up with
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 327/358
probability p or down with probability 1 p . The value of the option at time n if the yield moves up is Pn , j 1 ,
and if the yield moves down is Pn , j . The present value of the expected future value of the option is, therefore,
( p ) Pn , j 1 (1 p) Pn , j
Pn1, j (15.14)
r*
where r * e rt . Using this present value formulation, all of the option values in column n I may be
identified.
FIGURE 15.2 Possible paths that yield/bond price may follow under the binomial model where the
current price of the 8 percent, 20-year bond is 100, the riskless rate of interest is 6 percent annually,
and the yield volatility is 50 percent annually. The time to expiration of the option is 90 days, and the
number of time steps is also 90. The time increment, t , is, therefore, 1 day or 0.00274 years.
Yield at the end of Interval:
0 1 2 3 4 n (even) n (odd)
8.43/95.94 8.43/95.94
8.88/91.93
8.65/93.93
84.34/9.63 84.34/9.63
8.21/97.97 8.21/97.97 8.21/97.97
8.00/100.00 8.00/100.00 8.00/100.00 8.00/100.00
7.79/102.08 7.79/102.08 7.79/102.08
7.59/104.16 7.59/104.16
7.40/106.26
7.20/108.36
0.76/237.38 0.76/237.38
Before proceeding back another time increment, t , in the valuation, it is necessary to see if any of
the computed option values are below their early exercise proceeds at the respective nodes, X Bn1, j . If the
exercise proceeds are greater than the computed option value, the computed value is replaced with the early
exercise proceeds. If they are not, the value is left unchanged. Note that if this step is not performed, the
procedure will produce the value of a European put option.
Once the checks are performed, we go to column n 2 and repeat the steps, and so on back through
time. Eventually, we will work our way back to time 0, and the current value of the American put option (in
column 0) will be identified.
To complete the binomial method illustration, suppose that the bond price lattice shown in Figure
15.2 underlies a 90 -day American put option with an exercise price of 100. Applying the yield-based
binomial method, the value of the American put is $6.157. The value of the corresponding European-style
put option using the yield-based method is $5.713. The early exercise premium of the American put is,
therefore, worth about 44.40.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 328/358
dB / B
Dm (15.15)
dy
and convexity indicates how duration changes for a given change in yield,
1 d 2B 1
Convexity (15.16)
2 dy 2 B
The keys to these expressions are the first and second derivatives of bond price with respect to a change in
yield, that is, dB / dy and d 2 B / dy 2 , respectively. In this sense, duration and convexity are like the delta and
gamma of an option. In fact, we will now show that bond option deltas and gammas enable a fixed-income
portfolio manager to control the duration and convexity of the portfolio.
To understand how to tailor the duration and convexity exposure of a fixed-income portfolio, we need to
develop expressions for bond (or bond futures) option price changes as a function of yield changes. The first
derivative of option price with respect to a change in yield is
O O B B
(15.17)
y B y y
where is the delta value of the option. Thus, to change the dollar value exposure of a bond portfolio, we
simply combine the exposure in bonds, dB / dy , with the exposure in n bond options, dB / dy , that is,
dB
Desired dollar risk exposure = (1 n) . (15.18)
dy
To reduce the portfolio’s risk exposure to zero (for small changes in yields), the optimal number of options is
1
n
In the general case, one may use N different options to hedge a bond portfolio, that is,
Note that the dollar risk exposure is exactly zero only at the point where the derivative is taken.
EXAMPLE 15.5
Assume that a fixed income portfolio manager holds a 9-percent, 20-year bond whose current yield to
maturity is 8 percent. Its current market value is 109.82, its modified duration is 9.606, and its convexity is
70.450. Given the uncertainty about economic events, the manager decides to hedge the interest rate risk of
his portfolio by writing call options on this bond. The call has an exercise price of 110 and a time to
expiration of 3 months. The current price of this option is 2.900, its delta is 0.472, and its gamma is 0.0474.
Compute the number of call options to sell against this bond to immunize it from movements in the bond’s
yield.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 329/358
1 1
The optimal number of calls to mitigate the duration exposure is: n 2.119 .
0.472
If we sell this number of options against the underlying bond, the bond portfolio value for given changes in
the yield is shown in Figure 15.3.
Figure 15.3 demonstrates that the delta-neutral hedge reduces the range of possible portfolio values.
The unhedged portfolio ranges in value from 90 to 135 over the range of yields shown, while the hedged
portfolio ranges in value from 90 to 110. But, even with the delta-neutral hedge, the range of possible
portfolio values is large over this somewhat limited yield range. In addition, the hedged portfolio has reduced
value if the yield rises or falls. To improve upon this hedge, it is possible to use more than one option to
hedge both duration and convexity risk.
The zero-risk portfolio given by (15.19) is analogous to the delta-neutral portfolio discussed in Chapter 12.
In that chapter, we also noted that a change in delta brought about by a commodity price change introduces
gamma risk. A similar situation arises in the case of the hedged bond portfolio. Not only does the bond
option delta change as the yield changes, but so does the duration of the bond portfolio. To compensate for
these effects, we must consider the second partial derivative of the bond option price with respect to yield,
that is
2O Oy B y
O B
O d 2 B dB 2O
d 2 B dB OB
d 2 B dB B
y 2 y y y dy 2 dy By dy 2 dy y dy 2 dy y y
2
d 2B dB
2 (15.20)
dy dy
where / B is the gamma value of the option. Combining the convexity exposure of the bond with a
portfolio that consists of N bond options, we get dollar convexity exposure
Dollar convexity =
d 2B
dy 2
N
i 1
ni i
d 2B
2
dy
N
i 1
ni i
dB d 2 B
dy
dy 2
1
N
i 1
ni i f
N
n
i 1 i i (15.21)
/ .
2
where f dB
dy
d 2B
dy 2
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 330/358
EXAMPLE 15.6
Unsatisfied with the effectiveness of the hedge portfolio indicated in Example 15.5, the portfolio manager
decides to evaluate the effectiveness of a hedge of both the duration and the convexity risk of his position.
Aside from the 110 call described in the last exercise, a 105 call with three months to expiration is available.
The current price of the 105 call Is 5.720, its delta is 0.703, and its gamma is 0.0403. Compute the number of
calls to buy/sell against this bond to immunize it from movements in the bond’s yield.
To neutralize the duration and the convexity risk of the bond portfolio, we need to solve simultaneously the
following equations:
These can be obtained from the modified duration and the convexity figures reported for the bond. That is,
from equations (15.15) and (15.16), we know
dB
BDm = (109.82)(9.606) = 1, 054.93 ,
dy
d 2B
and 2 BConvexity = 2 109.82 70.450 = 15, 473.64 .
dy 2
The value of f is, therefore, 71.921. The deltas and gammas of the individual options are known, and the
remaining task is only computational. The optimal composition of the duration-convexity hedge is to sell
3.315 105 calls and to buy 2.819 110 calls.
The effectiveness of this hedge relative to the unhedged portfolio and the duration-hedged portfolio from
Example 15.5 can be seen in Figure 15.4. The range of outcomes has been further diminished. Using both
calls generates a curve that is much more horizontal at 110, ranging from 103 to 116. In addition, the hedged
portfolio rises if the yield falls and falls more slowly if the yield rises. Clearly, this second hedge is more
effective than the hedge discussed in Example 15.5.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 331/358
15.6 SUMMARY
This chapter focuses on the valuation of interest rate options. After reviewing the designs of the U.S.
exchange-traded interest option contracts, we discuss short-term interest rate option valuation. Using the
standard lognormal price distribution assumption is inappropriate for these options. In its place, we use the
assumption that yield is lognormally distributed and rederive the European option valuation equations. In the
third section, T-bond and T-bond futures option valuation under the lognormal price distribution assumption
is presented. The fourth section describes the same valuation but under the lognormal yield assumption. The
binomial method is also used. Finally, we show how T-bond option contracts can be used to control the
duration and the convexity of a fixed-income portfolio.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 332/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 333/358
16.4 SUMMARY
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 334/358
The last three chapters have focused on specific option contracts on stocks, stock indexes, and interest rate
instruments. Where the valuation procedures of Chapters 10 and 11 did not directly apply to these specific
option contracts, the procedures were modified. For example, we showed how the quadratic approximation
could be used to price American-style index options and how the binomial method could be used to value
options on assets with discrete cash flows during the option's life.
In this chapter, we discuss options on currencies and on physical commodities. Since no new valuation
procedures are needed for these contracts, sections 1 and 2 review only the nature of the exchange-traded
options on currencies and physical commodities, and refer the reader to the appropriate valuation equations
from the previous chapters. The third section discusses some exotic OTC options that are currently traded.
Examples of these are options on options, options on the maximum and the minimum of two risky
commodities, lookback options and barrier options. For these four types of options, we present valuation
equations. Other types of exotic options include options on the average price of a commodity, deferred-start
options, deferred-payment American options, and all-or-nothing options. For these instruments, we describe
only the essence of the option contracts. The chapter concludes with a summary.
Option contracts on spot currencies have been actively traded on the Philadelphia Exchange since late 1982.
Both European-style and American-style option contracts are traded, with the American-style contracts
having the greatest trading volume and open interest. The most active contracts are for British pounds,
German marks, Japanese yen, and Swiss francs, although options on other currencies also trade. Table 16.1
contains a listing of the currently active, exchange-traded currency option contracts. The spot currency must
be delivered upon exercise of these options. The current spot prices are also reported in Table 16.1. Currency
options expire on the Saturday before the third Wednesday of the contract month.
146
For an approach to foreign currency option valuation that permits interest rates to be stochastic, see Grabbe (1983).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 335/358
TABLE 16.1 Foreign currency exchange rates and foreign currency options.
CURRENCY TRADING
EXCHANGE RATES
Wednesday, November 13, 1991. The New York foreign exchange selling rates below apply
to trading among banks In amounts of $1 million and more, as quoted at 3 p.m. Eastern time
by Bankers Trust Co. and other sources. Retail transactions provide fewer units of foreign
currency Per-dollar.
U.S. $ equiv. Currency per U.S. $
Country Wed. Tues. Wed. Tues.
Argentina (Austral) 0.0001008 0.0001008 9918.67 9918.67
Australia (Dollar) 0.7860 0.7870 1.2723 1.2706
Austria (Schilling) 0.08681 0.08681 11.52 11.52
Bahrain (Dinar) 2.6539 2.6539 0.3768 0.3768
Belgium (Franc) 0.02966 0.02966 33.72 33.72
Brazil (Cruzeiro) 0.00144 0.00146 694.71 695.60
Britain (Pound) 1.7730 1.7725 0.5640 0.5642
30-Day Forward 1.7648 1.7640 0.5666 0.5669
90-Day Forward 1.7504 1.74% 0.5713 0.5716
180-Day Forward 1.7299 1.7291 0.5781 0.5783
Canada (Dollar) 0.8842 0.8838 1.1310 1.1315
30-Day Forward 0.8815 0.8814 1.1344 1.1346
90-Day Forward 0.8784 0.8779 1.1384 1.1391
180-Day Forward 0.8737 0.8733 1.1445 1.1451
Chile (Peso) 0.002844 0.002780 351.56 359.65
China (Renminbi) 0.185642 0.185642 5.3867 5.3867
Colombia (Peso) 0.001753 0.001753 570.38 570.38
Denmark (Krone) 0.1573 0.1573 6.3570 6.3555
Ecuador (Sucre) Floating rate 0.000966 0.000966 1035.00 10335.00
Finland (Markka) 0.24984 0.24941 4.0025 4.0095
France- (Franc) 0.17881 0.17879 5.5925 5.5930
30-Day Forward 0.17813 0.17808 5.6140 5.6156
90-Day Forward 0.17690 0.17695 5.6529 5.6545
180-Day Forward 0.17510 0.17504 5.7110 5.7130
Germany (Mark) 0.6112 0.6111 1.6362 1.6365
30-Day Forward 0.6090 0.6088 1.6421 l.6426
90-Day Forward 0.6045 0.6044 1.6543 1.6544
180-Day Forward 0.5982 0.5902 1.6717 1.6718
Greece (Drachma) 0.005405 0.005405 185.00 185.00
Hong Kong (Dollar) 0.12884 0.12884 7.7615 7.7615
India (Rupee) 0.03890 0.03890 25.77 25.77
Indonesia (Rupiah) 0.0005056 0.0005056 1978.00 1978.00
Ireland (Punt) 1.6330 1.6318 0.6124 0.6128
Israel (Shekel) 0.4308 0.4321 2.3215 2.3142
Italy (Lira) 0.0006121 0.0006117 1231.41 1232.01
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 336/358
Special Drawing Rights (SDR) are based on exchange rates for the U.S., German, British, French and
Japanese currencies. Source: International Monetary Fund.
European Currency Unit (ECU) is based on a basket of community currencies. Source: European
Community Commission.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 337/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 338/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 339/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 340/358
FINEX-Financial Instrument Exchange, a division of the New York Cotton Exchange. IMM-International
Monetary Market at the Chicago Mercantile Exchange. MCE MidAmerica Commodity Exchange.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 341/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 342/358
Source: Reprinted by permission of Wall Street Journal, © (November 14, 1991) Dow Jones & Company,
Inc. All Rights Reserved Worldwide.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 343/358
options. For example, Table 16.2 indicates it would have been possible to buy call options on the D-mark
futures with an exercise price of 0.61 at a cost of 1.30 cents per mark.
Since each contract at the Chicago Mercantile Exchange is DM 1 25,000, four contracts are
necessary, and the total premium is $6,500. This option position would cap the total dollar cost at
(0.61)(500,000) = $305,000, while retaining the possibility of gain if the D-mark should depreciate.
Increases in the D-mark above 0.61 would be offset by profits on the futures option position.
Options can provide a useful hedge if there is uncertainty about the underlying import or export
contract. For example, consider a U.S. company that bids a price of 350,000 pounds to install a computer
system in Great Britain and suppose the British company has a month in which to accept or reject the bid.
The U.S. company is concerned about a depreciation of the British pound, but if it sells futures to hedge the
foreign exchange risk, and the bid is not accepted, the company is left with an open currency futures position
that may have to be liquidated at a loss. An alternative hedge is for the U.S. company to buy put contracts on
350,000 British pounds. By purchasing puts, the company guarantees the price at which pounds can be sold
if the bid is accepted. If the bid is rejected, the put option is not exercised and is sold. In effect, the U.S.
company is using an option to hedge a contract that has an option feature. The U.S. company has given the
British company the put option to sell 350,000 pounds to the U.S. company in return for the computer
system. The U.S. company hedges that risk by buying a put.147
In addition to hedging import and export contracts, currency options are useful in international
investment and portfolio management. Investment in a fore-_ii country exposes a portfolio to exchange rate
risk as well as the usual risk of capital losses. Currency options can be used to modify that risk. For example,
an investor in Australian bonds could hedge principal and/or interest payments by purchasing puts on the
Australian dollar. Over-the-counter options written by banks are frequently used to tailor such hedges to the
needs of the investor, particularly when longer maturities are necessary and/or when a sequence of options is
required (as when a stream of coupon payments is hedged). Some fixed-income securities are offered with
imbedded currency options. For example, a bond might offer to pay interest and/or principal in either of two
currencies at a fixed exchange rate, with the investor having the option to choose the currency. Complex or
exotic options, which are discussed below, are often created to deal with currency risk. For example, a bond
could offer to pay principal and interest in dollars or in two other currencies at fixed exchange rates
established in the bond indenture. The holder of the bond thus has a dollar bond plus the option of choosing
the most valuable of the three currencies in which payment may be received.
147
The hedging uses of currency options in the kind of situation described here are also discussed in Giddy (1983) and in Feiger
and Jacquillat (1979).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 344/358
options are complex and are often tailored to the needs of the customer, they are available primarily in the
OTC market.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 345/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 346/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 347/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 348/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 349/358
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 350/358
In this section, we apply the lognormal price distribution mechanics used in Chapters 11 and 12 to price
compound options (or, more commonly, “options on options”), options on the maximum or the minimum,
lookback options, and barrier options. Illustrations are provided. Following these discussions, we describe
some other types of exotic options that currently trade. Our list of exotic options is necessarily incomplete,
since new option contracts are designed and traded almost every day. The descriptions included will give a
flavor for the ingenuity of some current option contract designs.
Options on Options
Compound options or options on options fall in the category of exotic options. Call (put) options providing
the right to buy (sell) call options (i. e., calls on calls or puts on calls) and call (put) options providing the
right to buy (sell) put options (i.e., calls on puts or puts on puts) are the most common forms. To value these
options on options,148 we adopt the assumptions and notation used in Chapter 11. The critical assumptions
are that the terminal commodity price distribution is lognormal and that the principles of risk-neutral
valuation apply. The call and put options that we are valuing are assumed to be European-style with exercise
prices, ct* and pt* , respectively, and with time to expiration t . The notation representing the right to buy a
call at time t (i.e., a call on a call) is c(c, t ; ct* ) , the right to sell a call at time t (i.e., a put on a call) is
p(c, t ; ct* ) , the right to buy a put at time t (i.e., a call on a put) is c(c, t ; pt* ) , and the right to sell a put at time
t (i.e., a put on a put) is p(c, t ; pt* ) . The option received or delivered at expiration from the exercise of an
option on an option has exercise price X and time to expiration T . The notation used to describe the
underlying options is c( St , t ; X ) and p( St , t ; X ) , respectively. Conditional upon knowing St , these
European-style options can be valued using equations (11.25) and (11.28) from Chapter 11.
To demonstrate how to value a compound option, we use a call on a call. The first step in the risk-
neutral valuation approach is to formulate the option's payoff contingencies. For the call on a call, the payoff
contingencies at time t are
That is, if the value of the call to be received at time t , c( St , T ; X ) , is greater than the exercise price, ct* , the
call option holder will exercise his right to buy the call. If the value is less, he will let it expire worthless.
The second step involves restating the contingent payoffs in (16.1) in terms of the underlying commodity
price at time t , St , in order to make the problem more tractable mathematically. The commodity price above
which the call option holder will choose to exercise his call at time t is given by
c( S t* , T ; X ) ct* (16.2)
where c( St , T ; X ) represents the European-style option valuation equation (11.25) evaluated at St S t* . Note
that the value of St* may be solved iteratively in the same manner that we have computed critical commodity
prices in earlier chapters.149 With St* known, the payoff contingencies expressed in (16.1) may be written as
148
The models presented in this section are based on the work of Geske (1979).
149
See, for example, the valuation of American-style call options on dividend-paying stocks in Chapter 13 or the valuation of
American-style options using the quadratic approximation method in Chapter 14.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 351/358
Call on Call. Under risk-neutral valuation, the value of a call on a call may be written as the present value of
the expected terminal value of the option, where the discount rate is the riskless rate of interest, r :
Expressing c( St , T ; X ) in terms of its terminal commodity price payoffs and isolating the cost of exercising
the option at time t , equation (16.4) becomes
Under the assumption that future commodity prices are lognormally distributed, the value of a European-
style call on a call is
c(ct , t ; ct* ) Se( br )( t T ) N 2 a1 , b1 ; t /(t T ) Xe r ( t T ) N 2 a2 , b2 ; t /(t T ) e rt ct* N1 b2 (16.6)
where
ln( S / X ) (b 0.5 2 )(t T )
a1 , a2 a1 (t T )
(t T )
ln( S / St* ) (b 0.5 2 )t
b1 , b2 b1 t
t
and N1 () and N 2 () , are the cumulative univariate and bivariate unit normal density functions described in
Chapters 11 and 13, respectively.
In equation (16.6), the term N 2 a1 , b1 ; t /(t T ) is the delta value of the call option on a call option. It
describes the call option price movement for a small change in the commodity price. Recall that in Chapter
12 we showed how delta values are used for hedging purposes. The term N1 (b2 ) is the probability that the
commodity price will exceed the critical commodity price at time t . The term N 2 a2 , b2 ; t /(t T ) is the
probability that the commodity price will exceed St* at time t and the exercise price X at time t T .
Put on Call. The simplest way to derive the valuation equation for a put on a call is to deduce the valuation
formula from known results. In Chapter 12, we showed that a long-call/short-commodity position is
tantamount to a long-put position. Here, the underlying commodity position is a call option, so a long-call-on
a-call/short-call position should be tantamount to a put on a call. Since we have the valuation equation for a
call on a call (16.6) and for a European-call (11.25), the valuation equation for a put on a call is
p (ct , t ; ct* ) Se( br )(t T ) N 2 a1 , b1 ; t /(t T ) Xe r ( t T ) N 2 a2 , b2 ; t /(t T ) e rt ct* N1 b2
,
Se( br )( t T ) N1 a1 Xe r ( t T ) N1 a2 e rt ct*
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 352/358
p (ct , t ; ct* ) Xe r ( t T ) N 2 a2 , b2 ; t /(t T ) (16.7)
Se( br )( t T ) N 2 a1 , b1 ; t /(t T ) e rt ct* N1 b2
Call on Put. The risk-neutral valuation framework shown above can also be applied to value a call on a put.
The value of a European-style call on a put is
c(ct , t ; pt* ) Xe r (t T ) N 2 a2 , b2 ; t /(t T ) (16.8)
Se(br )( t T ) N 2 a1 , b1 ; t /(t T ) e rt pt* N1 b2
The critical commodity price below which the call option holder will choose to exercise the call to buy the
put at time t is determined by solving
p ( St , T ; X ) represents the European-style option valuation equation (11.28) evaluated at S St* . All other
notation is as previously defined. The term N 2 a1 , b1 ; t /(t T ) is the delta value of a call option on a
put option delta value. The term N 2 (b2 ) is the probability that the commodity price will be below the critical
commodity price at time t . The term N 2 a2 , b2 ; t /(t T ) is the probability that the commodity price
will be below St* at time t and the exercise price X at time t T .
Put on Put. A put on a put has the same payoff contingencies as a long-call on-a-put/short-put position.
Using equations (11.28) and (16.8), it can be shown that the value of a put on a put is
p (ct , t ; pt* ) Se( br )( t T ) N 2 a1 , b1 ; t /(t T ) , (16.10)
Xe r ( t T ) N 2 a2 , b2 ; t /(t T ) e rt pt* N1 b2
EXAMPLE 16.1
Consider a call option that provides its holder with the right to buy a put option on the S&P 500 index
portfolio. The put that would be delivered against the call if the call is exercised has an exercise price of
$400 and a time to expiration of six months. The call has an exercise price of $10 and a time to expiration of
three months. The S&P 500 index is currently at 390, pays dividends at a constant rate of 4 percent annually,
and has a volatility rate of 28 percent. The riskless rate of interest is 7 percent.
The first step in valuating the compound option is to compute the critical commodity price below
which the call will be exercised to take delivery of the put. This is done by solving
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 353/358
The critical commodity price, St* , is 497.814. The next step is to apply the valuation formula (16.8). Here,
we get
c(ct , t ; pt* ) 400e0.07(0.250.5) N 2 a2 , b2 ; 0.25 / 0.75
390e0.03(0.250.5) N 2 a1 , b1 ; 0.25 / 0.75 e0.07(0.25)10 N1 b2 27.722
where
ln(390 / 400) (0.03 0.5(0.28)2 )(0.75)
a1 0.1096, a2 0.1096 0.28 0.75 0.1329
0.28 0.75
ln(390 / 497.814) (0.03 0.5(0.28) 2 )0.25
b1 1.620, b2 1.620 0.28 0.25 1.760
0.28 0.25
The probability that the commodity price will be below the critical commodity price at time t , N1 (b2 ) , is
0.961. The probability that the commodity price will be below St* at time t and below the exercise price, X ,
at time t T , N 2 a2 , b2 ; t /(t T ) , is 0.453. The value of a call on a call with the same terms as the put is
27.012. (The critical index price is 342.424.)
cmax ( S1 , S2 ; X ) e rT E ( S1,T | S1,T X and S1,T S2,T ) Pr( S1,T X and S1,T S2,T )
e rT E ( S2,T | S2,T X and S2,T S1,T ) Pr( S2,T X and S2,T S1,T )
(16.11)
Xe rT Pr( S1,T X or S2,T X )
Under the assumption that future commodity prices are lognormally distributed, the value of a European-
style call on the maximum is
cmax ( S1 , S 2 ; X ) S1e( b1 r )T N 2 ( d11 , d1' ; 1' ) S 2 e( b2 r )T N 2 ( d12 , d 2' ; 2' ) Xe rT [1 N 2 ( d 21 , d 22 ; 12 )] (16.12)
where
ln( S1 / X 1 ) (b1 0.5 12 )T
d11 , d 21 d11 1 T
1 T
ln( S 2 / X 2 ) (b2 0.5 22 )T
d12 , d 22 d12 2 T
2 T
ln( S1 / S2 ) (b1 b2 0.5 2 )T '
d1' , d 2 (d1' T )
T
150
Other names for the option on the maximum are “the better of two assets” or “outperformance options.” The models presented
here are on the maximum or the minimum of two risky commodities and the valuation models are based on Stulz (1982). To
generalize these models to three or more risky assets, see Johnson (1987).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 354/358
1 12 2 2 12 1
2 12 22 2 12 1 2 , 1' and 2'
In equation (16.12), the term [1 N 2 (d 21 , d 22 ; 12 )] is the probability that one of the two commodity prices
will exceed the exercise price at time T or, alternatively, one minus the probability that neither commodity
will have a price greater than the exercise price at the option's expiration.
EXAMPLE 16.2
Consider a call option that provides its holder the right to buy $100,000 worth of the S&P 500 index
portfolio at an exercise price of $400 or $100,000 worth of a particular T-bond at an exercise price of $ 1 00,
whichever is worth more at the end of three months. The S&P 500 index is currently priced at $360, pays
dividends at a rate of 4 percent annually, and has a return volatility of 28 percent. The T-bond is currently
priced at $98, pays a coupon yield of 10 percent, and has a return volatility of 15 percent. The correlation
between the rates of return of the S&P 500 and the T-bond is 0.5. The riskless rate of interest is 7 percent.
Before applying the option on the maximum formula, it is important to recognize that there are two
exercise prices in this problem: $400 for the S&P index portfolio and $ 1 00 for the T-bond. What this
implies is that we can buy $ 1 00,000/ $400 = 250 “units” of the index portfolio or $100,000/$100 = 1,000 T-
bond “units” at the end of three months, depending on which is worth more. At this juncture, we must decide
whether to work with the valuation equation (16.12) in units of the S&P 500 index portfolio, in which case
we multiply the current T-bond price and its exercise price by 4, and then multiply the computed option price
by 250, or to work with the valuation equation (16.12) in units of the T-bond, in which case we divide the
current S&P 500 price and the option's S&P 500 exercise price by 4, and then multiply the computed option
price by 1,000.151 In this exercise, we choose to work in units of the S&P 500 index portfolio, so we adjust
the T-bond prices: the current T-bond price is assumed to be 392, and the T-bond exercise price is 400. With
the units of the two underlying assets comparable, we now apply equation (16.12):
cmax 360e0.04(0.25) N 2 (d11 , d1' ; 1' ) 392e0.10(0.25) N 2 (d12 , d 2' ; 2' )
400e0.07(0.25) [1 N 2 (d 21 , d 22
'
; 12 )] 11.962
where
ln(360 / 400) (0.07 0.04 0.5(0.28) 2 )(0.25)
d11 0.6290,
(0.28) 0.25
d 21 0.6290 (0.28) 0.25 0.7690
ln(392 / 400) (0.07 0.10 0.5(0.15)2 )(0.25)
d12 0.3319,
(0.15) 0.25
d 22 0.3319 (0.15) 0.25 0.4069
ln(360 / 392) (0.06 0.5(0.2427) 2 )(0.25)
d1' 0.5175,
(0.2427) 0.25
d 2' 0.5175 (0.2427) 0.25 0.6388
0.282 0.152 2(0.5)(0.28)(0.15) 0.2427,
0.28 0.5(0.15) 0.15 0.5(0.28)
1' 0.8447, and 2' 0.0412
0.2427 0.2427
151
These types of adjustments can be made freely because the option price is linearly homogeneous in both the commodity price
and the exercise price. See Merton (1973).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 355/358
The computed option price is 11.962, which implies the value of the option contract is $11.962 x 250, or
$2,990.50. The probability that either or both components of the option are in-the-money at expiration is
1 N 2 (0.7690,0.4069;0.5) , or 42.72 percent.
Under the same assumptions, the value of a European-style call on the minimum is
Lookback Options
Aside from compound options and options on the maximum and the minimum, many other exotic options
trade in OTC markets. Some of the options are backward looking. A lookback call option provides its holder
with settlement proceeds equal to the difference between the highest commodity price during the life of the
option less the exercise price, and a lookback put option provides its holder with settlement proceeds equal to
the difference between the exercise price and the lowest commodity price during the life of the option. It
should come as no surprise, therefore, that these options are sometimes referred to as “no-regret options.”
In a sense, lookback options are like American-style options because the option holder is guaranteed
the most advantageous exercise price. Lookback call options can be valued analytically using the risk-neutral
valuation mechanics.152 The reason for this is that it never pays to exercise a lookback option prior to
expiration. Independent of how low the commodity price has been thus far during the option's life, there is
always some positive probability that it will fall further. For this reason, the option holder will always defer
early exercise in the hope of recognizing higher exercise proceeds in the future.
Under the assumptions of risk-neutral valuation and lognormally distributed future commodity prices, the
value of a lookback call may be written as style call option whose exercise price is the current minimum
value of the underlying commodity. This is the least the lookback call can be worth since the commodity
price may fall below X, thereby driving the “exercise price” down further.
b[T
2 ln( S / X )
]
cLB Se(br )T N1 (d1 ) Xe rT N1 (d 2 ) Se( br )T e 2
N1 ( d3 ) N1 ( d1 ) (16.14)
where X is the current minimum price of the commodity during the life of the option, 0.5 2 / b ,
ln( S / X ) (b 0.5 2 )T (b 0.5 2 )T
d1 , d 2 d1 T , and d3 . Note that the first two terms of the
T T
option are the value of a European style call option whose exercise price is the current minimum value of the
underlying commodity. This is the least the lookback call can be worth since the commodity price may fall
below X , thereby driving the “exercise price” down further.
152
The pricing equations provided here are based on the work of Goldman, Sosin, and Gatto (1979).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 356/358
EXAMPLE 16.3
Consider a lookback call option that provides its holder with the right to buy the S&P 500 index at any time
during the next three months. The S&P 500 index is currently at a level of 390, pays dividends at a constant
rate of 4 percent annually, and has a volatility rate of 28 percent. The riskless rate of interest is 7 percent.
The cost-of-carry rate is 0.07 - 0.04 = 0.03. The value of the lookback call is, therefore,
Note that the price of the lookback call is considerably higher than an at-the-money index call option. The
value of a European-style call (i.e., the sum of the first two ten-ns in the valuation equation) is only 22.941.
2 ln( S / X )
b[ T ]
pLB Xe rT N1 ( d 2 ) Se( br )T N1 ( d1 ) Se( br )T N1 ( d1 ) e 2
N1 ( d3 ) (16.15)
where all notation is as defined for the lookback call. Note that a standard European-style put option is the
lower bound for the price of the lookback put option. The third term is necessarily positive. Using the same
parameters as in Example 16.3, the value of a lookback put option is $43.468, with the underlying ordinary
European-style put being valued at $20.056.
Other backward-looking options are also traded. For example, average price or Asian options are
based on the average (either arithmetic or geometric) commodity price during the option's life. The average
commodity price may be used as the exercise price of the option, in which case the settlement value of the
call will be the terminal commodity price less the average price, or it may be used as the terminal commodity
price, in which case the settlement value will be the average price less the exercise price. Unfortunately,
most Asian options do not have closed-form valuation equations. Accurate pricing involves the use of
numerical methods.153
Barrier Options
Barrier options are options whose existence depends on the underlying commodity price. A down-and-out
call, for example, is a call that expires if the commodity price falls below a prespecified “out” barrier, H .154
At that time, the option buyer may receive a cash rebate, R . A down-and-in call is a call that comes into
existence if the commodity price falls below the “in” barrier at any time during the option's life. Note that if
153
There are a number of useful background readings for those interested in pricing Asian options. Among them are Boyle (1977)
and Boyle and Emanuel (1985).
154
The valuation equation for the down-and-out call option was first provided in Cox and Rubinstein (1985, Ch. 7). The valuation
equation presented here is a modified version of the formula presented in Rubinstein (1990).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 357/358
we buy a down-and-out call and a down-and-in call with the same barrier price, H , exercise price, X , and
time to expiration, T , the portfolio has the same payoff contingencies as a standard call option. For this
reason, we automatically know how to value a down-and-in call if we can value a down-and-out call.
Under the assumptions of risk-neutral valuation and lognormally distributed commodity prices, the
valuation equation for a down-and-out call option is
H is the barrier commodity price below which the call option life ends; R is the rebate, if any, received by
the option buyer should the option terminate,
b 0.5 2 ln( S / X )
; 2 (2r / 2 ); a1 (1 ) T , a2 a1 T
2
T
ln( H 2 / SX ) ln( H / S )
b1 (1 ) T ; b2 b1 T ; c1 T ; c2 c1 T
T T
The valuation equation for a down-and-in call is simply equation (11.25) less (16.16).
EXAMPLE 16.4
Consider a down-and-in call option that provides its holder with the right to buy the S&P 500 index at 380
any time during the next three months, should the index level fall below 375. The S&P 500 index is currently
at a level of 390, pays dividends at a constant rate of 4 percent annually, and has a volatility rate of 28
percent. The riskless rate of interest is 7 percent.
he cost-of-carry rate is 0.07 - 0.04 = 0.03. The value of the down-and-out call is
cDO 390e(0.030.07)(0.25) N1 (a1 ) 380e0.07(0.25) N1 (b2 ) 390e(0.030.07)(0.25) (375 / 390)2( 1) N1 (b1 )
380e0.07(0.25) (375 / 390)2 N1 (b2 ) 14.817
where
0.03 0.5(0.28)2
2
0.1173; (0.1173) 2 (2(0.07) /(0.28) 2 ) 1.3414;
(0.28)
ln(390 / 380)
a1 (1 0.1173)(0.28) 0.25 0.3091, a2 0.3091 (0.28) 0.25 0.1691
(0.28) 0.25
ln((375) 2 / 390 380)
b1 (1 0.1173)(0.28) 0.25 0.2512; b2 0.2512 (0.28) 0.25 0.3912;
(0.28) 0.25
ln( H / S )
c1 (1.3414)(0.28) 0.25 0.0924; and c2 0.0924 (0.28) 0.25 0.4680
(0.28) 0.25
The value of a standard European-style call option is 28.151, using equation (11.25). The value of the down-
and-in call is, therefore, 28.151 - 14.817 = 13.334.
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Derivados Avanzados enero de 2010 Página: 358/358
An up-and-out put and an up-and-in put can be valued in a similar manner. An up-and-out put is a put that
expires if the commodity price rises above the out barriers Its valuation equation is
An up-and-in put comes into existence when the commodity price rises above H . Its valuation equation is
simply (11.28) less (16.17).
16.4 SUMMARY
This chapter concludes the presentation of option valuation principles and applications. First, we discussed
currency and currency futures options. Contract specifications were provided, and we noted that the
valuation of these options is a straightforward application of the constant cost-of-can-y framework developed
in Chapters 10 and 11. The cost-of-carry rate for currency options is the domestic rate of interest less the
foreign rate of interest, and the cost-of-carry rate for currency
futures options is zero. We discussed, as well, the use of currency options in hedging the currency risk that
arises in international trade or investment. Second, we discussed physical commodity futures options. In
general, no options on physical commodities trade, only options on physical commodity futures. Hence, the
valuation principles for these options also follow straightforwardly from the constant cost-of-carry
framework of the earlier chapters. The cost-of-carry rate for physical commodity futures is zero.
The remainder of the chapter focuses on exotic options. These are not exchange-traded options but
are unusual options that trade in OTC markets. We show how options on options, options on the maximum
and the minimum of two commodities, lookback options, and barrier options may be valued within a
lognormal price distribution framework. But, these are only four of a myriad of option contract designs that
exist in the OTC markets. We discuss others; however, the list is certainly incomplete given the pace with
which these new contracts are introduced.
155
For a brief review of a range of exotic options, see Hudson (1991).
Francisco A. Delgado: Francisco_Delgado@Hotmail.Com Versión Junio 2010 Impresión January 19, 2010
Chapter 2: Mechanics of Future Markets
Section Description
2.0 Introduction
2.1 Background
2.2 Specifications of the Futures Contract
2.3 Convergence of Futures Price to Spot Price
2.4 Daily Settlement and Margins
2.5 Newspaper Quotes
2.6 Delivery
2.7 Types of Traders and Types of Orders
2.8 Regulation
2.9 Accounting and Tax
2.10 Forward Contracts vs. Futures Contracts
2.0 Introduction
This chapter covers:
what futures are,
how the futures markets operate,
the convergence of futures prices to the spot price of the underlying asset, and
the key differences between futures and forward contracts.
Futures and forward contracts are agreements to buy or sell an asset at a future time at a certain price.
2.1 Background
In March, an investor in NY calls a broker to buy 5,000 bushels of corn for delivery in July.
i. The broker passes these instructions to a trader on the floor of the CBOT.
ii. The broker requests a long position in one contract (each corn contract comprises exactly 5,000 bushels).
In March, another investor in Kansas might instruct a broker to sell 5,000 bushels of corn for July delivery. The broker
would pass instructions to short one contract to a trader on the floor of the CBOT.
The two floor traders meet, agree on a price to be paid for corn in July, and the deal is done; in this example:
The investor in New York who agreed to buy has a long futures position in one contract.
The investor in Kansas who agreed to sell has a short futures position in one contract.
The price agreed to is the current futures price for July corn, which like any price, is determined by the laws of
supply and demand.
Closing Out Positions
The vast majority of futures contracts do not lead to delivery.
i. Most traders close out (take opposite trades from the original one) their positions prior to the delivery.
ii. For example, the New York investor can close out the position by selling (i.e., shorting) one July corn futures,
while the Kansas investor can close out the position on by buying one July contract.
In these cases, the exchange uses a formula to adjust the price received according to the coupon and maturity date of
the bond delivered. This is discussed in Chapter 6 (Interest Rate Futures).
2. The Contract Size:
A. Specifies the amount of the asset that has to be delivered under one contract.
B. Is important since contracts that are too large or too small exclude many traders from hedging their exposures.
3. Delivery Arrangements
A. Although the vast majority of the futures contracts do not lead to delivery, delivery arrangements are important
in understanding the relationship between the futures price and the spot price of the asset (This is particularly
important for commodities where there may be significant transportation costs).
B. When alternative delivery locations are specified, the price received by the party with the short position is
sometimes adjusted according to the location chosen by that party.
4. Delivery Months
A. A futures contract is referenced by its delivery month.
B. Delivery months vary from contract to contract. For example:
1. currency futures have delivery months of March, June, September, and December;
2. corn futures traded on the CBOT have delivery months of March, May, July, September, and December.
Note: the exchange specifies the last day on which trading can take place, which is generally a few days before the last
day on which delivery can be made.
5. Price Quotes
A. Are given in ways that are convenient and easy to understand.
B. Examples:
1. Crude oil futures prices are quoted in dollars per barrel.
2. Bond and T–notes are quoted in dollars and 32nds of a dollar.
6. Daily Price Movement Limits (specified by the exchange)
Its purpose is to prevent large price movements from occurring because of speculative excesses.
A. If the price moves down by an amount equal to the daily price limit, the contract is said to be limit down (vice
versa for limit up).
B. Normally, trading ceases for the day once the contract is limit up or limit down.
7. Position Limits (specified by the exchange)
Its purpose is to prevent speculators from exercising undue influence on the market.
A. Position limits are the maximum number of contracts that a speculator may hold.
B. Bona fide hedgers are not affected by position limits.
Futures Price
Spot Price
Time
Conclusion: A futures price (above or below the spot price) will converge to the spot price of the underlying asset as
the delivery month of a futures contract draws near.
Further Details
A. Some brokers allow investors to earn interest on the balance in their margin accounts.
B. To satisfy the initial margin requirements (but not subsequent margin calls), an investor can sometimes deposit
securities with the broker.
1. T–bills are used in lieu of cash (at about 90% of their face value).
2. Shares are also sometimes accepted in lieu of cash (at about 50% of their face value).
Additional Notes:
Minimum levels for initial and maintenance margins are set by the exchange.
A bona fide hedger (a company that produces the commodity on which the futures contract is written) is
often subject to lower margin requirements than a speculator.
A day trade is one in which the position will be closed out in the same day.
A spread transaction is one where the trader simultaneously takes a long position in a contract with one
delivery month and a short position in a contract on the same underlying asset with another delivery month.
Margin requirements are the same on short futures positions as they are on long futures positions.
Clearinghouse and Clearing Margins
The exchange clearinghouse:
acts as an intermediary or middleman in futures transactions.
guarantees the performance of the parties to each transaction.
has as its main task to track all the transactions that take place during a day so that it can calculate the net position
of each of its members.
A. A clearinghouse member is required to maintain a margin (known as a clearing margin) account with the
clearinghouse.
B. Every day, the account balance for each contract must equal the original margin times the number of contracts
outstanding.
C. The clearinghouse calculates the number of contracts outstanding on either a gross or a net basis.
1. The gross basis adds the total of all long positions to the total of all the short positions entered by clients.
2. The net basis allows these to be offset against each other (most exchanges currently use net margining).
Credit Risk
The purpose of the margining system is to ensure that traders do not walk away from their commitments.
Overall the system has been successful as investors have always had their contracts honored.
While some investors walked away from their long positions in S&P futures after the 20% decline in the S&P 500
index on October 19, 1987, which also caused some brokers to become bankrupt due to their inability to meet
margin calls on contracts they entered into on behalf of their clients, everyone who had a short futures position on
the S&P 500 got paid off.
Collateralization in OTC Markets
Credit risk has been a feature of the over–the–counter markets.
There is always a chance that the party on the other side of an OTC trade will default.
To reduce credit risk, the OTC market is now imitating the margining system adopted by exchanges with a
procedure known as collateralization.
Example: Consider company A and B who have an outstanding OTC counter contract.
They could enter into a collateralization agreement where they value the contract each day using a pre–agreed
valuation method.
Over time if the value of the contract to company A increases, company B is required to pay collateral equal to this
increase to company A.
If the value of the contract to company A decreases, company A is required to pay collateral equal to the decrease
to company B.
Collateralization significantly reduces the credit risk in OTC contracts and is discussed further in Chapter 20.
Note: In the 1990s, collateralization agreements were used by a hedge fund, Long–Term Capital Management
(LTCM). This allowed LTCM to become highly leveraged, and although the contracts did provide credit risk
protection, the high leverage left the hedge fund vulnerable to other risks. LTCM would buy bond × (less liquid) and
short Bond Y (more liquid) and would wait for the prices of the two bonds to converge at some future time. However,
due to a major Russian default and its impact on illiquidity spread on bonds, the prices of the bonds LTCM had bought
went down and the prices of those it had shorted increased. It was required to post collateral on both. LTCM could not
make payments required under their collateralization agreements and this lead to their demise.
2.6 Delivery
To close a position, the broker of the party with the short position sends a notice of intention to deliver to the
exchange. The exchange then selects a party with an outstanding long position to accept delivery.
Cash Settlement
Financial futures are settled in cash, since it is inconvenient or impossible to deliver the underlying asset (e.g. not
feasible to deliver a portfolio of 500 stocks in the case of a futures contract on the S&P 500).
The final settlement price is set equal to the spot price of the underlying asset at either the opening or close of trading
on that day (e.g. in the S&P 500 futures contract, all contracts are declared closed on the third Friday of the delivery
month and the final settlement price is the opening price of the index on that day).
2.8 Regulation
Futures markets are regulated federally by the Commodity Futures Trading Commission (CFTC) which:
i. is responsible for licensing futures exchanges and approving contracts. For contracts to be approved, the contract
must have some useful economic purpose (e.g. it must serve the needs of hedgers as well as speculators).
ii. looks after the public interest, by ensuring that prices are communicated to the public and that futures traders
report their outstanding positions (if they are above certain levels)
iii. licenses all individuals who offer their services to the public in futures trading, conducts backgrounds checks and
determines minimum capital requirements.
iv. deals with complaints brought by the public and ensures that disciplinary action is taken when appropriate.
In 1982, the National Futures Association was formed and assumed some of the CFTC responsibilities. The NFA’s
objective is to prevent fraud and to ensure that the market operates in the best interests of the public. NFA requires its
members to pass an exam, and is authorized to monitor trading and take disciplinary action when appropriate.
The Securities and Exchange Commission (SEC), the Federal Reserve Board, and the U.S. Treasury Department also
have jurisdictional rights over some aspects of futures trading. These bodies are concern with the effects of futures
trading on the spot markets for securities such as stocks, Treasury bills, and Treasury bonds. Further, the SEC
currently has an effective veto over the approval of new stock or bond index futures contracts.
Trading Irregularities
One type of trading irregularity occurs when an investor group tries to "corner the market".
i. A huge long futures position is taken to exercise some control over the supply of the underlying commodity.
ii. As the maturity of the futures contracts approaches, an investor group may not close out its position, and thus, the
number of outstanding futures contracts would exceed the amount of the commodity available for delivery.
iii. Investors with short positions would realize that it is difficult to deliver, become desperate to close out their
positions, and this would result is a large rise in both futures and spot prices.
Regulators deal with this type of abuse by increasing margin requirements, imposing stricter position limits,
prohibiting trades that increase a speculator's open position, and forcing market participants to close out their positions.
Other types of trading irregularities involve traders on the floor of the exchange.
In 1989, the FBI carried out a two year investigation, due to offenses including overcharging customers, not paying
customers the full proceeds of sales, and traders using their knowledge of customer orders to trade first for themselves.
Tax laws define hedging as a transaction entered-into in the normal course of business primarily for two reasons:
1. To reduce the risk of price changes/currency fluctuations with respect to property held or to be held by the
taxpayer for the purposes of producing ordinary income.
2. 2. To reduce the risk of price/interest rate/ currency fluctuations with respect to borrowings made by the taxpayer.
a. Gains or losses are treated as ordinary income.
b. The timing of the recognition of gains or losses matches that of income or deduction from the hedged items.
Special rules apply to foreign currency futures transactions. A taxpayer can elect to treat gains and losses as ordinary
income, regardless of whether the contract is entered into for hedging or speculative purposes.
If a taxpayer does not make this election, then foreign currencies futures transactions are treated in the same way as
other futures transactions.
2.10 Forward Contracts vs. Futures Contracts
Both forwards and futures are agreements to buy or sell an asset for a certain price at a certain future time.
The main differences between forward and futures contracts are summarized below.
Forward Futures
Private contract between two parties Traded on an exchange
Not standardized Standardized contract
Usually one specified delivery date Range of delivery dates
Settled at end of contract Settled daily
Delivery or cash settlement usually takes place Contract is usually closed out prior to maturity
Some credit risk Virtually no credit risk
Profits from Forward and Futures Contracts
Keep in mind the following:
Under a forward contract, the whole gain or loss is realized at the end of the life of the contract.
Under a futures contract, the gain or loss is realized day by day because of the daily settlement procedures.
Example: Assume the following:
The sterling exchange rate for a 90–day forward contract is 1.6000 and that this is also the rate for the futures
price for a contract to be delivered in exactly 90 days.
Investor A is long £1 million in a 90–day forward contract and
Investor B is long £1 million in 90–day futures contracts. (Note: because each futures contract is for the
purchase or sale of £62,500, investor B must purchase a total of 16 contracts i.e. £62,500 × 16 contracts =
£1 million.)
The spot exchange rate in 90 days proves to be 1.8000 dollars per pound.
What is the difference between the gains and losses under the two contracts?
Investor A makes a gain of $200,000 on the 90th day [16 × £62,500 × ($1.80/£ – $1.60/£)].
Investor B makes the same gain–but spread out over the 90–day period. On some days investor B may
realize a loss, whereas on other days he or she makes a gain. However, in total, when losses are netted
against gains, there is a gain of $200,000 over the 90–day period.
Foreign Exchange Quotes
a. Futures prices are always quoted as the number of U.S. dollars per unit of the foreign currency or as the number of
U.S. cents per unit of the foreign currency.
b. Forward prices are always quoted in the same way as spot prices. For other than British pound, the euro, the
Australian dollar, and the New Zealand dollar, forward quotes show the number of units of the foreign currency per
U.S. dollar (USD).
Example: Consider the Canadian dollar (CAD).
A futures price quote of 0.7050 USD per CAD corresponds to a forward price quote of 1.4184 CAD per USD (1.4184
= 1/0.7050).
Section Description
3.0 Introduction
3.1 Basic Principles
3.2 Arguments For and Against Hedging
3.3 Basis Risk
3.4 Cross Hedging
3.5 Stock Index Futures
3.6 Rolling the Hedge Forward
Appendix Proof of Minimum Variance Hedge Ratio Formula
3.0 Introduction
Many participants in the futures markets are hedgers. The purpose of a futures hedge is to reduce risk by making the
outcome more certain. A perfect hedge is one that completely eliminates the risk. Perfect hedges are rare.
Some of the questions addressed in this chapter are:
a. When is a short or long futures position appropriate?
b. Which futures contract should be used?
c. What is the optimal size of the futures position for reducing risk?
Two additional notes:
a. The strategies discussed by the author can be described as hedge–and–forget strategies (i.e. no attempt is made to
adjust the hedge once it has been put in place).
b. The author treats futures contracts as forward contracts.
Future Price
Time
t1 t2
Example:
Assume a hedge is put in place at time t1 and closed out at time t2.
Given the following: S1 = 2.50, F1 = 2.20, S2 = 2.00, and F2 = 1.90.
Compute the basis as b1 = S1 – F1 = 2.50 – 2.20 = 0.30 and
b2 = S2 – F2 = 2.00 – 1.90 = 0.10
Scenario 1: Short Hedge: (the asset will be sold at time t2 and a short futures position is taken at t1).
a. The price realized for the asset is S2 and the profit on the futures position is F1 – F2.
The effective price obtained for the asset with hedging is S2 + F1 – F2 = F1 + b2 = 2.30
b. A perfect hedge (i.e., a hedge eliminating all uncertainty about the price obtained) results when b2 is known at the
time of the hedge. However, this is usually not the case.
Thus, basis risk is the uncertainty of the value of b2
Scenario 2. Long Hedge: (the asset will be bought at time t2 and a long futures position is taken at t1).
The price paid for the asset is S2 and the loss on the hedge is F1 – F2
The effective price obtained for the asset with hedging is S2 + F1 – F2 = F1 + b2 = 2.30
The basis risk for an investment asset stems from uncertainty as to the level of the risk–free interest rate and the asset’s
future yield.
Consider the case when the asset being hedged and the asset underlying the futures contract may not be the same. In
this case, the basis risk is then usually greater.
Define S 2* as the price of the asset underlying the futures contract at time t2.
Recall that S2 is the price of the asset being hedged at time t2.
By hedging, a company ensures that the price paid (or received) for the asset is S2 + F1 – F2 (which can be written as)
F1 ( S2* F2 ) ( S2 S2* )
Thus, two components of the basis can be defined:
1. ( S 2* F2 ) is the basis if the asset being hedged were the same as the asset underlying the futures contract.
2. S 2 S 2* is the basis arising from the difference between the two assets.
Note: basis risk can lead to an improvement or a worsening of a hedger’s position.
Choice of Contract
The choice of the futures contract is a key factor affecting basis risk. The choice has two components:
1. The choice of the asset underlying the futures contract.
If the asset being hedged does not match the asset underlying a futures contract, it is necessary to determine which of
the available futures contracts has futures prices most closely correlated with the price of the asset being hedged.
2. The choice of the delivery month (which is influenced by the following factors).
a. Price volatility during the month of delivery. Thus, a contract with a later delivery month is usually chosen.
b. Taking actual possession. A long hedger runs the risk of taking delivery of the physical asset if the contract is
held during the delivery month, and this can be expensive and inconvenient.
Thus, a good rule of thumb is to choose a delivery month that is as close as possible to, but later than, the expiration of
the hedge.
Example: Assuming the following information is given:
On 3/1, a U.S. company is notified that it will receive 50 million Japanese yen at the end of July.
Yen futures contracts have delivery months of March, June, September, and December, and one contract is for the
delivery of 12.5 million yen.
The company shorts four September yen futures contracts on 3/1.
The company closes out its position when the yen are received at the end of July.
The futures price on 3/1 in cents per yen is 0.7800 and the spot and futures prices when the contract is closed out
are 0.7200 and 0.7250, respectively.
1. Compute the gain on the futures contract. 0.7800 – 0.7250 = 0.0550 cents per yen.
2. Compute the basis when the contract is closed out.
Basis equals 0.7200 – 0.7250 = –0.0050 cents per yen
3. Compute the effective price obtained in cents per yen:
Effective price equals the final spot price plus the gain on the futures: 0.7200 + 0.0550 = 0.7750
Effective price also equals the initial futures price plus the final basis: 0.7800 – 0.0050 = 0.7750
The total amount received by the company for the 50 million yen is 50 × 0.00775 million dollars, or $387,500.
Example: Assuming the following information:
On 6/8, a company knows it will need to purchase 20,000 barrels of crude oil some time in October or November.
The futures price on 6/8 is $18.00 per barrel.
Oil futures contracts are traded for delivery every month and the contract size is 1,000 barrels.
The company decides to use the December contract and takes a long position in 20 December contracts.
The company closes out its futures contract on 11/10, when the spot price and futures price are $20.00 per barrel
and $19.10 per barrel.
1. Compute the gain on the futures contract: 19.10 – 18.00 = $1.10 per barrel.
2. Compute the basis when the contract is closed out: 20.00 – 19.10 = $0.90 per barrel.
3. Compute the effective price paid (in dollars per barrel):
Effective price paid equals the final spot price less the gain on the futures: 20.00 – 1.10 = 18.90.
Effective price paid also equals the initial futures price plus the final basis: 18.00 + 0.90 = 18.90
The total price paid is 18.90 × 20,000 = $378,000.
3.4 Cross Hedging
Cross hedging occurs when the two assets are different. If an airline is concerned about the future price of jet fuel and
there is no futures contract on jet fuel, it might choose to use heating oil futures contracts to hedge its exposure.
A hedge ratio:
is the ratio of the size of the position taken in futures contracts to the size of the exposure.
of 1.0 is not necessarily optimal, if the objective of the hedger is to minimize risk.
The following notation is used to explain the hedge ratio:
S: the change in spot price, S, during the life of the hedge.
F: the change in futures price, F, during the life of the hedge.
S: the standard deviation of S
F: the standard deviation of F
: coefficient of correlation between S and F
h*: hedge ratio that minimizes the variance of the hedger’s position
Appendix A proves that h* is determined as follows: h* = (S /F).
If = 1 and F = S, the hedge ratio, h*, is 1.0 and the futures price mirrors the spot price.
If = 1 and F = 2S, the hedge ratio h* is 0.5 so the futures price always changes by twice as much as the spot price.
The optimal hedge ratio, h*, is the slope of the best fit line when S is regressed against F. The hedge effectiveness is
F2
defined as the proportion of the variance that is eliminated by hedging. This is 2, or h*2
S2
Optimal Number of Contracts
To determine the optimal number of contracts, we first define the following variables:
NA: Size of position being hedged (units) QF: Size of one futures contract (units)
N *: Optimal number of futures contracts for hedging
The futures contracts used should have a face value of h*NA. The number of futures contracts is required is therefore
h* N A
given by N *
QF
Example: Hedging the price of fuel by buying futures contracts on heating oil given:
A company will buy 2 million gallons of jet fuel in 1 month.
S (standard deviation of the change in the price per gallon of jet fuel) = 0.0263.
F (standard deviation of the change in the futures price) = 0.0313
(correlation coefficient between the one month change in the price of jet fuel and the futures price) = 0.928
Compute the optimal hedge ratio: h 0.928 0.0263 0.78
0.0313
Compute the optimal number of contracts to buy:
Given that a heating oil futures contract is on 42,000 gallons, the company should therefore buy
2 M gal
0.78 37.14 37 contracts
42 K gal/contract
10. The US Dollar Index is a trade–weighted index of the values of six foreign currencies (the euro, yen, pound,
Canadian dollar, Swedish krona, and Swiss franc).
11. The Nikkei 225 Stock Average is based on a portfolio of 225 of the largest stocks trading on the Tokyo Stock
Exchange. Stocks are weighted according to their prices. One futures contract (traded on the CME) is on $5 times
the index.
12. The Share Price Index is the All Ordinaries Share Price Index, a broadly based index of Australian stocks. The
CAC–40 Index is based on 40 large stocks trading in France.
13. The Xetra DAX Index is based on 30 stocks trading in Germany.
14. The FTSE 100 Index is based on a portfolio of 100 major UK stocks listed on the London Stock Exchange.
15. The DJ Euro Stoxx 50 Index and the DJ Stoxx 50 Index are two different indices of blue–chip European stocks
compiled by Dow Jones and its European partners. The futures contracts on these indices trade on Eurex and are
on 10 times the values of the indices measured in euros.
16. The other indices shown in Table 3.3 are not stock indices.
The DJ–AIG commodity index and the GSCI index futures contract track commodity prices.
The TRAKRS long–short tech index is designed to reflect the performance of a portfolio that is long
individual technology stocks and short financial instruments representing technology sectors.
Recall:
Futures contracts on stock indices are settled in cash, not by delivery of the underlying asset.
All contracts are marked to market to either the opening price or the closing price of the index on the last trading
day, and the positions are then deemed to be closed. For example, contracts on the S&P 500 are closed out at the
opening price of the S&P 500 index on the 3rd Friday of the delivery month.
Stock index futures can be used to hedge an equity portfolio. We begin by defining the following notation:
P: the current value of the portfolio.
A: the current value of the stocks underlying one futures contract.
: the appropriate hedge ratio for a portfolio.
is the slope of the best fit line when excess return on the portfolio over the risk–free rate is regressed against the
excess return of the market over the risk–free rate.
N*: the optimal number of futures contracts for hedging.
1. If the portfolio mirrors the index, a hedge ratio of 1.0 is appropriate, and the number of futures contracts that
P
should be shorted is N *
A
P
2. When the portfolio does not mirror the index, the N *
A
When = 1.0, the return on the portfolio tends to mirror the return on the market
When = 2.0, the excess return on the portfolio tends to be twice as great as the excess return on the market.
When = 0.5, it tends to be half as great.
Example: Compute the number of contracts that need to be shorted to hedge a portfolio of $1 million, given that the
current value of the index is 1.000, and each futures contract is on $250 times the index.
$1M
N * 1.0 4
$250 $1,000
The computation of Beta using the Capital Asset pricing model:
If the expected return–beta relationship holds, then the expected rate of return (as a %) on any portfolio i is
Cov( ri , rM )
E ( ri ) rf i [ E ( Rm ) rf ] , where i
M
2
Example: A futures contract on the S&P 500 with 4 months to maturity will be used to hedge the value of the portfolio
over the next 3 months. You are given the following information:
The value of S&P 500 index: 1,000; The value of portfolio: $5,000,000; The risk–free interest rate: 10% pa (2.5% per
3 months); The dividend yield on index: 4% pa (1.0% per 3 months); The Beta of the portfolio: 1.5.
One futures contract is for delivery of $250 times the index.
1. Compute the current futures price: 1,000e (0.10 – 0.04) × 4/12 = 1,020.20
5,000,000
2. Compute the number of futures contracts that should be shorted to hedge the portfolio: 1.5 30
250,000
3. If the index turns out to be 900 in 3 months, compute the futures price: 900e(0.10 – 0.04) × 1/12 = 904.51
4. Compute the gain from the short futures position: 30 × (1,020.20 – 904.51) × 250 = $867,676
5. Compute the expected return (as a %) on the portfolio.
a. The return on the index over a 3 month period is a loss of 10% [(900/1000) – 1.0] = –0.10
b. The return from the dividend yield on index is 1%.
c. The net return on the index is –0.10 + 0.01 = –0.09.
d. The expected rate of return on the portfolio can be computed using: E ( ri ) rf i [ E ( Rm ) rf ]
The expected return (in percentage) on the portfolio equals 2.5 + [1.5 × (–9.0 – 2.5)] = –14.75
6. Compute the expected value of the portfolio (inclusive of dividends) at the end of the 3 months:
$5,000,000 × (1 – 0.1475) = $4,262,500.
7. Compute the expected value of the hedger’s position (including the gain on the hedge):
$4,262,500 + $867,676 = $5,130,176
The table below shows that the value of the position in 3 months depends on the value of the index in 3 months:
Value of index In three months 900.00 950.00 1,000.00 1,050.00 1,100.00
Three months total value of position ($000) 5,130,176 5,128,296 5,126,416 5,124,537 5,122,657
Conclusions:
a. The hedging scheme results in a value for the hedger’s position close to $5,125,000 at the end of 3 months.
b. The effect of the hedge results in the hedger’s position growing at the risk–free rate (2.5% per 3 months).
Reasons for Hedging an Equity Portfolio
Q. Why the hedger use futures contracts to hedge when the hedger can sell the portfolio and buy Treasury bills?
A1. The hedger feels that the stocks in the portfolio have been chosen well and will outperform the market.
Therefore, a hedge using index futures removes the risk arising from market moves and leaves the hedger exposed only
to the performance of the portfolio relative to the market.
A2. The hedger is planning to hold a portfolio for a long period of time and requires short–term protection.
Changing Beta of the hedgers’ portfolio
Based on the example above, the purchase of 30 futures contracts reduces the beta of the portfolio to zero.
To alter the beta of the portfolio to something other than zero, we adjust the formula to determine the number of
contracts purchased. Therefore,
a. To change the beta of the portfolio from to *, where > * a short position in ( – *)(P/A) number of
contracts is required.
b. To change the beta of the portfolio from to *, where < * a long position in ( – *)(P/A) number of
contracts is required.
Exposure to the Price of an Individual Stock
Some exchanges trade futures contracts on selected individual stocks, but in most cases a position in an individual
stock can only be hedged using a stock index futures contact.
Example: You are given the following information.
An investor holding 20,000 IBM shares, worth $100 per share, is concerned about the market’s volatility over the
next month and chooses to hedge the position using an August futures contract on the S&P 500.
Each contract is for delivery of $250 times the index.
The of IBM is 1.1.
The current level of the index is 900, and the current futures price for the August contract is 908.
20,000 $100 2,000,000
Q1. Compute the number of contracts that should be shorted: 1.1 1.1 9.78 10
900 250 225,000
Assume IBM rises to $125 during the month, and the futures price of the S&P 500 rises to 1080.
Q1. Compute the effect of the hedge:
The investor’s gains $500,000 [20,000 × ($125 – $100)] on the shares of IBM; the investor loses $430,000 [10 × $250
× (1080 – 908)] on the futures index; the net effect is a gain to the investor of $70,000
Q2. What was the purpose of the hedge?
A. To reduce risk making unfavorable outcomes less unfavorable and favorable outcomes less favorable.
Section Description
5.0 Introduction
5.1 Investment Assets vs. Consumption Assets
5.2 Short Selling
5.3 Assumptions and Notation
5.4 Forward Price of an Investment Asset
5.5 Known Income
5.6 Known Yield
5.7 Valuing Forward Contracts
5.8 Are Forward and Future Prices Equal?
5.9 Futures Prices of Stock Indices
5.10 Forward and Future Contracts on Currencies
5.11 Futures on Commodities
5.12 Cost of Carry
5.13 Delivery Options
5.14 Futures Prices and the Expected Future Spot Price
5.0 Introduction
This chapter examines how forward and futures prices relate to the spot price of the underlying asset.
Forward contracts are settled at the end of the contract while futures contracts are settled daily. The settlement of a
forward contract by a single payment at maturity makes forward contracts easier to analyze than futures contracts.
However, forward and futures prices of an asset are equal when the maturities of the two contracts are the same.
Interest is paid on the balance in margin accounts and, if the interest rate offered is unacceptable, marketable
securities such as Treasury bills can be used to meet margin requirements.
The proceeds of the sale of the asset belong to the investor and normally form part of the initial margin.
Examples of cash flows from short sale and purchase of shares.
Purchase of shares
April: Purchase 500 shares for $120 –$60,000
May: Receive dividend +$500
July: Sell 500 shares for $100 per share +$50,000
Net profit = –$9,500
Short sale of shares
April: Borrow 500 shares and sell them for $120 +$60,000
May: Pay dividend –$500
July: Buy 500 shares for $100 per share (replace shares and close position) –$50,000
Net profit = +$9,500
Note: Securities used to be sold short only on an uptick (when the most recent movement in the price is up).
2. The present value of the first coupon payment is ( 40 e 0.090.50 38.24 ). 38.24 is borrowed and invested at 9%
so that in six months it repays the first coupon payment.
3. 900 – 38.24 = 861.76 is borrowed at 10% for one year and 861.76 × 1.1 = $952.39 needs to be repaid at the end of
the year.
4. The second coupon provides $40 toward this amount, and $930 is received for the bond under the terms of the
forward contract. A net profit of $40 + $930 – $952.39 = $17.61 is made.
Scenario 2: Assume the forward price equals $905. Since this is below its theoretical value ($912.39):
1. An arbitrager who holds the bond can sell the bond and enter into a long forward contract.
2. Using the values in the previous scenario, the arbitrager’s gain is $952.39 – $40 – $905 = $7.39
Conclusions regarding the scenarios:
The first scenario a profit of $17.61 when the forward price ($930) is greater its theoretical price ($912.39).
The 2nd scenario produces a profit of $7.39 when the forward price ($905) is less than $912.39.
Therefore, when no arbitrage opportunities exist, the forward price must be $912.39.
A Generalization (for investments generating known income)
When an investment asset generates a known cash income during the life of a forward contract:
F0 = (S0 – I )erT , where I equals the present value of the income.
In the previous example, S0 = 900.00; I = 40e–0.09 × 0.5 + 40e–0.10 × 1 = 74.433; r = 0.10, and T = 1.0
Thus, F0 = (900.00 – 74.433)e0.10 × 1 = $912.39
5.6 Known Income
Example: Consider a 10–month forward contract on a stock with a price of $50.
Assume: The risk–free rate of interest (continuously compounded) is 8% per annum for all maturities.
Dividends of $0.75 per share are expected after 3, 6, and 9 months. Thus, S0 = 50, r = 0.08, T = 10/12
Compute the forward price.
Step 1: Compute the PV of the dividends, I = 0.75e–0.08 × 3/12 + 0.75e–0.08 × 6/12 + 0.75e–0.08 × 9/12 = 2.162
Step 2: Using the formula above and the value from Step 1, find the forward price: F0 = (50 – 2.162)e0.08 × 10/12 = $51.14
The risk–free rate of interest (with continuous compounding) is 10% per annum.
The stock price is $25, and the delivery price is $24.
Thus, S0 = 25, r = 0.10, T = 0.5, and K = 24.
Compute:
The 6 month forward price. F0 S0 e rT 25 e0.100.50 26.28
The value of the forward contract (using the 1st formula): f (26.28 24) e 0.100.50 2.17
The value of the forward contract (using the 2nd formula): f 25 24 e 0.100.50 2.17
Other Formulas:
The value of a short forward contract with delivery price K: f ( K F0 )e rT
The value of a long forward contract on an investment asset with the present value a known income (I):
f S0 I Ke rT
The value of a long forward contract on an investment asset that provides a known dividend yield at rate q:
f S0 e qT Ke rT
5.8 Are Forward and Future Prices Equal?
Forward and futures prices in theory are no longer the same when interest rates vary. The following argument for the
relative prices of forwards and futures is based on the premise that S, the price of the underlying asset, is strongly
correlated with interest rates.
When S increases, an investor holding a long futures position makes an immediate gain because of the daily settlement
procedure. The gain is invested at a higher than average rate of interest.
When S decreases, an investor holding a long futures position incurs a loss. The loss is financed at a lower than
average rate of interest.
Based on the above argument,
When S is strongly positively correlated with interest rates, futures prices tend to be higher than forward prices.
When S is strongly negatively correlated with interest rates, forward prices tend to be higher than futures prices.
Additional views of the equality of forward and futures prices:
i. The differences between forward and futures prices for contracts that last only a few months are small.
ii. As the life of a futures contract increases, the differences between forward and futures contracts are liable to
become significant.
iii. Factors not reflected in theoretical models that may cause forward and futures prices to be different include taxes,
transactions costs, and the treatment of margins.
iv. Empirical research provides conflicting evidence as to whether statistical differences exist between forward and
futures prices.
Note: the value of interest paid in a foreign currency depends on the value of the foreign currency.
Example: Assume the interest rate on British pounds is 5% per annum.
To a US investor the British pound provides an income equal to 5% of the value of the British pound per annum. In
other words it is an asset that provides a yield of 5% per annum.
7.0 Introduction
A swap is an agreement by two companies to exchange cash flows in the future.
A forward contract may be viewed as a swap. For example:
On 3/1/02 a company enters into a 1 year forward contract to buy 100 ounces of gold for $300 per ounce.
This is equivalent to a swap where the company agrees that on 3/1/03, it will pay $30,000 and receive 100S, where S
is the market price of one ounce of gold on that date.
Unlike forwards, swaps typically lead to multiple cash flow exchanges taking place on several future dates.
Pays interest at 5%
The cash flows involved in the Swap are shown on the next page.
Cash flows (million of dollars) to Microsoft in a $100 million three–year interest rate swap when a fixed rate of 5% is
paid and LIBOR is received
7.3 Confirmations
A confirmation is the legal agreement underlying a swap and is signed by representatives of the two parties.
The confirmation specifies that the following business day convention is to be used and that the US calendar
determines which days are business days and which days are holidays.
If a payment date falls on a weekend or a US holiday, the payment is made on the next business day.
September 5, 2004, is a Sunday.
The first exchange of payments in a swap is therefore on Monday September 6, 2004.
7.4 The Comparative–Advantage Argument
Consider the use of an interest rate swap to transform a liability.
Some companies have a comparative advantage when borrowing in fixed–rate markets, whereas others have a
comparative advantage in floating–rate markets. Thus, a company may borrow fixed when it wants floating, or borrow
floating when it wants fixed.
The swap is used to transform a fixed–rate loan into a floating–rate loan, and vice versa.
Example:
AAA-Corp and BBB-Corp (named after their credit rating) wish to borrow $10 million for 5 years
BBB-Corp wants to borrow at a fixed rate of interest; AAA-Corp wants to borrow at a floating rate.
Borrowing rates that provide a basis for the comparative–advantage argument
Fixed Floating
AAA-Corp 10.0% 6–month LIBOR + 0.3%
BBB-Corp 11.2% 6–month LIBOR + 1.0%
Notes: BBB-Corp pays 1.2% more than AAA-Corp in fixed–rate markets and only 0.7% more than AAA-Corp in
floating–rate markets.
BBB-Corp appears to have a comparative advantage in floating–rate markets, whereas AAA-Corp appears to have a
comparative advantage in fixed–rate markets.
They enter into a swap agreement to ensure that AAA-Corp ends up with floating–rate funds and BBB-Corp ends up
with fixed–rate funds.
Terms of the swap: AAA-Corp agrees to pay BBB-Corp interest at six–month LIBOR on $10 million.
In return, BBB-Corp agrees to pay AAA-Corp interest at a fixed rate of 9.95% per annum on $10 million.
Section Description
8.0 Introduction
8.1 Types of Options
8.2 Option Positions
8.3 Underlying Assets
8.4 Specifications of Stock Options
8.5 Newspaper Quotes
8.6 Trading
8.7 Commissions
8.8 Margins
8.9 The Options Clearing Corporation
8.10 Regulation
8.11 Taxation
8.12 Warrants, Executive Stock Options, and Convertibles
8.13 Over the Counter Markets
8.0 Introduction
Options are different from forward, futures, and swap contracts:
An option gives its holder the right to do something (which may include not exercising this right)
In forward, futures, or swap contracts, the two parties have committed themselves to some action.
General definitions:
A call option is the right to buy an asset for a certain price.
A put option is the right to sell an asset for a certain price.
A European option can be exercised only at the end of its life.
An American option can be exercised at any time during its life.
Keep in mind that there are four types of option positions: a long position in a call, a long position in a put, a short
position in a call, and a short position in a put.
In general, call options should always be exercised at the expiration date if the stock price is above the strike price.
Figure 1: Profit from buying a European call option on one eBay share. Option price = $5; strike price = $100
Put Options
The purchaser of a put option is hoping that it will decrease. Consider the following:
An investor who buys a European put option to sell 100 shares in IBM with a strike price of $70.
The current stock price is $65, the expiration date of the option is in 3 months, and the price of an option to sell
one share is $7.
The initial investment is $700 (i.e. $7 × 100).
The option is European, so it will be exercised only if the stock price is below $70 on the expiration date.
Suppose that the stock price is $55 on the expiration date.
The investor buys 100 shares for $55 per share and, sells the same shares for $70 to realize a gain of $15 per share,
or $1,500 (transactions costs ignored.)
1,500 minus the $700 initial cost of the option yields the investor a net profit of $800.
If the final stock price is above $70, the put option expires worthless, and the investor loses $700.
Figure 2: Profit from buying a European put option on one IBM share. Option price = $7; strike price = $70.
Early Exercise
Exchange–traded stock options are generally American rather than European. Thus, investors would not have to wait
until the expiration date before exercising the option.
Later in this chapter, the circumstances under which it is optimal to exercise American options prior to maturity will be
described.
Figure 3: Profit from writing a European call option on eBay. Option price = $5; strike price = $100.
It is useful to characterize European option positions in terms of the terminal value or payoff to the investor at
maturity. The initial cost of the option is then not included in the calculation.
If K is the strike price and ST is the final price of the underlying asset, the payoff from a long position in a
European call option is max(ST – K, 0)
This reflects the fact that the option will be exercised if ST > K and will not be exercised if ST K.
The payoff to the holder of a short position in the European call option is
–max(ST – K, 0) = min(K – ST, 0)
The payoff to the holder of a long position in a European put option is max(K – ST, 0)
The payoff from a short position in a European put option is –max(K – ST, 0) = min (ST – K, 0)
Figure 5: Payoff’s from positions in European options: (a) long call; (b) short call; (c) long put; (d) short put.
Strike price = K; price of asset at maturity = ST.
The intrinsic value of a call and put option are defined as follows:
For a call option, it is defined as the max of (S – K, 0).
For a put option, it is defined as the max of (K – S, 0).
The total value of an option can be thought of as the sum of its intrinsic value and its time value.
4. Flex Options
The CBOE offers flex options on stock and stock indices.
These options give traders on the floor of the exchange the ability to agree to nonstandard terms (e.g. strike prices
or expiration dates different from those usually offered by the exchange).
Flex options are an attempt by the exchanges to regain business from the over–the–counter markets.
5. Dividends and Stock Splits
Both exchange–traded and over–the–counter options are not generally adjusted for cash dividends.
Exchange–traded options are adjusted for stock splits: In general, an n–for–m stock split should cause:
a. the stock price to go down to m/n of its previous value.
b. the exercise price is reduced to m/n of its previous value, while the number of shares covered by one contract is
increased to n/m of its previous value.
Example: Consider a call option to buy 100 shares of a company for $30 per share.
After a 2–for– 1 stock split, the option holder has the right to purchase 200 shares for $15 per share.
Stock options are adjusted for stock dividends.
Example: Consider a put option to sell 100 shares of a company for $15 per share.
After a 25% stock dividend is declared (which is equivalent to a 5–for–4 stock split), the option holder has the right to
sell 125 shares for $12.
6. Position Limits (Exercise Limits)
Position limits and exercise limits are designed to prevent investors or groups of investors from having undue
influence on the market. Therefore,
The CBOE specifies position limits for option contracts. This defines the maximum number of option contracts
that a trader can hold on one side of the market. Long calls and short puts are on the same side of the market.
Options on the largest traded stocks have positions limits of 75,000 contracts, while options on smaller
capitalization stocks have position limits of 60,000, 31,500, 22,500, or 13,500 contracts.
8.5 Newspaper Quotes
An extract of stock option quotations from Wall Street Journal is shown and described as follows:
Column (1) list the company and its stock price
Columns (3) and (4) identify the strike price and the expiration month for the option.
Columns (4) and (5) show the volume of trading and price at last trade for the call option.
Columns (6) and (7) show the volume of trading and price at last trade for the put option.
(1) (2) (3) (4) (5) (6) (7)
Call Put
Option StrikeP Exp Vol last Vol Last
AT&T
23.35 22.50 Apr 2933 2 12689 1.10
8.6 Trading
Options exchanges use market makers to facilitate trading. A market maker will:
quote both a bid and an offer price on the option.
i. The bid is the price at which the market maker is prepared to buy.
ii. The offer is the price at which the market maker is prepared to sell.
The offer is always higher than the bid, and the amount by which the offer exceeds the bid is referred to as the
bid–offer spread.
The exchange sets upper limits for the bid–offer spread (e.g. the spread can be no more than $0.25 for options
priced at less than $0.50, etc).
ensure that buy and sell orders can always be executed at some price without delays.
Offsetting Orders (used to close out positions)
Buyers of options can close out their positions by issuing an offsetting order to sell the same option.
Option writers can close out their position by issuing an offsetting order to buy the same option.
8.7 Commissions
Commissions for option trades vary dependent upon broker type (e.g. discount vs. full service).
The actual amount charged is usually a fixed cost plus a proportion of the dollar amount of the trade, subject to
maximum and minimum commissions.
Based on the discount brokerage commission table below, the purchase of 8 contracts when the option price is $3
would cost $20 + [0.02 × ($3 × 8 × 100)] = $68 in commissions.
Dollar amount of Trade Commission*
< $2,500 $20 + 0.02 of the dollar amount
$2,500 to $10,000 $45 + 0.01 of the dollar amount
> $10,000 $120 + 0.0025 of the dollar amount
*Maximum commission is $30 per contract for the first five contracts plus $20 per contract for each additional
contract.
Minimum commission is $30 per contract for the first contract plus $2 per contract for each additional contract.
The market maker’s bid–offer spread is a hidden cost in option trading (and in stock trading). For example:
Given a bid price of $4.00, an offer price of $4.50, we can assume that a “fair” price for the option is halfway
between the two amounts, or $4.25.
The cost to the buyer and to the seller of the market maker system is the difference between the fair price and the
paid price (in this case, the cost is $4.50 – $4.25 = $0.25 per option, or $25 per contract).
8.8 Margins
Differences exist in how payment is rendered when purchasing stocks and options in the U.S.
When shares are purchased, an investor can either pay cash or borrow using a margin account.
The initial margin required is usually 50% of the value of the shares.
The maintenance margin is usually 25% of the value of the shares.
When call and put options are purchased, the option price must be paid in full since options are highly leveraged
investments. Buying on margin would raise this leverage to an unacceptable level.
Investors who write options are required to maintain funds in a margin account.
The size of the margin required depends on the circumstances.
Writing Naked Options
A naked option is one that is not combined with an offsetting position in the underlying stock.
The initial margin when writing a naked options is the greater of the results of the following two calculations:
1. 100% of the proceeds of the sale + 20% of the underlying share price – the amount by which the option is out of
the money (if any).
2. 100% of the proceeds of the sale + 10% of the underlying share price.
For options on a broadly based index, 15% (instead of 20% in the above calculation) is generally used, because an
index is usually less volatile than the price of an individual stock.
Example:
An investor writes 4 naked call option contracts on a stock.
The option price is $5, the strike price is $40, and the stock price is $38.
Note: the option is $2 out of the money.
The second calculation gives 400 × [$5 + (0.10 × $38)] = $3,520
The first calculation gives 400 × [$5 + (0.2 × $38) – $2] = 4,240 (and this is the initial margin requirement).
Notes:
If the investor wrote a put option, the margin requirement would be $5040 [400 ($5 + 0.2 x $38)] (since the put is
already $2 in the money).
In both cases, the proceeds of the sale, $2,000 (400 × $5), can be used to form part of the margin account.
Writing Covered Calls
A covered call occurs when call options are written on shares that might have to be delivered, but are already owned.
Covered calls are less risky than writing naked calls because the worst case the investor faces is selling shares already
owned at below their market value.
If covered call options are out of the money, no margin is required.
Example: An investor decides to buy 200 shares of a certain stock on margin and to write two call options.
The stock price is $63, the strike price is $65, and the price of the option is $7.
Compute the minimum cash required from an investor given the following trades.
To determine the minimum cash required, note the following:
i. the margin account allows the investor to borrow 50% of the price of the stock, or $6,300.
ii. the price received for the options is $1400 ($7 × 200) are used to purchase of the shares.
iii. the cost of the shares is $12,600 ($63 × 200).
Thus, the minimum cash required for his/her trades is: $12,600 – $6,300 – $1,400 = $4,900.
8.10 Regulation
Regulation occurs in a number of ways:
Both the exchange and the OCC have rules governing the behavior of traders.
Both federal and state regulatory authorities monitor activities.
a. The SEC regulates options markets in stocks, stock indices, currencies, and bonds at the federal level.
b. The Commodity Futures Trading Commission regulates markets for options on futures.
c. In Illinois and New York (major options markets), state laws enforce acceptable trading practices.
In general, options markets have demonstrated a willingness to regulate themselves, as no major scandals or defaults
by OCC members have occurred.
8.11 Taxation
For tax purposes, a gain or loss is recognized when:
a. the option expires unexercised, or
b. the option is sold.
Wash Sale Rule
To prevent investors from selling stock at a loss and immediately repurchasing it so that a loss can be realized for tax
purposes, tax authorities have ruled that when the repurchase is within 30 days of the sale (i.e., between 30 days before
the sale and 30 days after the sale), any loss on the sale is not deductible.
Constructive Sales
Prior to 1997, a taxpayer selling a security short while holding a long position in a substantially identical security faced
no gain or loss until the short position was closed out.
This means that short positions could be used to defer recognition of a gain for tax purposes. As a result of the Tax
Relief Act of 1997, an appreciated property is now treated as “constructively sold” when the owner:
1. Enters into a short sale of the same or substantially identical property.
2. Enters into a futures or forward contract to deliver the same or substantially identical property; or
3. Enters into one or more positions that eliminate substantially all of the loss and opportunity for gain.
Warrants:
often come into existence as a result of a bond issue.
are added to the bond issue to make it more attractive to investors.
Executive stock options:
are issued to motivate company personnel to act in the best interests of the company’s shareholders.
are usually at–the money when first issued.
cannot be traded (unlike warrants and exchange–traded stock options).
Convertible bonds
can be considered an embedded call option on the company’s stock.
can be converted into equity at certain times using a predetermined exchange ratio.
lead to more shares being issued by the company when they are exercised.
Section Description
9.0 Introduction
9.1 Factors Affecting Option Prices
9.2 Assumptions and Notation
9.3 Upper and Lower Bounds for Option Prices
9.4 Put–Call Parity
9.5 Early Exercise: Calls on a Non–Dividend Paying Stock
9.6 Early Exercise: Puts on a Non–Dividend Paying Stock
9.7 Effect of Dividends
9.0 Introduction
This chapter covers stock option properties.
Different arbitrage arguments are used to explore the relationships between European option prices, American option
prices, and the underlying stock price.
The most important relationship is put–call parity, which is a relationship between European call option prices and
European put option prices.
The chapter examines American options. The author’s conclusions are:
i. it is never optimal to exercise American call options on non–dividend–paying stocks before expiration.
ii. early exercise of an American put option on such a stock can be optimal.
The effect on the price of a stock option of increasing one variable while keeping all others fixed
Variable European Call European Put American Call American Put
Stock price + – + –
Strike price – + – +
Time to expiration ? ? + +
Volatility + + + +
Risk–free rate + – + –
Dividends – + – +
+ indicates that an increase in the variable causes the option price to increase
Figure 1: Effect of changes in stock price, strike price, and expiration date on option prices when S0 = 50,
K = 50, r = 5%, = 30%, and T = 1.
Figure 2: Changes in volatility & risk–free interest rate on option prices if S0 = 50, K = 50, r = 5%, = 30%, & T = 1.
3. Time to Expiration
A. American options:
Both put and call options become more valuable as the time to expiration increases. Since the owner of a
longer–life option has all the exercise opportunities available to the owner of a shorter–life option (and more),
an increase in time to expiration causes the option price to increase.
B. European options:
Put and call options do not necessarily become more valuable as the time to expiration increases.
1. As stated earlier, European options can only be exercised at the maturity.
2. Consider two European call options on a stock: one with an expiration date in 1 month, and the other with an
expiration date in 2 months. Suppose a large dividend is expected in 6 weeks. Since the dividend will cause the stock
price to decline, the short–life option will be worth more than the long–life option.
4. Volatility
The owner of a call benefits from price increases, and is limited in downside risk to the price of the option.
The owner of a put benefits from price decreases and has limited downside risk in the event of price increases.
Therefore, the values of both calls and puts increase as volatility increases.
The cash amount is invested at the risk–free interest rate and it will grow to K at time T.
Q. What is Portfolio A and B worth at time T?
If ST > K, the call option is exercised at time T and portfolio A is worth ST
If ST < K, the call option expires worthless and the portfolio is worth K.
A. At time T, portfolio A is worth max (ST, K), and portfolio B is worth ST
Thus, at time T, Portfolio A is always worth at least as much as (sometimes more than) portfolio B.
At time 0, this can be represented as c + Ke–rT S0 or c S0 – Ke–rT
The worst case scenario for a call option is for it to expire worthless. Since its value cannot be negative, the formula
for the lower bound on a European call option is: c max(S0 – Ke–rT , 0)
Example: Let S0 = 51, K = 50, T = 0.5, and r = 0.12.
Since the price of a European call option on a non–dividend–paying is S0 – Ke–rT, the lower bound equals
51 – 50×e–0.12×0.50 = 3.91
3. Lower Bound for European Puts on Non–Dividend–Paying Stocks
A lower bound for the a European put option on a non–dividend–paying stock is Ke–rT – S0
Example 1: Let S0 = 37, K = 40, T = 0.5, and r = 0.05.
The lower bound is 40×e–0.05×0.50– 37 = 2.01
Example 2: Consider p = $1.00 (which is less than its theoretical minimum of $2.01). An arbitrager can
Borrow $38.00 for six months to buy both the put ($1) and the stock ($37).
Repay 38×e0.05×0.50 = 38.96. At the end of the six months,
If the stock price < $40.00, exercise the option to sell the stock for $40.00, repay the loan. This results in a profit of
$40.00 – $38.96 = $1.04
If the stock price is > $40.00, (say $42.00), discard the option, sell the stock, and repay the loan. This results in a profit
of $42.00 – $38.96 = $3.04
For a more formal argument, consider the following two portfolios:
Portfolio C: one European put option plus one share
Portfolio D: A cash amount equal to Ke–rT
Q. What are Portfolios C and D worth at time T?
If ST < K, the option is exercised, and the portfolio is worth K.
If ST > K, the put option is worthless and portfolio C is worth ST
A. At time T portfolio C is worth max (ST, K) and portfolio D is worth K.
At time T, portfolio C is always worth at least as much as (sometimes more than) portfolio D.
At time 0, this can be represented as p + S0 Ke–rT or p Ke–rT– S0
The worst case scenario for a put option is for it to expire worthless. Since its value cannot be negative, the formula for
the lower bound on a European put option is
p max(Ke–rT– S0, 0)
Example: Assume S0 = 38, K = 40, T = 0.25, and r = 0.10.
Compute the lower bound for a European put option on a non–dividend–paying stock.
Step 1: Write a formula for the price European put option on a non–dividend–paying stock: Ke–rT– S0
Step 2: Determine the lower bound based on the formula p max(Ke–rT– S0, 0).
Thus, the lower bound equals 40×e–0.10×0.25– 38 = 1.01
Figure 5 shows the variation of the European put price with the stock price.
Note that point B in Figure 5, at which the price of the option is equal to its intrinsic value, must represent a higher
value of the stock price than point A in Figure 4.
Point E in Figure 5 is where S0 = 0 and the European put price is Ke–rT
Figure 5: Variation of price of a European put option with the stock price, S0.
S0 – D – K C – P S0 – D – Ke–rT
Section Description
10.0 Introduction
10.1 Strategies Involving a Single Option and a Stock
10.2 Spreads
10.3 Combinations
10.4 Other Payoffs
10.0 Introduction
This chapter covers the range of profit patterns obtainable using options.
Although the underlying asset used is a stock, similar patterns can be obtained for other underlying assets such as
foreign currencies, stock indices, and futures contracts.
The key attraction to options is their use in creating a very wide range of payoff patterns.
In the figures below, profit will be shown as the final payoff minus the initial cost (note: it should be calculated as the
present value of the final payoff minus the initial cost).
10.2 Spreads
A spread trading strategy:
involves taking a position in two or more options of the same type (i.e., two or more calls or two or more puts).
includes Bull Spreads, Bear Spreads, Box Spreads, Butterfly Spreads, Calendar Spreads, and Diagonal Spreads:
Total payoff is computed as the sum of the payoffs on each option and varies dependent upon the stock’s price when
the option is exercised.
Profit is calculated by subtracting the initial investment from the total payoff.
A bull spread strategy limits the trader’s upside, as well as downside, risk. Three types of bull spreads are:
1. Both calls are initially out of the money.
a. This is the most aggressive type.
b. Setup cost is small, and there is a small probability of receiving high payoff (K2 – K1).
2. One call is initially in the money; the other call is initially out of the money.
3. Both calls are initially in the money.
Spreads become more conservative as one moves from type 1 to type 3.
Assume an investor buys a call for $3 with a strike price of $30 and sells for a call for $1 with a strike price of $35.
The cost of the strategy is $3 – $1 = $2.
Figure 1: Profit patterns (a) long position in a stock combined with short position in a call; (b) short position in a stock
combined with long position in a call; (c) long position in a put combined with long position in a stock; (d) short position
in a put combined with short position in a stock.
Bull spreads created by puts involve buying a put with a low strike price and selling a put with a high strike price.
Unlike bull spreads created from calls, bull spreads created from puts involve:
a positive cash flow to the trader up front, and
a payoff that is either negative or zero.
Figure 3: Profit from bull spread created using put options.
2. Bear Spreads: Buying a Call (with a higher strike price) and Selling a Call
An investor entering a bear spread is hoping that the stock price will fall.
A bear spread is created by buying a call with one strike price and selling a call with another strike price. However,
the strike price of the option purchased is greater than the strike price of the option sold.
It involves an initial cash inflow (since the price of the call sold is greater than the price of the call purchased).
Box Spreads
A box spread is a combination of a bull call spread with strike prices K1 and K2 and a bear put spread with the same
two strike prices.
Table 3 shows that the payoff from a box spread is always K2 – K1.
The value of a box spread is always the present value of this payoff or (K2 – K1)e–rT.
If it has a different value there is an arbitrage opportunity.
If the market price of the box spread is too low, it is profitable to buy the box; Buy a call with strike price K1, buy a put
with strike price K2, sell a call with strike price K2, and sell a put with strike price K1.
If the market price of the box spread is too high, it is profitable to sell the box. Buy a call with strike price K2, buy a
put with strike price K1, sell a call with strike price K1, and sell a put with strike price K2.
A box–spread arbitrage only works with European options.
Table 3: Payoff from a box spread.
Stock Price Range Payoff From bull call spread Payoff From bear put spread Total Payoff
ST K2 K2 – K1 0 K2 – K1
K1 < ST < K2 ST – K1 K2 – ST K2 – K1
ST K1 0 K2 – K1 K2 – K1
3. Butterfly Spreads involve positions in options with three different strike prices:
i. buying a call option with a high strike price (K3) and a call with a low strike price (K1)
ii. selling 2 call options with a strike price K2 (close to stock price and halfway between K1 and K3)
A butterfly spread:
leads to a profit if the stock price stays close to K2
gives rise to a small loss if there is a significant stock price move in either direction.
Payoffs from a butterfly spread.
Stock Price Range Payoff From First Payoff From Second Payoff From Total
Long Call Long Call Short Calls
ST < K1 0 0 0 0
K1 < ST < K2 ST – K1 0 0 ST – K1
K2 < ST < K3 ST – K1 0 –2 (ST – K2) K3 – ST
ST > K3 ST – K1 ST – K3 –2 (ST – K2) 0
Assume a stock is currently worth $61 (and a significant price move in the next 6 months is unlikely).
4. Calendar Spreads
This strategy involves options having the same strike price but different expiration dates.
A calendar spread is created by selling a call option and buying a longer–maturity call option with the same strike
price. It requires an initial investment.
A profit is made if the stock price at the expiration of the short–maturity option is close to the strike price of the short–
maturity option.
A loss is incurred if the stock price is significantly above or significantly below this strike price.
Types of calendar spreads (and relative strike price choices) are as follows:
A Neutral calendar spread (chosen with a strike price close to the current stock price).
A Bullish calendar spread (involves the use of a higher strike price).
A Bearish calendar spread (involves the use of a lower strike price).
Calendar spreads can be created using put options when an investor buys a long–maturity put option and sell a short–
maturity put option.
5. Diagonal Spreads
Bull, bear, and calendar spreads can all be created from a long position in one call (put) and a short position in another
call (put). In the case of bull and bear spreads, the calls have different strike prices and the same expiration date. In the
case of calendar spreads, the calls have the same strike price and different expiration dates. A diagonal spread requires
that both the expiration dates and the strike prices of the calls (puts) be different. This increases the range of profit
patterns that are possible.
10.3 Combinations
Combination strategies involve taking a position in both calls and puts on the same stock.
Four types of combinations are Straddles, Strips, Straps, and Strangles.
1. The Straddle:
The straddle involves buying a call and a put with the same strike price and expiration date.
If the stock and strike price are close at the option's expiration, a loss ensues.
If there is a large price swing in either direction, a significant profit will result.
A straddle is appropriate when there is an expectation of a large move in a stock price but the direction is not known.
Payoffs from a straddle.
Stock Price Range Payoff From Call Payoff From Put Total Payoff
ST < K 0 K – ST K – ST
ST > K ST – K 0 ST – K
A top straddle or straddle write is the reverse position. It is created by selling a call and a put with the same exercise
price and expiration date. If the stock price on the expiration date is close to the strike price, a significant profit results.
2 and 3. Strips and Straps
A strip: A long position in one call and two puts with the same strike price and expiration date. The bet is that a big
stock price move occurs, but a decrease in stock price is considered more likely.
A strap: A long position in two calls and one put with the same strike price and expiration date. The bet is that a big
stock price move will occur, but a increase in stock price is considered more likely.
Profit patterns from a Strip (a) and a Strap (b)
K1 K2 K3 ST
11.0 Introduction
The construction of a binomial tree is a useful technique to price a stock option. The binomial tree diagrams the
different paths a stock price might take over the life of an option.
(r – q)t e ( r q ) t d
This means that pS0u + (1 – p)S0 d = S0e so that p
ud
As in the case of options on non–dividend–paying stocks, match volatility by setting u e t and d 1/ u.
Options on Stock Indices
It is assumed that the valuation of an option on a stock index is similar to the valuation of an option on a stock paying a
known dividend yield.
Example: Consider the following:
A Two–step tree to value a European 6–month call option on an index when the index level is 810, strike price is 800,
risk–free rate is 5%, volatility is 20%, and dividend yield is 2%.
At each node:
Upper value = Underlying Asset Price
Lower value = Option Price
Strike price = 800
Discount factor per step = 0.9876
Time step, dt = 0.2500 years, 91.25 days
Growth factor per step, a = 1.0075
Probability of up move, p = 0.5126
Up step size, u = 1.1052
Down step size, d = 0.9048
In this case.
t 0.25, u e0.20 0.25
1.1052,
(0.05-0.02)0.25
d 1/ u 0.9048, ae 1.0075
p (1.0075 0.9048) /(1.1052 0.9048) 0.5126
The value of the option is 53.39.
Options on Currencies
As discussed in Section 5.10, a foreign currency can be regarded as an asset providing a yield at the foreign risk–free
rate of interest, rf. By analogy with the stock index case we can construct a tree for options on a currency by using
ad ( r r ) t
u e t and d 1/ u and p and setting a e f
ud
Example: Consider the following:
The Australian dollar is worth 0.6100 U.S. dollars and this exchange rate has a volatility of 12%.
The Australian risk–free rate is 7% and the U.S. risk–free rate is 5%.
A three–step tree to value an American 3–month call option on a currency when the value of the currency is 0.6100,
strike price is 0.6000, risk–free rate is 5%, volatility is 12%, and foreign risk–free rate is 7%.
At each node:
Upper value = Underlying Asset Price
Lower value = Option Price
Strike price = 0.6
Discount factor per step = 0.9958
Time step, dt = 0.0833 years, 30.42 days
Growth factor per step, a = 0.9983
Probability of up move, p = 0.4673
Up step size, u = 1.0352
Down step size, d = 0.9660
Section Description
12.0 Introduction
12.1 The Markov Property
12.2 Continuous–Time Stochastic and Stock Processes
12.3 The Process for A Stock Price
12.4 The Parameters
12.5 Itô’s Lemma
12.6 The Lognormal Property
12.0 Introduction
A Stochastic Process:
A stochastic process is one in which the value of a variable changes over time in an uncertain way.
A stochastic process can be classified as discrete time or continuous time.
1. For a discrete–time stochastic process, the value of the variable can only change at certain fixed times.
2. For a continuous–time stochastic process, changes can take place at any time.
A stochastic variable can be classified as continuous variable or discrete variable.
1. For a continuous–variable process, the underlying variable can take any value within a certain range.
2. For a discrete–variable process, only certain discrete values are possible.
At the end of year 1, the value of the variable is normally distributed with a mean of 25 and a standard deviation of 1.0.
At the end year 5, it is normally distributed with a mean of 25 and a standard deviation of 5 2.236
Based on the above, we see that the uncertainty about the value of the variable increases as the square root of how far
we are looking ahead.
Two noteworthy properties of Wiener processes, related to the t property, are:
The expected length of the path followed by z in any time interval is infinite.
The expected number of times z equals any particular value in any time interval is infinite.
2. A Generalized Wiener Process
A generalized Wiener process for a variable x can be defined in terms of dz as follows: dx = adt + bdz, where a and b
are constants.
The change in the value of x in any time interval T is normally distributed with:
mean of change in x = aT, standard deviation of change in x b T , and variance of change in x = b2T
The generalized Wiener process has an expected drift rate (i.e., average drift per unit of time) of a and a variance rate
(i.e., variance per unit of time) of b2.
Example: Assume that the cash position of a company, initially 50 (in 000s), follows a generalized Wiener process
with a drift of 20 per year and a variance rate of 900 per year.
At the end of year 1, the cash position has a normal distribution with mean of 70 [50 + (20 × 1)] and a standard
deviation of 30 ( 900 ). At the end of 6 months, it has a normal distribution with a mean of 60 [50 + (20 × 0.50)] and
a standard deviation of 30 0.5 21.21
Notes:
The uncertainty about the cash position at some time in the future increases as the square root of how far ahead we are
looking.
The cash position can become negative (we can interpret this as a situation where the company is borrowing funds).
3. An Ito Process:
An Ito process is a generalized Wiener process where the parameters a and b are functions of the value of the
underlying variable, x, and time, t. An Ito process can be written dx = a (x, t) dt + b (x, t) dz. An Ito process assumes
that the drift and variance rate of x are equal to a(x, t) and b(x, t)2 during the time between t and t + t.
Section Description
13.0 Introduction
13.1 Lognormal Property of Stock Prices
13.2 The Distribution of the Rate of Return
13.3 The Expected Return
13.4 Volatility
13.5 Concepts Underlying the Black–Scholes–Merton Differential Equation
13.6 Derivation of the Black–Scholes–Merton Differential Equation
13.7 Risk–Neutral Valuation
13.8 Black–Scholes Pricing Formulas
13.9 Cumulative Normal Distribution Function
13.10 Warrants and Executive Stock Options
13.11 Implied Volatilities
13.12 Dividends
13.0 Introduction
Topics:
a. Derivation of the Black–Scholes (BS) model for valuing European call and put options on a non–dividend–paying
stock, and how the BS model deals with dividend–paying stocks.
b. Volatility either estimated from historical data or implied from option prices using the model.
c. Risk–neutral valuation argument introduced in Chapter 11 can be used.
d. Pricing of American call options on dividend–paying stocks is determined.
Example: Consider a stock with a current price of $20, an expected return of 20% per annum, and volatility of 40% per
annum. Thus, S0 = 20; = 0.20; = 0.40; T = 1.0 Then:
E ( ST ) 20e0.21 24.43 , Var ( ST ) 400e 20.21[e0.4
2 1
1] 103.54 , and the 103.54 10.18
13.4 Volatility
When T is small, T is the standard deviation of the proportional change in the stock price in time T .
Example: Assume = 0.30 (30% per annum) and the current stock price is $50.
The standard deviation of the proportional change in the stock price in one week is 0.30 1 0.0416
52
A one standard deviation move in the stock price in one week is $50 × 0.0416 = $2.08.
The standard deviation of the stock price in four weeks is ~ twice the standard deviation in one week.
Estimating Volatility from Historical Data
To estimate the volatility of a stock price empirically, we first define the following: Define:
n + 1: Number of observations
Si: Stock price at end of ith interval (i = 0, 1,…, n)
: Length of time interval in years
Si
Let, ui ln .
Si 1
The standard deviation of the ui is s = ; can be estimated as * where * s with a standard error of
*
2n
Note: Empirical research indicates that time should be measured in trading days (days when the exchange is closed
should be ignored).
Example (see the data in table 13.1 on page 287). Table 13.1 shows a possible sequence of stock prices during 21
consecutive trading days. In this situation,
ui 0.09531 and
ui2 0.00326
2
an estimate of the standard deviation of the daily return is 0.00326 0.09351 0.01216
19 380
an estimate for the volatility per annum is * 0.01216 0.193 (based on 252 trading days).
1/252
the standard error of this estimate is * 0.193 0.031 per annum.
2n 220
This analysis can be adapted to accommodate dividend–paying stocks:
S D
The return during a time interval that includes an ex–dividend day is given by ui ln i , where D is the amount
Si 1
Si
of the dividend. The return in other time intervals is ui ln
Si 1
Trading Days vs. Calendar Days
Practitioners tend to ignore days when the exchange is closed when estimating volatility from historical data and when
calculating the life of an option. The volatility per annum is calculated from the volatility per trading day using the
formula:
volatility per annum = volatility per trading day number of trading days per annum .
It is assumed that there are 252 trading days for stocks per year.
The life of an option is also usually measured using trading days rather than calendar days. It is calculated as T years,
trading days until option maturity
where T
252
13.12 Dividends
The BSM can be modified to take account of dividends. It is assumed that the amount and timing of the dividends can
be predicted with certainty.
A dividend–paying stock can be expected to follow the stochastic process developed in Chapter 11 except when the
stock goes ex–dividend. At this point, the stock’s price goes down by an amount reflecting the dividend paid per share.
Notes:
For tax reasons, the stock price may decrease by a smaller amount than the cash amount of the dividend.
“Dividend” should be interpreted as the reduction in the stock price on the ex–dividend date caused by the
dividend.
Example: If a dividend of $1 per share is anticipated, and the share decreases in price by 80% of the dividend on the
ex–dividend date, the dividend should be assumed to be $0.80.
European Options:
European options can be analyzed by assuming that the stock price is the sum of 2 components:
1. a riskless component (the present value of the dividends during the life of the option discounted from the ex–
dividend dates to the present at the risk–free rate).
2. a risky component.
The BS formula is correct if S0 is equal to the risky component of the stock price and , is the volatility of the process
followed by the risky component.
European options can be used with the BSM formula provided that the stock price is reduced by the present value of
all the dividends during the life of the option.
Example: Consider a European call option on a stock paying dividends of $0.50 and having ex–dividend dates in 2
months and 5 months. Let S0 = 40; K = 40; = 0.30; T = 0.5; r = 0.09.
The present value of the dividends is 0.50e 0.16670.09 0.50e 0.41670.09 0.9741
The option price using the BSM and the following inputs [S0 = 40 – 0.9741 = 39.0259; K = 40; = 0.30; T = 0.5; r =
0.09 is shown below:
ln(39.0259/40)(0.09 0.32 /2)0.50
d1 0.2017 ; d2 d1 T 0.2017 0.3 0.5 0.0104
0.3 0.5
N(d1) = 0.5800; N(d2) = 0.4959
Using the equation c = S0N(d1) – Ke–rT N(d2), the call price is 39.0259 × 0.5800 – 40e–0.09×0.5 × 0.4959 = 3.67
American Options
In section 9.5, Hull states that in the absence of dividends, American options should never be exercised early.
In this section, Hull claims that when stocks pay dividends, it is optimal to exercise only at a time immediately before
the stock goes ex dividend. For example, assume that:
i. n ex–dividend dates are anticipated and that t1, t2, …tn, are moments in time immediately prior to the stock going
ex–dividend, with t1 < t2 < t3 < ... < tn.
ii. the dividends corresponding to these times are D1, D2,…, Dn respectively.
Consider the early exercise of an American call option just prior to the final ex–dividend date.
i. If the option is exercised at tn, the investor receives S(Tn) – K
ii. If the option is not exercised at tn, the stock price drops to S(Tn) – Dn
Hull goes on to show that:
r (T tn )
i. If Dn K (1 e ) , it is not optimal to exercise at tn.
r (T tn )
ii. If Dn K (1 e ) , it is optimal to exercise at tn.
r ( ti 1 ti )
iii. If Di K (1 e ) , for any i < n , it is not optimal to exercise at ti.
Hull’s final concluding statements in this section are as follows:
i. The only time that needs to be considered for the early exercise of an American call is the final ex–dividend date,
tn.
r ( t t ) r ( T tn )
ii. If the inequality Di K (1 e i1 i ) holds for i = 1, 2, ... , n – 1 and inequality Dn K (1 e ) holds, then
early exercise is never optimal.
Black’s Approximation
Black’s approximation procedure for taking account of early exercise in call options involves:
i. Calculating the prices of European options that mature at times T and tn, and then
ii. Setting the American price equal to the greater of the two.
Example: Consider the prior example of a European call option on a stock paying $0.50 and having ex–dividend dates
in 2 months and 5 months. Thus, S0 40, K 40, 0.30, T 0.5, r 0.09
Suppose the option is American rather than European. In this case, D1 = D2 = 40; S0 = 40; K = 40; = 0.20; t1 = 1/12;
and t2 = 5/12; r = 0.10.
Between t1 and t2, K (1 e r (t2 t1 ) ) 40(1 e 0.09(0.25) ) 0.89 which is greater than 0.50 and thus, the option should
not be exercised immediately before the first ex–dividend date.
Between T and t2, K (1 e r (T t2 ) ) 40(1 e 0.09(6 /125 /12) ) 0.30 which is less than 0.50 and thus, the option should
be exercised immediately before the second ex–dividend date.
Now use Black’s approximation to value the option.
i. The present value of the first dividend is 0.50e 0.16670.09 0.4926
ii. The option price using the BSM and the following inputs
[S0 = 40 – 0.4926 = 39.5074; K = 40; = 0.30; T = 5/12 = 0.41667; r = 0.09, is $3.52.
iii. Black’s approximation involves taking the greater of $3.52 and the value of the option when it can only be
exercised at the end of six months ($3.67). $3.67 is the value of the American call option.
Section Description
14.1 Results for a Stock Paying a Known Dividend
14.2 Option Pricing Formulas
14.3 Options on Stock Indices
14.4 Currency Options
14.5 Futures Options
14.6 Valuation of Futures Options using Binomial Trees
14.7 Futures Price Analogy
14.8 Black’s Model for Valuing Futures Options
14.9 Futures Options vs. Spot Options
Binomial Trees
Here, Hull examines the effect of a dividend yield equal to q on the binomial model in Chapter 11.
i. Once again, the total return provided by a stock in a risk–neutral world must be r.
ii. Since dividends provide a return equal to q, the return in the form of capital gains must be r – q.
Thus p, the probability of an up movement in a risk–neutral world must satisfy pSu (1 p ) Sd e( r q ) t or
e( r q ) t d
p
ud
Relationship between value of index and value of portfolio for beta = 2.0
Value of index Value of portfolio
in three months in three months ($)
1,080 570,000
1,040 530,000
1,000 490,000
960 450,000
920 410,000
880 370,000
Suppose a portfolio manager is looking to ensure that the value of the portfolio does not drop below $450,000.
Based on the table above, the appropriate strike price for the put options purchased is 960.
Since 100S0 = $100,000 and beta = 2.0, two put contracts are required for each $100,000 in the portfolio. Since the
portfolio is worth $500,000, 10 contracts should be purchased.
Note: If the value of the index falls to 880, the table above indicates that the value of the portfolio is 370,000.
However, the put options pay off (960 – 880) × 10 × 100 = $80,000, which restores the portfolio to the insured level.
Note: There are two reasons why the cost of hedging increases as the beta of a portfolio increases:
1. more put options are required, and
2. the options have a higher strike price.
i. A company knowing it will receive sterling at a certain time in the future can hedge its risk by buying put options
on sterling that mature at that time. The benefits to a company using this strategy are:
a. a guaranty that the value of the sterling will not be less than the exercise price
b. rendering any favorable exchange–rate movements to the company.
ii. A company knowing that it will pay sterling at a certain time in the future can hedge by buying calls on sterling
that mature at that time. The benefits to a company employing this strategy are:
a. a guaranty that the cost of the sterling will not exceed a certain amount
b. rendering favorable exchange–rate movements to the company.
Q. What is the difference when a forward contract is used to hedge vs. an option approach to hedge?
A1. A forward contract locks in the exchange rate for a future transaction, while an option provides a type of
insurance.
A2. A forward transaction costs nothing to enter into. An options require a premium to be paid up front.
Valuation
Define S0 as the spot exchange rate (the value of one unit of the foreign currency measured in the domestic currency).
rf is the foreign risk–free interest rate, and is the yield received by the owner of foreign currency.
Therefore,
i. the European call price, c, and put price, p, are therefore given by
r T r T
c S0 e f N ( d1 ) Ke rT N ( d 2 ); p Ke rT N ( d 2 ) S0 e f N ( d1 )
where S0 is the value of the exchange rate at time zero, and
In( S0 / K ) ( r rf 2 / 2)T In( S0 / K ) ( r rf 2 / 2)T
d1 ; d2 d1 T
T T
ii. both the domestic interest rate, r, and the foreign interest rate, rf , are the rates for maturity T.
Note: Put and call options on a currency are symmetrical in that
i. a put option to sell XA units of currency A for XB units of currency B is the same as
ii. a call option to buy XB units of currency B for XA units of currency A.
Example. Consider a 4 month European call option on the British pound.
i. Suppose the current exchange rate is 1.600, the strike price 1.600, and the risk free rate in the U.S. is 8% per
annum, the risk–free interest rate in Britain is 11 % per annum, and the option price is 4.3 cents.
ii. Thus, S0 = 1.6, K = 1.6, r = 0.08, rf = 0.11, T = 4/12, and c = 0.043.
Calculate the implied volatility.
Note: The implied volatility can be calculated iteratively.
i. A volatility of 20% gives an option price of 0.0639
ii. A volatility of 10% gives an option price of 0.0285.
iii. The iterative procedure ultimately results in an implied volatility of 14.1%.
Example: Suppose it is 8/15 and an investor has 1 Sept. futures call option contract on copper with a strike price of 70
cents/ pound. 1 futures contract is on 25,000 pounds of copper.
Assume that the futures price of copper for delivery in Sept is 81 cents, and at the close of 8/14 trading it was 80 cents.
If the option is exercised, the investor receives:
Cash = 25,000 × (80 – 70) cents = $2,500 + a long position to buy 25,000 pounds of copper in Sept.
i. If the position in the futures contract was closed out immediately, the investor would have a $2,500 cash payoff +
25,000 × (81 – 80) cents = $250
ii. The total payoff from exercising the option on 8/15 = $2,750 = 25,000(F – K), where F is the futures price at the
time of exercise and K is the strike price.
Example: An investor has one Dec. futures put option on corn with a strike price of 200 cents per bushel.
One futures contract is on 5,000 bushels of corn. Assume the current futures price of corn for delivery in Dec. = 180
cents, and the most recent settlement price is 179 cents.
If the option is exercised, the investor receives 5,000 × (200 – 179) cents = $1,050 + a short position in a futures
contract to sell 5,000 bushels of corn in Dec.
If the position in the contract is closed, the investor has 1,050 cash payoff – 5,000 × (180 – 179) cents (i.e. $50). The
net payoff from exercise = $1,000 = 5,000(K – F), where F is the futures price at the time of exercise and K is the
strike price.
Quotes
Most futures options are American, and are referred to by the month in which the underlying futures contract matures
(not by the expiration month of the option).
The maturity date of a futures option contract is usually on (or a few days before), the earliest delivery date of the
underlying futures contract.
Options on Interest Rate Futures
The most actively traded futures options in the U.S. are those on T–bond futures, T–note futures, and Eurodollar
futures.
i. A T–bond futures option is an option to enter a T–bond futures contract. one T–bond futures contract is for the
delivery of Treasury bonds with a $100,000 face value. The price of a T–bond futures option is quoted as a percent
of the face value of the underlying T–bonds to the nearest 1/64 of 1%.
The table below gives the price of the April call futures option on Treasury bonds as 2–11 (2 11/64%) of the debt
principal when the strike price is 104 (implying that one contract costs $2,171.87). Quotes for options on T–notes
are similar.
ii. An option on Eurodollar futures is an option to enter into a Eurodollar futures contract. Recall from chapter 5,
when the Eurodollar futures quote changes by 1 basis point (0.01), there is a gain or loss on a Eurodollar futures
contract of $25. Similarly, in the pricing of options on Eurodollar futures, 1 bp = $25.
In the shortest maturity contract, prices are quoted to the nearest ¼ of a bp. For the next 2 mos., they are quoted to
the nearest ½ bp. The WSJ quote for the CME Eurodollar futures contract shown below should be multiplied by 10
to get the CME quote in bp. For example, the 5.92 quote for the CME March call futures option when the strike
price is 94.50 in indicates that the CME quote is 59.25 bp. and 1 contract costs 59.25 × $25 = $1,481.25.
iii. Interest rate futures contracts work in the same way as other futures contracts.
The payoff from a call is max (F – K, 0), where F is the futures price at the time of exercise and K is the strike price.
Further, the option holder obtains a long position in the futures contract at exercise and the option writer obtains a
short position (in addition to the cash payoff).
Since interest rate futures prices increase when bond prices increase (i.e., when rates fall), and vice versa, investors
who thinks that short–term interest rates will rise speculate by buying put options on Eurodollar futures (an investor
who thinks rates will fall speculates by buying call options on Eurodollar futures).
An investor who thinks that long–term interest rates will rise speculates by buying put options on T– note futures or T–
bond futures, and an investor who thinks they will fall can speculate by buying call
options on these securities.
Example: Suppose that it is Feb and the futures price for the Jun Eurodollar contract is
33 93.82 (which corresponds to a 3 mo. Eurodollar interest rate of 6.18% per annum).
The price of a call option on the contract with a strike price of 94.00 is quoted as 0.20.
The option is attractive to an investor who feels that interest rates are likely to fall.
Assume short–term interest rates fall about 100 bps over the next 3 mos., and
the investor exercises the call when the Eurodollar futures price is 94.78 (which
corresponds to a 3 mo Eurodollar interest rate of 5.22% per annum.)
The payoff = 25 × (94.78–94) = $1,950. The cost of the contract = 20 × 25 = $500.
The profit = $1,450.
Example: Suppose that it is Aug and the futures price for the Dec. T–bond contract is 96–09 (or 96 9/32 = 96.28125).
The yield on long–term government bonds is about 6.4% per annum.
If an investor feels that the yield will fall by Dec., buy Dec. calls with a strike price of 98. Assume that the price of
these calls is 1–04 (1 4/64 = 1.0625% of the principal).
If long–term rates fall to 6% per annum, and the T–bond futures price rises to 100–00, the net profit per $100 of bond
futures = 100.00 – 98.00 – 1.0625 = 0.9375
Since 1 option contract is for the purchase/sale of instruments with a face value of $100,000, the profit = $937.50 per
option contract purchased.
Reasons for the Popularity of Futures Options (vs options on the underlying asset).
1. The main reason is that a futures contract is (generally) more liquid and easier to trade.
2. A futures price is known immediately from trading on the futures exchange (vs. the spot price not being so readily
available).
Example: T–bonds.
i. The market for T–bond futures is more active than the market for any particular T–bond.
ii. A T–bond futures price is known immediately from trading (vs. the current market price of a bond known only by
contacting one or more dealers.
iii. Investors would rather take delivery of a T–bond futures contract than T–bonds.
For commodities futures, it is easier and more convenient to make or take delivery (e.g. of a live–hogs futures
contract than it is to make or take delivery of hogs).
iv. Exercising a futures option usually does not lead to delivery of the underlying asset, because the underlying
futures contract is closed out prior to delivery.
v. Futures options are eventually settled in cash, which is appealing to many investors (with limited capital) who may
find it difficult to come up with the funds to buy the underlying asset when an option is exercised.
vi. Futures and futures options are traded in pits side by side in the same exchange, which aids hedging, arbitrage, and
speculation, making the markets more efficient.
vii. Futures options tend to entail lower transactions costs than spot options in many situations.
Put–Call Parity for European futures options
Assumption: there is no difference between the payoffs from futures and forward contracts. See Chapter 8 for the
derivation of the put–call parity relationship for European stock options.
Consider European call and put futures options, both with strike price K and time to expiration T.
Form two portfolios:
Portfolio A: a European call futures option + cash equal to Ke–rT. In this case, cash is invested at the risk–free rate, r,
and grows to K at time T. Let FT be the futures price at maturity of the option.
If FT > K, the call option is exercised and the portfolio is worth FT.
If FT < K, the call is not exercised and portfolio A is worth K.
The value of portfolio A at time T is therefore the max (FT , K)
Portfolio B: a European put futures option + a long futures contract + cash equal to F0e–rT. In this case, cash is invested
at the risk–free rate and grows to F0 at time T.
The put option has a payoff of max(K – FT , 0). The futures contract has a payoff of FT – F0.
Further, for all future prices, the expected growth rate in a futures price in a risk–neutral world is zero.
This applies in the world where interest rates are stochastic as well as where they are constant.
i. ( F ) , where
Because the expected growth rate of the futures price is zero, F0 E T
FT is the futures price at the maturity of the contract, and
F0 is the futures price at time zero, and
denotes the expected value in a risk–neutral world.
E
ii. Because FT = ST, where ST is the spot price at time T, F0 E (F )
T
Thus, for all assets the futures price equals the expected future spot price in a risk–neutral world.
It is not generally true that an American futures option equals the corresponding American spot option when the
futures and options contracts have the same maturity.
Example: Suppose a market exists with futures prices consistently higher than spot prices prior to maturity (e.g. as is
the case with most stock indices, gold, silver, low–interest currencies, etc.)
An American call futures option must be worth more than the corresponding American spot call option. In some
situations the futures option will be exercised early, which will provide a greater profit to the holder. Thus, an
American put futures option must be worth less than the corresponding American spot put option.
In an inverted market, with futures prices lower than spot prices, as is the case with high–interest currencies and
some commodities, the reverse must be true.
American call futures options are worth less than the corresponding American spot call option, whereas American put
futures options are worth more than the corresponding American spot put option.
The differences just described hold true when the futures contract expires later than the options contract as well as
when the two expire at the same time.
In fact, the differences tend to be greater the later the futures contract expires.
Section Description
15.1 The Details of Selling a Hypothetical Call Option
15.2 Naked and Covered Positions
15.3 A Stop–Loss Strategy
15.4 Delta Hedging
15.5 Theta
15.6 Gamma
15.7 Relationship among Delta, Theta, and Gamma
15.8 Vega
15.9 RHO
15.10 The Realities of Hedging
15.11 Scenario Analysis
15.12 Portfolio Insurance
15.13 Stock Market Volatility
Introduction
Financial institutions manage the risk they incur when selling options or other derivatives that have been tailored to
meet the needs of their clients. The exposure created cannot be neutralized by buying on an exchange the same option
as it has sold. Since the option has been tailored to the needs of a client, it does not correspond to the standard products
traded by exchanges. Thus, hedging the exposure is difficult.
Alternative approaches to solving this problem are discussed in this chapter. Topics covered include:
The “Greeks” measure a different dimension of the risk in an option position.
The creation of synthetic options. Options do not necessarily need to be purchased, but can be created
synthetically. This process is very closely related to the hedging of options.
A trader’s goal is to manage the Greeks so that all risks are acceptable.
Conclusions:
Neither a naked position nor a covered position provides a satisfactory hedge.
A perfect hedge would ensure that the cost is always $240,000 (the price computed using the Black–Scholes
formula).
Under a perfect hedge, the standard deviation of the cost of writing the option and hedging it is zero.
t1 t2 t3 t4 Time (t)
One might think that the total cost, Q, of writing and hedging the option is Q Max [ S0 K , 0] , since all purchases
and sales made, subsequent to time 0, are at price K. If this were correct, the hedging scheme would work perfectly in
the absence of transactions costs. Further, the cost of the option would always be less than its Black–Scholes price (the
Black–Scholes Model can be used to determine its theoretical price). Therefore, one could earn riskless profits by
writing options and hedging them.
Two reasons why equation Q Max [ S0 K , 0] is incorrect.
1. The cash flows to the hedger occur at different times and must be discounted.
2. Purchases and sales cannot be made at exactly the same price, K.
Since we cannot assume that both purchases and sales are made at the same price, purchases must be made at must
be made at K + and sales must be made at K – . Thus, every purchase and subsequent sale involves a cost
(apart from transaction costs) of 2 . Assuming stock prices change continuously, can be made arbitrarily
small, but this is offset by the increasing frequency of trading.
Conclusions
The stop–loss strategy does not work particularly well as a hedging scheme, particularly in the case of an
out–of–the–money option.
If the stock price never reaches the strike price, K , the hedging scheme costs nothing.
If the path of the stock price crosses the strike price level many times, the hedging scheme becomes quite
expensive.
To assess the overall performance of the scheme, Monte–Carlo simulation can be used.
The table below shows the performance of stop–loss strategy.
t (weeks) 5 4 2 1 0.5 0.25
Hedge performance 1.02 0.93 0.82 0.77 0.76 0.76
The hedge performance measure is the ratio of the standard deviation of the cost of writing the option
and hedging it to the Black–Scholes price of the option.
A performance measure below 0.70 was impossible to attain (regardless of how small t is).
Q: If the delta of a call option on a stock is 0.6, what does this mean?
A: When the stock prices changes by a small amount, the option price changes by about 60% of that amount.
A Delta Hedging Example:
Givens: (1) An investor has sold 20 call options contracts (each option contract controls 100 shares).
(2) The stock price is $100 and the option price is $10.
By doing nothing, the investor is subject to incurring a loss should the stock price rise above $100.
To “delta hedge” this position, buy 0.60 × 2,000 = 1,200 shares.
Why? Because the gain (loss) on the option position can be offset by the gain (loss) on the stock position.
For example, if the stock price rises by $1.00, the investor’s realizes (if the total position is terminated):
a gain of $1,200 ($1.00 × 1,200 shares) on the shares purchased, and a loss of $1,200 ($0.60 × 2,000 shares) on the
options written. The option prices rises and falls by its delta (in this case 0.60 × $1.00 = $0.60)
In summary:
The delta of the investor’s option position is 0.60 × (–2,000) = –1,200.
The delta of the investor’s long position is 1.0 × (1200) = 1,200 (since the delta of the stock is 1.0).
The delta of the investor’s overall position is = 0 (a zero delta position is a.k.a. “delta neutral”).
Rebalancing:
Since delta changes, the delta hedged (or delta neutral) position will remain in tact for only a short time period.
Therefore, the hedge must be adjusted or rebalanced.
For example, if the stock price rises to $110, then delta rises from 0.60 to 0.65. To maintain the hedge, an extra 100
shares must be purchased [0.05 (e.g. 0.65 – 0.60) × 2,000 shares = 100 shares].
Definitions:
i. a dynamic hedging scheme: one that requires the hedge position to be adjusted periodically.
ii. a static hedging scheme: one that once set up, is never adjusted (a.k.a. hedge–and–forget schemes).
Relationship to Black–Scholes analysis:
Black and Scholes showed that one can set up a riskless portfolio consisting of a position in a derivative on a stock
and a position in the stock. Essentially, they set up a delta–neutral position and argue that the return on the position
over a short period of time equals the risk–free interest rate.
Delta of European Calls and Puts:
1. For a European call option on a non–dividend paying stock:
ln( S0 / K ) ( r 2 / 2)T
a. The Black–Scholes formula shows that (call) = N(d1), where d1
T
b. Delta hedging for a short position in a European call option requires keeping a long position of N(d1) shares
at any given time.
c. Delta hedging for a long position in a European call option involves maintaining a short position of N(d1)
shares at any given time.
During the 9 week period, the overall wealth of the financial institution changed as follows:
Value of shares owned at week nine: $4,171,000 (78,700 × $53)
Cost of shares purchased at week zero – 2,557,800
Value of option at week nine – 414,500
Value of option at week zero + 240,000
Difference in cumulative cost (week 9 – week 0) –1,442,700 (4,000,500 – 2,557,800)
–$ 4,000
Delta of a Portfolio
is the delta of an options portfolio, dependent on a single asset whose price is S, where is the value of the
S
portfolio. The delta of the portfolio can be calculated from the deltas of the individual options in the portfolio. If a
portfolio consists of an amount, wi of option i (1 i n), the delta of the portfolio is given by:
15.5 Theta
Definition: The theta of a portfolio of derivatives, :
a. is the rate of change of the portfolio value with respect to the passage of time (or with respect to a decrease in the
times to maturity of the derivatives in the portfolio).
b. is also referred to as the time decay of the portfolio.
S N ( d1 )
A. For a European call option on a non–dividend–paying stock, 0 rKe rT N ( d 2 ) , where:
2 T
ln( S0 / K ) ( r 2 / 2)T 1 x2 /2
d1 , d 2 d1 T and N ( x ) e
T 2
S N ( d1 )
B. For a European put option on the stock, 0 rKe rT N ( d 2 )
2 T
C. For a European call option on a stock index paying a dividend at rate q,
S N ( d1 ) e qT ln( S0 / K ) ( r q 2 / 2)T
0 qS0 N ( d1 )e qT rKe rT N ( d 2 ) , where d1
2 T T
S N ( d1 ) e qT
D. For a European put option on the stock index 0 qS0 N ( d1 )e qT rKe rT N ( d 2 )
2 T
Notes:
With q equal to rf , equations C. and D. above give thetas for European call and put options on currencies.
With q equal to r and S0 equal to F0, equations C. and D. give thetas for European futures options.
In the formulas above, time is measured in years. However, theta is usually quoted in days. To obtain the theta
with time measured in days, the results of using the formulas must be divided by 252 (the number of trading
days in one year).
More about Theta:
a. Theta is usually negative for an option (since as time passes, the option becomes less valuable).
b. Theta is not the same type of hedge parameter as delta. Since there is no uncertainty about the passage of time, it
does not make any sense to hedge against the effect of the passage of time.
c. Theta is a proxy for gamma in a delta–neutral portfolio.
15.6 Gamma
Definition: The gamma, , of a portfolio of derivatives on an underlying asset:
a. is the rate of change of the portfolio's delta with respect to the price of the underlying asset.
b. is the second partial derivative of the portfolio value with respect to the asset price: 2 / S 2 .
Gamma measures the curvature of the relationship between the option price and the stock price.
Example:
When the stock price moves from S to S', delta hedging assumes that the option price moves from C to C’ when in
actual fact it moves from C to C”. The difference between C’ and C” leads to a hedging error. The error depends on
the curvature of the relationship between the option price and the stock price. Gamma measures this curvature.
C’’
C’
S S’
Recall that Theta is a proxy for gamma in a delta–neutral portfolio.
For a delta–neutral portfolio, t (1 / 2) S 2 (14.6).
Symbol Description
price change of the portfolio.
S price change of an underlying asset during a small interval of time, t.
Example: Suppose that the gamma of a delta–neutral portfolio of options on an asset is – 10,000. Equation (14.6)
shows that if a change of +2 or –2 in the price of the asset occurs over a short period of time, there is an unexpected
decrease in the value of the portfolio of approximately 0.5 × –10,000 × 22 = –$20,000.
The relationship between and S:
When gamma is positive, theta tends to be negative.
When gamma is negative, theta tends to be positive.
Making a Portfolio Gamma Neutral
A position in either the underlying asset itself or a forward contract on the underlying asset has zero gamma and cannot
be used to change the gamma of a portfolio.
To adjust the gamma of a portfolio, we must take a position in an instrument such as an option that is not linearly
dependent on the underlying asset.
Example: Suppose that a delta–neutral portfolio has a gamma equal to and a traded option has a gamma equal to
T.
If the number of traded options added to the portfolio is wT, the gamma of the portfolio becomes wT T . Hence,
the position in the traded option necessary to make the portfolio gamma neutral is , since T 0 .
T T
Example: Suppose that a portfolio is delta neutral and has a gamma of –3,000. The delta and gamma of a particular
traded call option are 0.62 and 1.50, respectively.
3,000
The portfolio can be made gamma neutral by including in the portfolio a long position of 2,000 in the
1.5
traded option.
However, the delta of the portfolio will then change from zero to 2,000 × 0.62 = 1,240. Thus, 1,240 of the
underlying asset must be sold from the portfolio to keep it delta neutral.
Notes:
Delta neutrality provides protection against relatively small stock price moves between rebalancing.
Gamma neutrality provides protection against larger movements in this stock price between hedge rebalancing.
Calculation of Gamma
N ( d1 )
The gamma of a European call or put option on a non–dividend–paying stock is , where
S0 T
ln( S0 / K ) ( r 2 / 2)T 1 x2 / 2
d1 , d 2 d1 T and N ( x ) e
T 2
For a European call or put option on a stock index paying a continuous dividend at rate q .
N (d1 )e qT
S0 T
Notes: This formula gives the gamma for:
a. a European option on a currency when q is put equal to the foreign risk–free rate, and
b. a European futures option with q = r and S 0 = F 0.
15.8 Vega
The vega of a portfolio of derivatives, V , is the rate of change of the value of the portfolio with respect to the
volatility of the underlying asset: V
If vega is high in absolute terms, the portfolio’s value is very sensitive to small changes in volatility (and vice
versa).
A position in the underlying asset or in a forward contract has zero vega.
The vega of a portfolio changes by adding a position in a traded option.
If V is the vega of the portfolio and VT is the vega of a traded option, a position of V / VT in the traded option
makes the portfolio instantaneously vega neutral.
A portfolio that is gamma neutral will not, in general, be vega neutral, and vice versa 0.
For a portfolio to be both gamma and vega neutral, at least two traded derivatives dependent on the underlying
asset must usually be used.
Example 6: You are given a portfolio that is delta neutral. In addition, you are given the following:
Symbol Description
5, 000 The gamma of the portfolio is –5,000
V 8, 000 The vega of the portfolio is –8,000
t1 0.5 The gamma of traded option 1 is 0.5
t 2 0.8 The gamma of traded option 2 is 0.8
Vt1 2.0 The vega of traded option 1 is 2.0
Vt 2 1.2 The vega of traded option 2 is 1.2
t1 0.6 The delta of traded option 1 is 0.6
t 2 0.5 The delta of traded option 2 is 0.5
wi The amount of traded option i.
V 8, 000
1. To make the portfolio vega neutral, take a long position of w1 4, 000 in traded options.
Vt1 2
2. To maintain delta neutrality, 2,400 units of the asset must be sold, since delta increases to 2,400.
Note: The gamma of the portfolio changes from –5,000 to –3,000.
3. To make the portfolio gamma and vega neutral, we need to determine the w1 and w2 to be included in the
portfolio:
–5,000 + 0.5w1 + 0.8w2 = 0
–8,000 + 2.0w1 + 1.2w2 = 0
The solution is w1 = 400, w2 = 6,000.
4. The delta of the portfolio after adding the positions in the two traded options is 400 × 0.6 + 6,000 × 0.5 =
3,240. Hence 3,240 units of the asset would have to be sold to maintain delta neutrality.
Formulas for Vega:
1. A European call or put option on a non–dividend–paying stock: V S0 T N ( d1 ) , where
ln( S0 / K ) ( r 2 / 2)T 1 x2 / 2
d1 , d 2 d1 T and N ( x ) e
T 2
2. A European call or put option on a stock or stock index paying a continuous dividend yield at rate q,
ln( S0 / K ) ( r q 2 / 2)T
V S0 T N ( d1 )e qT , where d1
T
The above equation gives the vega for:
a. a European currency option with q equal to rf
b. a European futures option with q equal to r and S0 equal to F0
Notes:
The vega of a long position in a call or put option is always positive.
Gamma neutrality protects against large changes in the price of the underlying asset between hedge rebalancing.
Vega neutrality protects against variations in .
15.9 RHO
The rho of a portfolio of derivatives measures the sensitivity of the portfolio value to interest rates.
rho = /r
A. For a European call option on a non–dividend–paying stock, rho = KTe–rTN(d2)
B. For a European put option on the stock, rho = –KTe–rTN(–d2), where d 2 d1 T and
ln( S0 / K ) ( r 2 / 2)T
d1 .
T
These formulas apply to European call and put options on stocks and stock indices paying a dividend yield at rate q.
In the case of currency options, there are two rhos corresponding to the two interest rates.
1. The rho corresponding to the domestic interest rate is given by previous formulas.
2. The rho corresponding to the foreign interest rate for a European call on a currency is given by
r T r T
rho Te f S0 N (d1 ) , and, for a European put it is rho Te f S0 N ( d1 )
Section 15.5 – 15.9 Review
Given the following: S0 = 305; K = 300; q = 0.03; r = 0.08; = 0.25; T = 1/3; compute the theta, gamma, vega, and rho
of a four Month European Put Option on a Stock Index:
Preliminary Calculations: compute d1 , d2, N’(d1), N’(–d1) and N’(–d2)
ln( S0 / K ) ( r q 2 / 2)T ln(305/ 300) (0.08 0.03 (0.25)2 / 2)0.3
d1 . d1 0.30215 ; d 2 d1 T 0.1578
T 0.25 0.3
(0.30215) 2
1
N (0.30215) e 2 0.3811 . N ( 0.30215) 0.3974 and N ( 0.1578) 0.4370
2 3.1415
S N ( d1 ) e qt 1 2x
2
B. If the portfolio is worth K1 times the index and each index futures contract is on K2 times the index, then the
number of futures contracts shorted at any given time should be
eq (T *T ) e rT *[1 N ( d1 )]( K1 / K 2 )
Example 10
Suppose that in the previous example futures contracts on the S&P 500 maturing in nine months are used to create the
option synthetically. In this case, initially T = 0.5, T* = 0.75,
Assume: S0 = 900; K = 870; q = 0.03; r = 0.09; = 0.25; T = 0.5; K1 = 100,000; K2 = 250; T* = 0.75;
Question: a. compute the proportion of the portfolio to be sold to create a six month put option.
b. compute the number of futures contracts that should be shorted if nine–month futures contracts on the S&P 500
are used.
2
ln(900/870)(0.090.03 0.25 )0.50 0.03390.045625
a. First compute N(d1). d1 2 0.449; N (0.449) 0.6736 .
0.25 0.50 0.1767
Thus, the proportion of the portfolio that should be initially sold is e0.030.50 [1 0.6736] 0.322
b. To compute the number of futures contracts that should be shorted, use the formula:
e0.03(0.750.50) e0.090.75 [1 0.6736] (100,000 / 250) 123
Frequency of Rebalancing and October 19, 1987
Important notes:
a. Knowing the frequency of portfolio adjustment or rebalancing is important when creating put options
synthetically.
b. Creating put options on the index synthetically does not work well if the volatility of the index changes rapidly
or if the index exhibits large jumps. After the stock market crash of 10/19/87, portfolio managers who insured
themselves by:
i. buying put options survived this crash well.
ii. creating put options synthetically found that they were unable to sell either stocks or index futures fast
enough to protect their position.
Section Description
18.1 The VaR Measure
18.2 Historical Simulation
18.3 Model Building Approach
18.4 Linear Model
18.5 Quadratic Model
18.6 Monte Carlo Simulation
18.7 Comparison of Approaches
18.8 Stress Testing and Back Testing
18.9 Principal Components Analysis
18.10 Appendix 18A: Cash Flow Mapping
The risk measures discussed in Chapter 15, The Greek Letters, provide valuable information for a company’s trader,
but are of limited use to senior management. VaR attempts to provide a single number summarizing the total risk in a
portfolio of financial assets. Two approaches in calculating VaR are the historical simulation approach and the model–
building approach. Both are widely used and there is no agreement on which of the two is better.
The 10–day, 99% VaR for the portfolio of AT&T shares is $368,405.
The 10–day, 99% VaR for the portfolio of both IBM and AT&T shares is $1,622,657.
Thus, the benefit of diversification equals (1,473,621 + 368,405) – 1,622,657 = $219,369
Perfect correlation leads to no risk being “diversified away”.
Example: In the previous two asset portfolio, 1 = 10, 2 = 5, and the dollar change in the portfolio value in one day is
equal to P = 10×x1 + 5×x2.
a. To calculate VaR we, therefore, need to calculate only the mean and standard deviation of P. Assume that the
expected value of each xi is zero. This implies that the mean of P is zero.
b. To calculate the standard deviation of P, let:
1. i represent the daily volatility of the ith asset, which is the s.d. of xi
2. ij represent the coefficient of correlation between returns on asset i and asset j (xi and xj).
n n
The variance of P, P, is given by: P2
i 1 j 1
i i j i j .
n n n
This equation can also be written as P2 i2 i2 2 ij i j i j
i 1 i 1 j i
P 120 1,000 x1 30 20,000 x2 120,000 x1 600,000 x2 , where
x1 and x2 are the proportional changes in the prices of IBM and AT&T in one day and P is the resultant change in
the value of the portfolio.
Assuming that the daily volatility of IBM is 2% and the daily volatility of AT&T is 1% and the correlation between the
daily changes is 0.3,
a. the standard deviation of P (in thousands of dollars) is:
P (1200.02)2 (6000.01)2 20.31200.026000.01 7.099
b. the 5–day 95% value at risk is 1.65 5 7,099 26,193 . Note N (–1.65) = 0.05
2 S 2
Step 5. Subtract the value calculated in Step 1 from the value in Step 4 to determine a sample P.
Step 6. Repeat Steps 2 – 5 many times to build up a probability distribution for P.
Example: Assume 5,000 different sample values of P are computed as described above.
The 1–day 99% VaR is the value of P for the 50th (0.01 × 5,000) worst outcome.
The 1–day 95% VaR is the value of P for the 250th (0.05 × 5,000) worst outcome.
The N–day VaR is usually assumed to be the 1–day VaR multiplied by N
The drawback of Monte Carlo simulation is that it tends to be slow because a company’s complete portfolio has to be
revalued many times. A partial remedy to this problem is to use the partial simulation approach.
n n n
By assuming that P Si i xi Si S j ij xi x j describes the relationship between P and the xi’s
i 1 i 1 j 1
Steps 3 and 4 can be eliminated which avoids the need for a complete revaluation of the portfolio.
The first column shows the maturities of the rates, and the remaining 10 columns show the ten factors (or principal
components) describing the rate.
Factor Loadings (interest rate moves for a particular factor) for U.S. Treasury Data
PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3 mo. 0.21 –0.57 0.50 0.47 –0.39 –0.02 0.01 0.00 0.01 0.00
6 mo. 0.26 –0.49 0.23 –0.37 0.70 0.01 –0.04 –0.02 –0.01 0.00
12 mo. 0.32 –0.32 –0.37 –0.58 –0.52 –0.23 –0.04 –0.05 0.00 0.01
2 yr. 0.35 –0.10 –0.38 0.17 0.04 0.59 0.56 0.12 –0.12 –0.05
3 yr. 0.36 0.02 –0.30 0.27 0.07 0.24 –0.79 0.00 –0.09 –0.00
4 yr. 0.36 0.14 –0.12 0.25 0.16 –0.63 0.15 0.55 –0.14 –0.08
5 yr. 0.36 0.17 –0.04 0.14 0.08 –0.10 0.09 –0.26 0.71 0.48
7 yr. 0.34 0.27 0.15 0.01 0.00 –0.12 0.13 –0.54 0.00 –0.68
10 yr. 0.31 0.30 0.28 –0.10 –0.06 0.01 0.03 –0.23 –0.63 0.52
30 yr. 0.25 0.33 0.46 –0.34 –0.18 0.33 –0.09 0.52 0.26 –0.13
The factor loadings have the property that the sum of their squares for each factor is 1.0.
Factor score defined: The factor score for a day is defined as the quantity of a particular factor in the interest rate
changes on a particular day.
Standard Deviation of Factor Scores (measured in basis points (bps)) listed in order of importance
PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
17.49 6.05 3.10 2.17 1.97 1.69 1.27 1.24 0.80 0.79
Interpretation and analysis of Principle Component (PC) Factors shown above
The first factor in the first table (column PC1) corresponds to a roughly parallel shift in the yield curve.
Example: For 1 unit of factor 1, the 3–month rate increases by 0.21 basis points, the 6–month rate increases by 0.26
basis points, and so on.
The second factor (shown in the column labeled PC2) corresponds to a “twist” or “steepening” of the yield curve.
Example: Rates between 3 months and 2 years move in one direction; rates between 3 years and 30 years move in the
other direction.
The third factor corresponds to a “bowing” of the yield curve.
Example: Rates at the short end and long end of the yield curve move in one direction; rates in the middle move in the
other direction.
Interest rate changes, for any given day, can be expressed as a linear sum of the factors by solving a set of ten
simultaneous equations (note there are 10 rates and 10 factors).
Notes on the Standard Deviation of Factor Scores:
The importance of a factor is measured by the standard deviation (SD) of its factor score.
A 1 SD move in the first factor corresponds to a 3.67 (0.21 × 17.49) bps move in the 3–month rate, a 4.55 (0.26 ×
17.49) bps move in the 6–month rate, and so on.
The technical details of how the factors are determined are discussed by the authors. It is sufficient for us to note that
the factors are chosen so that the factor scores are uncorrelated.
For example, it is important to note that the first factor score (amount of parallel shift) is uncorrelated with the second
factor score (amount of twist) across the 1,543 days.
The variances of the factor scores add up to the total variance of the data. Thus, the total variance of the original
data is 17.492 + 6.052 + 3.102 + ... + 0.792 = 367.9
Most of the risk in interest rate moves is accounted for by the first two or three factors:
17.49 2
the first factor accounts for 83.1% of the variation in the original data.
367.9
17.49 2 6.052
the first two factors account for 93.1% of the variation in the data.
367.9
Thus, we can relate the risks in a portfolio of interest rate dependent instruments to movements in these factors (instead
of considering all ten interest rates).
Using Principal Components Analysis to Calculate VaR
Example: Assume we have a portfolio, with exposures shown in the table below, to interest rate moves.
Change in Portfolio Value for a 1 Basis Point Rate Move ($ millions)
1 yr rate 2 yr rate 3 yr rate 4 yr rate 5 yr rate
+10 +4 –8 –7 +2
a. Choose one month, three months, one year, two years, five years, seven years, ten years, and thirty years
maturities.
b. Regard each T–Bond as a portfolio of its constituent zero–coupon bonds. The position in each of the zero–coupon
bonds is then mapped into an equivalent position in the adjacent standard–maturity zero–coupon bonds.
The Mapping Procedure
Assume that $1,050,000 will be received in 0.8 years. Further assume we are given the following data:
Maturity 3–month 6–month 1 –year
Zero rate 5.50 6.00 7.00
Bond price volatility (% per day) 0.06 0.10 0.20
Solving this quadratic equation, one obtains = 0.3203 (thus, 32.03% of $997,622 should be allocated to a
6–month zero–coupon bond and 67.97% of the value should be allocated to a 1–year zero–coupon bond.)
Step 4: Replace the 0.8–year bond worth $997,662 with the following:
a. a 6-month bond worth 997,662 × 0.3203 = $319,589, and a
b. one–year bond worth 997,662 × 0.6797 = $678, 073
Step 5: For the $50,000 cash flow received at time 0.3 years, a similar mapping to a 3 month and 6 month security
can take place.
The results of the calculations are shown below.
$50,000 received $1,050,000 received
in 0.3 years in 0.8 years Total
Position in 3–month bond ($) 37,397 37,397
Position in 6–month bond ($) 11,793 319,589 331,382
Position in 1–year bond ($) 678,074 678,074
Notes:
a. Carry out similar calculations for the $50,000 cash flow. It can be mapped into a position worth $37,397 in a
three–month bond, a position worth $11,793 in a six–month bond
b. Use the following to determine the variance of the change in the price of the 0.8–year bond with n = 3, 1 =
37,397, 2 = 331,382, 3 = 678,074, 1 = 0.0006, 2 = 0.001, 3 = 0.002, 12= 0.9, 13 = 0.6, and 13 = 0.70. The
variance is 2,628,518. The standard deviation of the change in the price of the bond is 2,628,519 1,621 . The
10–day 99% VaR is 1,621 × 10 × 2.33 = 11,946 or about $11,950.
Advantages of cash flow mapping:
1. it preserves both the value and the variance of the cash flow.
2. the weights assigned to two adjacent zero–coupon bonds are always positive.
Section Description
20.0 Introduction
20.1 Credit Ratings
20.2 Historical Default Probabilities
20.3 Recovery Rates
20.4 Estimating Default Probabilities From Bond Prices
20.5 Comparison Of Default Probability Estimates
20.6 Using Equity Prices To Estimate Default Probabilities
20.7 Credit Risk In Derivatives Transactions
20.8 Credit Risk Migration
20.9 Default Correlation
20.10 Credit VaR
20.0 INTRODUCTION
Significant credit risk for banks and other financial institutions arise from defaults by borrowers, counterparties in
derivatives transactions, and bond issuers.
This chapter focuses on the following key issues associated with credit risk.
a. Different approaches to estimating the probability of default are discussed.
b. The key difference between risk–neutral and real–world probabilities of default is explained.
c. The nature of the credit risk in over–the–counter derivatives transactions and the clauses derivatives dealers write
into their contracts to reduce credit risk.
d. Default correlation, Gaussian copula models, and the estimation of credit value at risk.
Default Intensities
From Table 1:
The unconditional probability of a Caa bond defaulting during the 3rd year is 48.02 – 37.20 = 10.82%
The probability that the Caa–rated bond will survive until the end of year 2 is 100 – 37.20 = 62.80%.
The probability that it will default during the third year conditional on no earlier default is therefore
0.1082/0.6280, or 17.23%. Conditional default probabilities are referred to as default intensities or hazard rates.
The 17.23% just calculated is for a 1–year time period. Consider a short time period of length t.
The default intensity (t) at time t is defined so that (t) t is the probability of default between time t and t + t,
conditional on no earlier default.
If V(t) is the cumulative probability of the company surviving to time t (i.e., no default by time t), then V(t + t) –
V(t) = –(t) V(t) t
t
Taking limits
dV (t ) (t )V (t ) from which we get V (t ) 1 e ( ) d
0
dt
t
Suppose that the probability of default per year (assumed in this example to be the same each year) is Q.
Table 3 below calculates the expected loss from default in terms of Q assuming:
Defaults can happen at times 0.5, 1.5, 2.5, 3.5, and 4.5 years (immediately before coupon payment dates).
Risk–free rates for all maturities are 5% (with continuous compounding).
Table 3: Calculation of loss from default on a bond in terms of the default probabilities per year, Q.
Notional principal = $100.
Time Default Recovery Risk–free Loss given Discount PV of Expected
(years) probability amount ($) value ($) default ($) factor Loss ($)
0.5 Q 40 106.73 66.73 0.9753 65.08Q
1.5 Q 40 105.97 65.97 0.9277 61.20Q
2.5 Q 40 105.17 65.17 0.8825 57.52Q
3.5 Q 40 104.34 64.34 0.8395 54.01 Q
4.5 Q 40 103.46 63.46 0.7985 50.67Q
Total 288.48Q
Consider the 3.5 year row in the table above.
The expected value of the risk–free bond at time 3.5 years (calculated using forward interest rates) is
3 + 3e–0.05×0.5 + 3e–0.05×1.0 + 103e–0.05×1.5 = 104.34
The amount recovered if there is a default is 40, so that the loss given default is 104.34 – 40 = $64.34. The present
value of this loss is 54.01. The expected loss is therefore 54.01Q.
The total expected loss is 288.48Q.
Thus, 288.48Q = $8.75; Q = 3.03%.
The calculations just performed assume that:
the default probability is the same in each year and
defaults take place at just one time during the year.
Now, extend the calculations to assume:
that defaults can take place more frequently.
a constant default intensity or a particular pattern for the variation of default probabilities with time, instead of
assuming a constant unconditional probability of default.
With several bonds, several parameter estimates can be computed to determine the term structure of default
probabilities. Example: Suppose we have bonds maturing in 3, 5, 7, and 10 years.
Use the first bond to estimate a default probability per year for the first 3 years,
Use the second bond to estimate default probability per year for years 4 and 5,
Use the third bond to estimate a default probability for years 6 and 7, and
Use the fourth bond to estimate a default probability for years 8, 9, and 10 (see problems 20.15 and 20.27).
This is analogous to the bootstrap procedure in Section 4.5 for calculating a zero–coupon yield curve.
The Risk–Free Rate
When bond prices are used to estimate default probabilities, what is the meaning of the terms “risk–free rate” and
“risk–free bond”? The answer differs, as explained below:
the spread s is the excess of the corporate bond yield over the yield on a similar T–bond (see eq. 20.2).
the risk–free value of the bond must be calculated using the risk–free rate (see table 3).
the benchmark risk–free rate usually used in quoting corporate bond yields is the yield on similar Treasury bonds
(e.g. a bond trader might quote the yield on a particular corporate bond as being a spread of 250 basis points over
Treasuries).
traders use LIBOR/swap rates as proxies for risk–free rates when valuing derivatives (see section 4.1). traders also
use LIBOR/swap rates as risk–free rates when calculating default probabilities. Example: the spread s in eq. 20.2 is
the spread of the bond yield over the LIBOR/swap rate when they determine default probabilities from bond prices
the risk–free discount rates used in the calculations in Table 3 are LIBOR/swap zero rates.
Credit default swaps can be used to imply the risk–free rate assumed by traders.
The rate used is approximately equal to the LIBOR/swap rate minus 10 basis points on average, and is plausible.
The credit risk in a swap rate is the credit risk from making a series of 6–month loans to AA–rated counterparties
and 10 basis points is a reasonable default risk premium for a AA–rated 6–month instrument (see Section 7.5).
Asset Swaps
Traders often use asset swap spreads to extract default probabilities from bond prices.
Asset swap spreads provide a direct estimate of the spread of bond yields over the LIBOR/swap curve.
Consider an asset swap spread for a particular bond as 150 basis points. There are three possible situations:
1. The bond sells for its par value of 100. The swap involves one side (company A) paying the coupon on the bond and
the other side (company B) paying LIBOR plus 150 basis points.
2. The bond sells below its par value (e.g. for 95). The swap is structured so that in addition to the coupons, company
A pays $5 per $100 of notional principal at the outset.
3. The underlying bond sells above par (e.g. for 108). Company B then makes a payment of $8 per $100 of principal at
the outset.
Thus, the present value of the asset swap spread is the amount by which the price of the corporate bond is exceeded by
the price of a similar risk–free bond where the risk–free rate is assumed to be given by the LIBOR/swap curve (see
Problem 20.24).
Consider the example in Table 3 where the LIBOR/swap zero curve is flat at 5%.
Suppose it is known that the asset swap spread is 150 basis points (instead of knowing the bond's price).
Thus, the amount by which the value of the risk–free bond exceeds the value of the corporate bond is the present value
of 150 basis points per year for 5 years.
Assuming semiannual payments, this is $6.55 per $100 of principal.
The total loss in Table 3, in this case, would be set equal to $6.55.
The default probability per year is Q = 6.55/288.48 = 2.27%.
It shows the excess return over the risk–free rate (still assumed to be the 7–year swap rate minus 10 basis points)
earned by investors in bonds with different credit rating.
Table 5: Expected excess return on bonds (basis points)
Rating Bond yield spread over Spread of risk–free rate Spread for historical Expected excess
Treasuries over Treasuries defaults return
Aaa 83 43 2 38
Aa 90 43 4 43
A 120 43 8 69
Baa 186 43 28 115
Ba 347 43 144 160
B 585 43 449 93
Caa 1,321 43 1,014 264
Consider again an A–rated bond.
The average spread over Treasuries is 120 basis points.
Of this, 43 basis points are accounted for by the average spread between 7–year Treasuries and our proxy for the
risk–free rate.
A spread of 8 basis points is necessary to cover expected defaults. (This equals the real–world probability of
default from Table 4 times 1 minus the assumed recovery rate of 0.4.)
Thus, an expected excess return (after expected defaults have been taken into account) equals 69 basis points.
Tables 4 and 5 show that a large percentage difference between default probability estimates translates into a small
(but significant) expected excess return on the bond.
For Aaa–rated bonds, the ratio of the two default probabilities is 16.8, but the expected excess return is only 38
basis points.
The expected return tends to increase as credit quality declines.
Real–World vs. Risk–Neutral Probabilities
Default probabilities implied from bond yields are risk–neutral probabilities of default.
Consider the calculations of default probabilities in Table 3.
The calculations assume that expected default losses can be discounted at the risk–free rate.
The risk–neutral valuation principle shows that this is valid provided the expected losses are calculated in a risk–
neutral world (i.e. the default probability Q in Table 3 must be a risk–neutral probability).
By contrast, the default probabilities implied from historical data are real–world default probabilities (a.k.a. physical
probabilities). The expected excess return in Table 5 arises from the difference between real–world and risk–neutral
default probabilities.
If there were no expected excess return, then the real–world and risk–neutral default probabilities would be the same,
and vice versa. Why are there big differences between real–world and risk–neutral default probabilities? Stated
differently, why do corporate bond traders earn more than the risk–free rate on average?
Potential reasons include:
2. Corporate bonds are relatively illiquid and bond traders demand an extra return in compensation.
3. Subjective bond trader default probabilities may be higher than those in Table 1 (e.g. bond traders may be allowing
for depression scenarios worse than seen during 1970 to 2003).
4. Bonds do not default independently of each other.
This is the most important reason for the results in Tables 4 and 5.
There are periods of time when default rates are very low and periods when they are very high.
This gives rise to systematic risk (i.e., risk that cannot be diversified away) and bond traders should require an
expected excess return for bearing the risk.
The variation in yearly default rates may be due to overall economic conditions or because a default by one
company has a ripple effect resulting in defaults by other companies (the latter referred to as credit contagion).
5. Bond returns are highly skewed with limited upside.
Thus, it is more difficult to diversify risks in a bond portfolio than in an equity portfolio.
Since a large number of different bonds must be held, and since many bond portfolios are not fully diversified,
bond traders may require an extra return for bearing unsystematic risk in addition to the systematic risk
mentioned above.
Should real–world or risk–neutral default probabilities be used in the analysis of credit risk?
The answer depends on the purpose of the analysis.
When valuing credit derivatives or estimating the impact of default risk on the pricing of instruments, use risk–
neutral default probabilities.
The analysis calculates the present value of expected future cash flows and almost invariably (implicitly or
explicitly) involves using risk–neutral valuation.
When carrying out scenario analyses to calculate potential future losses from defaults, use real–world default
probabilities.
If VT < D, the company is likely to default on the debt at time T. The value of the equity is zero.
If VT > D, the company is likely repay the debt at time T. The value of the equity at this time is VT – D.
Thus, the value of the firm's equity at time T is ET = max(VT – D, 0)
The above equation shows that the firm’s equity is a call option on firm’s assets with a strike price equal to the
repayment on the debt.
ln V0 / D ( r V2 / 2)T
The Black–Scholes formula values a firm’s equity as E0 = V0 N(d1) – De–rt N(d2) where d1
V T
and d 2 d1 V T
The value of the debt today is V0 – E0.
The risk–neutral probability that the company will default on the debt is N(–d2). To calculate this, we need to
determine V0 and V.
neither of these are directly observable.
however, E0 is observable if the company is publicly traded.
This means that E0 = V0N(d1) – De–rt N(d2) provides one condition that must be satisfied by V0 and V.
E
We can estimate E using It ô 's lemma: E E0 V V0 N (d1 ) V V0
V
This provides another equation that must be satisfied by V0 and V. Thus, E0 = V0N(d1) – De–rt N(d2) and E E0 = N(d1)
V V0 provide simultaneous equations that can be solved for V0 and V.
Example 1. The value of a company's equity is $3 million. The volatility of its equity is 80%.
The debt to be paid in one year is $10 million. The risk–free rate is 5% per annum. Thus, E0 = 3, E = 0.80, r = 0.05, T
= 1, D = 10.
a. Solving E0 = V0N(d1) – De–rt N(d2) and E E0 = N(d1) V V0 yields V0 = 12.40 and V = 0.2123.
b. The parameter d2 is 1.1408, so that the probability of default is N(–d2) = 0.127 = 12.7%.
c. The market value of the debt is V0 – E0 = 12.40 – 3.0 = 9.40.
d. The present value of the promised payment on the debt is 10e–0.05×1 = 9.51.
e. The expected loss on the debt is therefore (9.51 – 9.40)/9.51 or about 1.2% of its no–default value.
Comparing this with the probability of default gives the expected recovery in the event of a default as (12.7 –
1.2)/12.7, or about 91%.
Q: How well do the default probabilities produced by Merton's model relate to actual default experience?
A: Merton's model produces a good ranking of default probabilities (risk–neutral or real–world). Thus, a monotonic
transformation can be used to convert the probability of default output from Merton's model into a good estimate of
either the real–world or risk–neutral default probability.
instrument that pays off the exposure on the derivative under consideration at time ti.
Consider the three categories of derivatives mentioned earlier.
n
For the first category, the value of fi is always negative, the total expected loss from defaults, u v
i 1
i i 0 , and the
Its loss is the payoff from a portfolio of call options on the contract values where each option has a strike price of zero.
N
With netting, the financial institution loses (1 R )max Vi ,0
i 1
Its loss is the payoff from a single option on the portfolio of contract values with a strike price of zero. Further, the
value of an option on a portfolio is never greater than, and is often considerably less than, the value of the
corresponding portfolio of options.
Reducing Exposure to Credit Risk
Two ways financial institutions reduce potential losses in the event of a default.
1. Collateralization. Collateralization mitigates credit risks.
Consider that a company and a financial institution have entered into a number of derivatives contracts.
A collateralization agreement specifies that the contracts be marked to market using a pre–agreed formula.
If the total value of the contracts to the financial institution is above a certain threshold level on a certain day, it
can ask the company to post collateral.
The amount of collateral posted, when added to collateral already posted by the company, is equal to the difference
between the value of the contract to the financial institution and the threshold level.
When the contract moves in favor of the company so that the difference between value of the contract to the
financial institution and the threshold level is less than the total margin already posted, the company can reclaim
margin.
In the event of a default by the company, the financial institution can seize the collateral. If the company does not
post collateral as required, the financial institution can close out the contracts.
Example: Assume the threshold level for the company is $10 million and contract is marked to market daily for the
purposes of collateralization.
If one day the value of the contract to financial institution is $10.5 million, it can ask for $0.5 million of collateral.
If the next day the value of the contract rises to $11.4 million it can ask for a further $0.9 million of collateral.
If the value of the contract falls to $10.9 million on the following day, the company can ask for $0.5 million of the
collateral to be returned.
The threshold (i.e. the $10 million) can be regarded as a line of credit that the financial institution is prepared to grant
to the company.
The threshold amount is not subject to protection.
Even when the threshold is zero, the protection is not total.
When a company gets into financial difficulties, it is likely to stop responding to requests to post collateral. By the
time the counterparty exercises its right to close out contracts, their value may have moved further in its favor.
2. Use of downgrade triggers.
Are clauses in contracts with counterparties that state that if the credit rating of the counterparty falls below a
certain level, then the contract is closed out with one side paying an amount to the other side.
Lead to a significant reduction in credit risk, but do not completely eliminate all credit risk.
If there is a large downward fall in the credit rating of the counterparty (e.g. from A to default) in a short period of
time, the financial institution will still suffer a credit loss.
These are “percentile–to–percentile” transformations. For example, the 5–percentile point in the probability
distribution for t1 is transformed to x1 = –1.645, which is the 5–percentile point in the standard normal distribution.
x1 and x2 have normal distributions with mean 0, and standard deviation 1.0.
Assume that the joint distribution of x1 and x2 is bivariate normal with correlation 12.
This assumption is referred to as using a Gaussian copula.
The assumption is convenient because it means that the joint probability distribution of t1 and t2 is fully defined by
the cumulative default probability distributions Q1 and Q2 for t1 and t2, together with a single correlation parameter
12.
The attraction of the Gaussian copula model is that it can be extended to many companies.
Suppose that we are considering n companies and that ti is the time to default of the ith company.
Transform each ti into a new variable, xi that has a standard normal distribution.
The transformation is the percentile–to–percentile transformation xi = N –1[Qi (ti)], where
where Qi is the cumulative probability distribution for ti. We then assume that the xi are multivariate normal. The
default correlation between ti and tj is measured as the correlation between xi and xj. This is referred to as the copula
correlation.
The Gaussian copula approach:
is a useful way representing the correlation structure between variables that are not normally distributed.
allows the correlation structure of the variables to be estimated separately from their marginal (unconditional)
distributions.
assumes that after a transformation is applied to each variable they are multivariate normal.
Example 3 – Use of a Gaussian copula approach
Suppose we wish to simulate defaults during the next five years in n companies.
For each company the cumulative probability of a default during the next 1, 2, 3, 4, and 5 years is 1 %, 3%, 6%,
10%, and 15%, respectively.
Sample from a multivariate normal distribution to obtain xi (1 i n).
Convert the xi to ti, a time to default.
Determine when defaults take place within a year. When the sample from the normal distribution is:
less than N – 1(0.01) = –2.33, a default takes place within the first year,
between –2.33 and N – 1(0.03) = –1.88, a default takes place during the second year;
between –1.88 and N – 1(0.06) = –1.55, a default takes place during the third year;
between –1.55 and N – 1(0.03) = –1.28, a default takes place during the fourth year;
between –1.28 and N – 1(0.03) = –1.04, a default takes place during the fifth year.
greater than –1.04, there is no default.
Using Factors to Define the Correlation Structure
To avoid defining a different correlation between xi and xj for each pair of companies i and j in the Gaussian copula
model, a one–factor model is often used.
The assumption is that xi ai M 1 ai2 Zi
M is a common factor affecting defaults for all companies and Zi is a factor affecting only company i.
The variables M and the Zi have independent standard normal distributions.
The ai are constant parameters between –1 and +1.
The correlation between xi and xj is ai aj.
Suppose that the probability that company i will default by a particular time T is Qi(T).
Under the Gaussian copula model, a default happens when N ( xi ) Qi (T ) or xi N 1 [Qi (T )]
N 1 [Qi (T )] ai M
From equation (20.6), this condition is xi ai M 1 ai2 Z i N 1[Qi (T )] , or Z i
1 ai2
N 1 [Q (T )] a M
Conditional on the value of the factor M, the probability of default is therefore Qi (T M ) N i i
Eq
1 a 2
i
(20.7)
A particular case of the one–factor Gaussian model is where the probability distributions of default are the same for all
i and the correlations between xi and xj is the same for all i and j.
Suppose that Qi(T) = Q(T) for all i and that the common correlation is , so that ai for all i.
N 1 [Q (T )] M
Eq (20.7) becomes Qi (T M ) N i
Eq (20.8)
1 2
Binomial Correlation Measures
Consider two companies A and B.
A default correlation measure is the coefficient of correlation between:
1. A variable that equals 1 if company A defaults between times 0 and T and 0 otherwise; and
2. A variable that equals 1 if company B defaults between times 0 and T and 0 otherwise
PAB (T ) QA (T )QB (T )
The measure is AB (T ) = , where
[QA (T ) QA (T )2 ][QB (T ) QB (T )2 ]
PAB(T) is the joint probability of A and B defaulting between time zero and time T.
QA(T) is the cumulative probability that company A will default by time T, and
QB(T) is the cumulative probability that company B will default by time T.
AB(T) typically increases as T increases.
Relationship between Measures
Define: M(a, b; ) as the probability in a standardized bivariate normal distribution that the first variable is less than a
and the second variable is less than b if the coefficient of correlation between the variables is .
Let AB be the default correlation between A and B in the Gaussian copula model.
M ( xA (T ), xB (T ); AB ) QA (T )QB (T )
It follows that PAB (T ) M ( x A (T ), xB (T ); AB ) so that AB (T )
[QA (T ) QA (T )2 ][QB (T ) QB (T )2 ]
This shows that, if QA(T) and QB(T) are known, AB(T) can be calculated from AB and vice versa.
AB is usually markedly greater than AB(T).
Example 4 Suppose the probability of company A defaulting in a one–year period is 1% and company B defaulting in
a one–year period is 1%. Further, let AB = 0.20.
Compute AB (1)
M ( x A (1), xB (1); AB ) QA (1)QB (1)
Step 1: Write an equation to determine AB (1). AB (1) , where
[QA (1) QA (1)2 ][ QB (1) QB (1)2 ]
Qi (ti ) as the cumulative probability distribution of ti and ui (ti ) N -1 [Qi (ti )] for 1 i N .
Step 2: Using the equations in Step 1, compute AB (1)
Compute xA (1) and xB (1); xA (1) = xB (1) = N – 1(0.01) = –2.326
Since AB is 0.20, then M ( x A (1), x B (1), AB ) = 0.000337 ; Thus, AB (1) = 0.024
A rough estimate of the credit VaR when an X % confidence level is used and the time horizon is T is therefore L(1–
R)V(X, T), where L is the size of the loan portfolio and R is the recovery rate.
The contribution of a loan of size Li to the credit VaR is Li(1–R)V(X, T).
Example 4: Consider the following:
A bank has a total of $100 million of retail exposures.
The 1–year probability of default averages 2% and the recovery rate averages 60%.
The copula correlation parameter is estimated as 0.1. Therefore,
N 1 [(0.02) 0.10 N 1 (0.999)
V (0.999,1) N 0.0128 . This shows that the 99.9% worst case default rate is
1 0.10
12.8%. The 1–year 99.9% credit VaR is therefore $100 million × 0.128 × (1 – 0.6) or $5.13 million.
CreditMetrics
One method for calculating credit VaR is CreditMetrics
a probability distribution of credit losses is estimated by simulating credit rating changes for each counterparty.
the probability distribution of losses over a one–year period can be determined as follows.
a. sample to determine credit rating changes of all counterparties throughout the year for each simulation trial.
b. re–value outstanding contracts to determine the total of credit losses from defaults and credit rating changes
a probability distribution for credit losses is obtained after a large number of simulation trials are conducted.
Comments:
This approach is computationally time intensive.
It has the advantage that credit losses are defined as those arising from credit downgrades and defaults.
The impact of credit mitigation clauses such as those described Section 20.–8 can be approximately incorporated
into the analysis.
Table 6 is typical of the historical data provided by rating agencies on credit rating changes and could be used as a
basis for a CreditMetrics Monte Carlo simulation.
It shows the percentage probability of a bond moving from one rating category to another in 1–year.
For example, a bond that starts with an A credit rating has a 91.84% chance of still having an A rating at the end of
1 year. It has a 0.02% chance of defaulting during the year, a 0.13% chance of dropping to B, and so on.
In sampling to determine credit losses, the credit rating changes for different counterparties should not be assumed to
be independent.
A Gaussian copula model can be used to construct a joint probability distribution of rating changes similarly to the
way it is used to describe the joint probability distribution of times to default.
The copula correlation between the rating transitions for two companies is usually set equal to the correlation between
their equity returns using a factor model similar to that in Section 20.9.
Example: Suppose the following:
we wish to simulate the rating change of a Aaa and a Baa company over a one–year period using the transition
matrix in Table 6.
the correlation between the equities of the two companies is 0.2.
sample two variables xA and xB from normal distributions so that their correlation is 0.2 (on each simulation trial).
the variable xA determines the new rating of the Aaa company and the variable xB determines the rating of the Baa
company.