ST/SG/AC.8/2001/CRP.

8
________________________________________________________________________________

19 July 2001

English
_________________________________________________________________________________

Ad Hoc Group of Experts on International

Cooperation in Tax Matters
Tenth meeting
Geneva, 10 - 14 September 2001

Innovative Financial Transactions: Tax Policy Implications

(A Report by the Special Sessions on Innovative
Financial Transactions of the OECD)1

1
The present paper was prepared by the Organisation for Economic Cooperation and
Development, Paris, France. The views and opinions expressed therein are those of the authors and
do not necessarily represent those of the United Nations.
 2

EXECUTIVE SUMMARY

The increased use of innovative financial transactions in recent years is primarily due to the
need for greater risk management by businesses and financial investors and the development of
sophisticated instruments tailored to meet such demands.

A broad group of innovative financial instruments, referred to as "derivative" contracts, call for
specified cash flows to be made between the counterparties over time. Unlike traditional debt and
equity securities, these instruments generally do not involve a return on an initial investment. Rather,
derivative contracts are constructed and priced by reference to values "derived" from an underlying
index, commodity, or other asset, and their value fluctuates with the market movement of that
referenced item.

The basic building blocks of derivative instruments -- forward contracts and options -- can be
combined in any number of ways as specified by the counterparties. In particular, derivative contracts
can separate each of the discrete economic attributes of a particular position or recombine them into
new forms. Significantly, they can also be constructed to replicate any specified set of economic
attributes (including those of debt or equity instruments) in a variety of forms.

Traditional patterns of investment have also been expanded through securities lending and
repurchase agreements. These arrangements permit the owner to transfer title or possession of
underlying securities, while retaining the economic attributes of the position.

Derivative instruments and other innovative financial transactions serve legitimate business and
investment purposes. The holder can use such products either to take a position carrying specifically
defined opportunities for profit and loss, or to offset (i.e., "hedge") the inherent risks of other
investments or business activities. This ability to shift, substitute, or transform risks through the use
of financial products is an essential tool of modern business and investment.

In addition to their critical role in risk management, innovative financial instruments also
present a number of serious challenges for income tax systems.

1) The traditional income tax issues of character, source, and the timing and amount of
income are generally based on an initial classification of the type of income in question.
These systems of categorisation are difficult to maintain and administer given the
emergence of instruments that can mirror economic attributes of investments in any
number of diverse forms.

2) Similarly, the fundamental distinctions in most income tax systems between debt and
equity are challenged by instruments providing for returns and risks that are economically
equivalent to the financial attributes of debt and equity investments, or any "hybrid"
combination thereof.
 3

3) The critical issue of determining the "owner" of an instrument for tax purposes is also
tested by contracts that replicate, shift or eliminate some or all of the returns and risks of
an investment. On the other hand, securities lending and repurchase arrangements that
transfer legal title but retain economic attributes of an investment present similar
problems of identifying the "tax owner" of a position.
4) Finally, tax systems are challenged by two broad and sometimes competing concerns --
(a) removing artificial tax barriers to effective risk management strategies; and (b)
limiting the opportunities for tax arbitrage.

Tax systems that have addressed these issues have adopted a number of different approaches.
In general terms, these approaches can be classified as follows:

(1) Reliance on Financial Accounting Rules. Aligning the tax and financial accounting
treatment of innovative financial transactions can in some cases offer greater consistency
and reduced compliance costs. Financial accounting standards, however, are far from
settled in this area, and in some circumstances allow greater subjectivity and discretion
than would be acceptable for tax purposes.

(2) Bifurcation. This approach relies on the disaggregation of financial instruments, treating
each of their discrete economic components separately for income tax purposes. The
purpose of this approach is to isolate and identify each element with the goal of applying
tax rules consistently to the particular components. This goal is often frustrated in
practice, however, since there is little agreement on the appropriate method of bifurcation,
or on the taxation of the constituent elements.

(3) Integration. Some tax systems allow taxpayers the election to "match" the tax attributes
of their "hedging instruments" with the attributes of specified business or investment
transactions. By integrating the two offsetting positions, the income tax system permits
the hedging strategy to be effective on an after tax basis. Similarly, some tax systems
impose mandatory integration of certain offsetting transactions to prevent potential
abusive transactions. In each case, integration of transactions raises difficult issues
regarding scope of the rules and identification of appropriately "matched" transactions.

(4) Mark-to-Market Systems. Many of the timing and classification problems described
above can be avoided to the extent that a mark-to-market system applies for income tax
purposes. Under such an approach, financial instruments are treated for income tax
purposes as if they are sold at year-end, and all resulting gain or loss is taken into account.
Difficulties in making such systems work in practice include valuation of illiquid
positions and the cash flow (and potential abuse) problems resulting from imposition of
tax effects in advance of market transactions. However, where a mark-to-market system
is limited to particular sectors that have adopted the method for accounting purposes
(such as financial intermediaries trading in financial instruments), its expansion to cover
taxation may offer an appropriate measure of profits and losses without the numerous
 4

practical concerns associated with bifurcation and integration methods.

(5) Anti-Abuse Measures. Some tax authorities have applied broad anti-abuse rules which
impose "substance over form" rules to combat tax arbitrage. To the extent such rules
target specific abuse transactions they are often adopted too late to be effective. On the
other hand, broad rules that give tax authorities discretion to consider the "facts and
circumstances" may result in uncertain and inconsistent administration of the rules.

All of the above approaches should be considered, either individually or in combination, in

developing appropriate responses to the challenges presented by innovative financial transactions.
New tax rules designed to meet these challenges must address the overall tax policy objectives of
neutrality and equity in order to promote the efficiency of the financial markets and protect the
revenue base. This must be accomplished, however, with appropriate attention to the goals of
certainty and administrability. Moreover, as the use of new financial products continues to expand,
they will inevitably test the viability of the fundamental rules and classifications on which income tax
systems and international tax agreements are currently based.
 5

Table of Contents

Page

Executive Summary 1

I. Background
6
A. Developments in Financial Markets
6
B. Risk Management
6
C. Participants
7

II. Overview of Innovative Financial Transactions

8
A. Traditional Financial Products
9
B. Building Blocks for Derivative Instruments
9
1. In General
9
2. Forwards/Futures Contracts
10
3. Options
10
C. Multiple Payment Period Derivatives - Notional Principal Contracts 10
1. In General
11
2. Swaps
11
3. Caps/Collars/Floors
11
4. Equity Derivatives and other Total Return Swaps
11
D. Replication and Synthetic Instruments
12
E. Securities Loans and Repurchase Agreements
14

III. Income Tax Issues and Problems Presented by Innovative

Financial Instruments
15
 6

A. In General
15
B. Traditional Tax Issues
16
1. Character of Income
16
2. Timing and Amount of Income
16
3. Source of Income
18
4. Treaty Classification
18
C. Instrument Classification
19
1. Debt/Equity/Derivative Distinctions
19
2. Synthetics and Hybrids
20
D. Breakdown of Ownership Concepts
21
E. Matching Positions for Effective Risk Management
22
F. Tax Arbitrage
23

IV. General Tax Policy Goals

23

V. Policy Options
25
A. General Approaches vs. Specific Rules
25
B. Reliance of Financial Accounting Rules
26
C. Bifurcation
27
1. In General
27
2. Problems with Bifurcation 28
3. Practical Issues 29
4. Alternative Methodology: Estimating the Yield 29
 7

D. Development of Elective Hedging Systems 29

1. The Demand for Risk Management 29
2. The Need for Effective Tax Rules 30
3. Characteristics of Elective Hedging Systems 30
4. Problems of Scope C Portfolio and Anticipatory Hedging 31
5. Partial vs. Full Integration 32
E. Non-Elective Integration 33
1. Use of Integration to Prevent Tax Avoidance 33
2. Problems with Mandatory Integration 34
3. Non-Elective Integration through Uniform Characterisation Rules 34
F. Are Bifurcation and Integration Compatible? 35
G. Mark-to-Market Valuation Systems 35
H. Anti-Abuse Measures 36

VI. Summary
37

Appendix I: Examples of Replication

39
Appendix II: Examples of Bifurcation
42
Appendix III: Hedging Examples
44
 8

I. Background

A. Developments in Financial Market

1. During the last twenty years, there has been a tremendous increase in the availability and variety
of innovative financial transactions. In many cases, domestic tax systems have proved ill-equipped to
keep pace with the dynamic evolution of these instruments.

2. Many related factors have contributed to the needs of business for more sophisticated financial
instruments, as well as to the financial world=s technological ability to meet that demand.
Developments in international financial markets that have influenced this dramatic change include the
following]

! increased volatility of interest and currency exchange rate

! expansion of multinational operations of corporations
! deregulation of financial markets
! elimination or liberalisation of exchange controls
! increasing speed of transmission of information and money
! the evolution in the power and availability of computer technology for use in pricing
and risk management

3. As a consequence of those changes, innovative financial transactions have been developed to

meet the specific risk-return needs of companies and others engaged in the financial markets.

4. Domestic tax legislation and international tax agreements generally lag behind developments
in the financial markets, with the consequence that the domestic and international tax treatment of
financial instruments is often uncertain. In many cases, tax rules and policies that were developed to
address more traditional financial instruments are not suited to deal with modern financial
instruments. That may hold true whether or not a country's tax rules are largely in harmony with
accounting standards and practice, or are largely independent of the accounting rules.

B. Risk Management

5. One of the central themes reflected in these developments is the evolution of the ways in
which market participants manage risk. As financial products become more tailored to discrete
business specifications, they have evolved into more sophisticated tools for allocating risks between
parties.
6. Indeed, the original purpose of derivative transactions was to allocate risk in ways that other
instruments did not. For example, commodity futures and options have allowed persons to hedge or
to take a position in commodities without the transaction costs of the market in physical commodities.
Similarly, the subsequent expansion of derivatives activities to financial markets, particularly the
over-the-counter market, has allowed participants to tailor the risks in a way that traditional debt
instruments did not.
 9

7. Because the prices of derivatives generally respond more quickly to changes in market
conditions than prices of primary instruments, derivatives allow investors to leverage their investment
in a particular risk. The more narrowly crafted the risk exposure, the more leverage is possible.

8. Market risk that is allocated through an instrument=s cash flows can be created by
intermediaries so that the participation of the object of the transaction is not necessary. All that is
required is a benchmark to determine which of the parties must pay the other. In response to this
demand, financial institutions and large pools of capital have emerged that are willing to take on this
risk and perform an intermediation function.

9. For example, over the past decade, many financial intermediaries issued cash-settled equity
warrants with respect to the shares of industrial companies or baskets of companies in particular
industrial sectors. Payment due upon exercise would be determined by the public trading price of the
stock. Although the intermediary frequently hedged its risk by owning shares of the underlying equity
securities, such hedging was not necessary and in any event did not need to involve the issuer since
the intermediary could purchase the shares in the public market.

10. Risk management involves not only the traditional protection against interest rate and currency
exchange fluctuations, but also recognises that "risk" may consist of failure to take advantage of a
business opportunity.

! Asset managers use derivative instruments to tailor the risk profile of a security or
portfolio to particular investment objectives. For example, they can choose to keep
long-term exposure to a particular security while swapping away the short-term exposure
by entering into a two-year (or shorter) single-stock equity swap. Alternatively, the
manager can use small "baskets" of currencies to hedge larger portfolio credit risk
because of correlations in movements between different currencies, reducing transaction
costs of hedging. Similarly, banks may use real estate swaps relating to a reasonably
small number of buildings to hedge a large portfolio of real estate.
! On the liability side, issuers of securities can lower cost of funds by issuing debt into
global capital markets and entering into a derivative transaction with an intermediary to
lay off the inherent risks in the cross-border positions.

11. Counterparty credit risk in some cases is addressed through exchanges. Standard contracts are
traded on exchanges. The contracts have regular performance dates and the parties negotiate only on
price. The exchange manages counterparty risks by requiring the posting of margins on a daily basis.
These types of contracts have generally been highly regulated.

12. In contrast, instruments offered over the counter have not been as fully regulated by
government authorities or industry. Over the counter contracts can be structured with any terms
required by the party. This flexibility comes with the cost of increased counterparty risk.

C. Participants
13. The tax treatment of financial instruments is sometimes determined based on the roles played
 10

by the various parties to the transaction. Moreover, there is a growing recognition that uniform tax
treatment among the various counterparties is not necessarily required, and that the function of each
separate party may be an important determining factor for the method of taxation. For example, the
availability of the mark-to-market valuation system for measuring taxable income with respect to
financial products has in some contexts been limited to certain types of activity, generally conducted
by financial intermediaries. Where such a limitation applies, tax "consistency" based on the functions
of the participants with respect to the transactions would take precedence over tax "consistency"
among the counterparties to a single transaction.

14. Participants with respect to innovative financial transactions can be classified to include the
following broad categories:

! End-Users typically use financial instruments to manage their levels of risk with respect
to their business or investment activities. End-users can be corporations, insurance
companies, mutual funds, pension funds, and other persons with business risks or assets
or liabilities subject to fluctuations in value.

! Financial Intermediaries, primarily banks and securities firms, generally take on only
counterparty credit risk, and fully or partially hedge the risks of one party by passing them
on through contracts to another. Previously, these institutions served as traditional
intermediaries between parties who had surplus capital for investment and parties who
were seeking additional capital. The development of innovative financial transactions has
expanded this role, allowing the financial institution to enter into transactions directly
with customers, and to modify, manage, and spread the resulting risks more efficiently.

II. Overview of Innovative Financial Transactions

A. Traditional Financial Products

15. Traditional financial instruments were generally limited to securities that could be classified as
either debt or equity investments. Moreover, the traditional rules of taxation presumed a sharp
distinction between these two forms of instruments. In particular, it was assumed that debt provided
for a relatively fixed return of income and principal, whereas returns on equity investments were
contingent on the performance of the issuing corporation. These classifications lead to tax rules
reflecting differences in the timing and character of the income from these discrete classifications.
For example, due to the greater perceived risk in the investment, equity instruments tended to be
taxed on a realisation basis and, in some cases, on capital account. In contrast, debt tended to be
taxed on an accruals basis and on revenue account.
16. The explosive growth in the use of new financial products during the last several decades has
revealed an underlying weakness in the above assumptions. Both the development of new
instruments and the application of financial engineering techniques to existing products have
demonstrated that the risk-return relationship between debt and equity is a continuum rather than a
difference in kind.
 11

17. One of the factors which fuelled this breakdown of traditional debt/equity distinctions was the
development of securities that combined elements of both categories. Examples of such instruments
include the following:

! Perpetual Debt. Debt instruments have been issued in a number of forms that have no
fixed maturity date or uncommonly long terms. In some cases, these instruments
developed in response to bank regulatory requirements. In other cases, the instruments
have been solely tax-motivated, employing interest prepayments and related party
repurchases to inflate the income tax benefits of interest deductions and other tax
attributes.

! Equity Notes. Debt instruments have been issued in a variety of forms that reflect
characteristics of equity investments while claiming the tax benefits associated with debt.
For example, interest returns can be linked to dividends (or otherwise contingent on
performance or earnings of the issuer), and the instruments could be subordinated to rank
below all other securities issued to creditors. Tax advantages may be multiplied by such
instruments where they are issued cross-border to take advantage of the differing tax
classifications between two jurisdictions.

! Bonds with Equity Warrants and Convertible Debt. In some debt issues, the holder
receives a package of a bond and a warrant entitling the holder to subscribe for shares of
the issuer or a related party. The warrant is a separate instrument traded independently
from the bond. The issuer benefits from lower financing costs, since the value of the
warrant is taken into account in the pricing of the bond. Other debt obligations consist of
a single instrument that is convertible into equity of the issuer. In both cases, the holder
receives a mixture of fixed and contingent returns. The distinction between the two
packages of rights, as well as the tax treatment of the different components embedded in
each, challenge the traditional tax classification system.

18. Although the issuance of these and other "hybrid" instruments have been a factor in the
continued erosion of the debt/equity distinction, they are not the sole cause. As discussed below,
challenges have also arisen from the introduction and market development of new types of
"derivative" instruments, from the increased ability to disaggregate and recombine the constituent
elements of all financial products, and from the breakdown in the concepts of ownership with respect
to particular instruments.

B. Building Blocks for Derivative Instruments

1. In General

19. A "derivative" instrument is simply a contractual right or obligation that "derives" its value
from the value of something else, such as a debt security, an equity or commodity, or a specified
index. In general terms, derivative instruments are constructed from two basic types of contracts
which are described in more detail in the following sections: (i) contracts that provide for fixed
 12

contractual rights that will be executed in the future are referred to as "forward contracts" (or, in their
standardised form, "futures contracts"); and (ii) contracts that are contingent on one party's decision to
execute in the future are referred to as "options contracts". These basic building blocks are used in
both developing and pricing the numerous derivative products available on the market today.

2. Forwards/Futures Contracts

20. A forward contract is an executory agreement to buy or sell a specified amount of an asset at a
specified price on a certain future date. The forward price is not based on a projection of the parties'
expectation of the value of the asset on that future date; rather, it generally reflects the current market
price of the asset plus the net costs of carrying the underlying asset during the term of the contract.

21. Where the forward contract is entered into at market prices, the parties' obligations are matched
and no cash is exchanged up-front. If the terms of the agreement reflect an "off-market" price,
however, one party will generally pay an up-front premium amount to reflect that differential. At the
close of the forward contract, the parties can settle the contract by exchanging the underlying asset for
the predetermined price, or simply offset their obligations by a net exchange of cash that reflects the
then current price of the underlying asset.

22. A futures contract is a type of forward contract that is traded through an exchange in a
standardised form. The futures exchange serves as a clearinghouse and an intermediate counterparty
between sellers and purchasers of futures contracts. Credit risks are reduced through the exchange by
requiring each party to post daily margin amounts reflecting the fluctuation in value of their positions.
Normally, the futures contract will be closed before delivery of the underlying asset through the entry
of offsetting contracts, so that settlement is effected through a cash payment.

3. Options

23. An option is an agreement which gives the purchaser (or "holder") the right, but not the
obligation, to buy (in the case of a "call option") or to sell (in the case of a "put option") an underlying
asset for a specified price during a certain period or on a specified future date. Options often involve
the future purchase of shares (an "equity warrant"), but they can involve any underlying asset, such as
foreign currency, a stock index, or interest rate futures. The seller (or "writer") of the option receives
an up-front premium in exchange for the obligation to perform at the specified future time. The
pricing of the option reflects the relationship between the current market price of the underlying asset
and the option price, as well as the duration of the agreement and the volatility of the price of the
underlying asset.

C. Multiple Payment Period Derivatives -- Notional Principal Contracts

1. In General

24. Derivative contracts can be written to cover more than one payment period. In general, such
contracts are written with reference to an index applied to a hypothetical (or "notional") amount of
 13

principal, which is not actually exchanged by the parties. These "notional principal contracts" can be
based on a series of forward agreements covering each of the payment periods ("swaps") or on a
series of options agreements ("caps", "collars", and "floors").

2. Swaps

25. A swap is a financial transaction in which two parties agree to make a series of payments to
each other calculated by reference to a notional principal amount that is not in fact exchanged. In an
interest rate swap, for example, one party may agree to pay an amount calculated as a specified fixed
rate of interest on the notional principal amount in exchange for the counterparty's agreement to pay
an amount calculated as a floating rate of interest on the same notional principal amount. The cash
streams received by each party can be used to meet its respective business needs and the transaction
as a whole can be structured to meet the complementary market strengths of both parties.

26. Swaps have been expanded beyond interest rate agreements to reflect a number of categories
of risks and returns in which the cash streams exchanged are calculated by reference to specified
indices. Equity swaps, for example, involve an exchange of payments based on the value of an equity
index or a notional amount of shares in a specific company. Currency swaps involve an exchange of
payment streams denominated in different currencies. Commodity swaps involve the exchange of
payments calculated by reference to notional amounts of a commodity without a physical exchange of
the commodity.
3. Caps/Collars/Floors

27. An agreement consisting of a series of options can provide for multiple payments depending on
movements in an index or in the price of the underlying asset during specified periods. In a "cap"
agreement, for example, the purchaser will receive periodic payments whenever the underlying index
rises above the level specified in the agreement. Similarly, a "floor" agreement will provide the
purchaser with payments whenever such value drops below the specified level. A "collar" agreement
protects both parties through a combination of a cap and a floor; one party makes periodic payments
to the other if the value rises above a specified range, and the counterparty makes periodic payments
if the value drops below the range. For example, an interest rate "collar" could be used by a borrower
of floating rate debt to limit the risk of higher interest rates through the "cap" element of the
agreement. The "collar" may be used to reduce the "cap" premium since the specified "floor" level
will allow the counterparty to share in the benefits if interest rates fall.

4. Equity Derivatives and other Total Return Swaps

28. Certain notional principal contracts, such as equity swaps, are designed to have the general
economic effect of a leveraged purchase of underlying securities. When the index referenced in the
swap consists of a single stock (so that the swap purchaser captures all appreciation value, as well as
dividend equivalent payments, and bears the risk of loss in value) the contract challenges the notion
of whether the holder should be treated as an "owner" of the underlying stock itself for tax purposes.
Such contracts can also be used by the counterparty to offset (or effectively to transfer) the economic
benefits and burdens of equivalent shares that the counterparty continues to hold, similarly raising
 14

ownership questions.

29. Where the swap is based on an index of multiple securities (such as an equity index swap),
there is less clearly a transfer of "ownership" rights in the underlying asset. Such transactions,
however, raise the issue of whether tax laws should have a separate definition of the owner of a
security applicable to single-stock equity swaps vs. equity index swaps.

30. Equity swaps can be combined with caps, collars, and floors so that the instrument conveys only
a specified range of risk with respect to the underlying stock or equity index.

D. Replication and Synthetic Instruments

31. In broad terms, each of the financial transactions described above involves contractual or
quasi-contractual rights to receive and obligations to pay money or money's worth, or to enter into
other financial transactions. The high degree of financial innovation that has occurred in recent years
is exemplified by the many different ways in which these rights and obligations have been structured.

32. At the heart of the innovation is the aggregation and disaggregation of some basic types of
rights and obligations to create new sets of rights and obligations. For example,

! an interest rate swap is economically equivalent to a series of cash settled interest rate
forward contracts

! caps, collars and floors options based on interest rates

! a swaption is an option on a swap.

33. Together with 'physical' securities such as debt and equity, forward contracts and option
contracts form the basic building blocks for financial transactions. The rights and obligations
involved can be combined in various ways with the result that cash flows can, for example, be:

! certain as to amount, timing and direction

! certain as to timing and direction but not amount

! certain as to timing but not direction or amount

! contingent on a specific event.

34. Cash flows can be denominated in a specified foreign currency or calculated by reference to, or
contingent on, variables such as exchange rates, interest rates, commodity prices, equity prices or
indices.

35. Reduction of financial instruments into separate flows of cash payments allows for the
construction of a series of equivalent relationships between them. These equations are simplified for
 15

presentation purposes to ignore the credit risks and transaction costs that would be taken into account
in pricing such instruments in the market. A comparison of such cash flows can be useful, however,
in explaining the relationship between and among more traditional instruments and innovative
financial transactions.

36. Starting with the most traditional instruments, the economic benefits of debt and equity
securities can be expressed in cash flow terms:

Debt = Principal + Interest

Equity = Principal + Dividends

37. The price set in a forward contract to purchase a specific asset is based on the current value of
the asset plus the net carrying costs for maintaining the asset during the term of the contract. For
example, the price for a one-year forward contract to purchase stock X (which has a current market
price of $100) should be equivalent to the net costs for maintaining ownership of the stock for one
year. An investor that borrows $100 and actually purchases stock X will bear the interest costs on the
borrowing (and will receive the benefits of stock X dividends for one year). The following formula
shows the relationship among the three instruments:

Forward Contract = Equity - Debt

Or substituting the cash flows described above for Equity and Debt,

Forward Contract = (Principal +Dividend) - (Principal + Interest)

Forward Contract = Dividend - Interest

38. The above relationship, however, is not the only approach for describing a forward contract. As
described above, a forward contract represents both a right and an obligation to enter into a sale of the
underlying asset. The right to purchase at the forward price is equivalent to the purchase of call
option for the asset exercisable on the forward date at that price. Similarly, the obligation to purchase
the asset is equivalent to the sale of a put option for that date and price. Expressed in terms of an
equation,

Forward Contract = Call Option - Put Option

39. Similar equivalencies can be demonstrated through the fundamental relationship between put
and call options and debt and equity. Under the "put-call parity" theorem, there is a fundamental
relationship between the value of debt, stock, and put and call options with respect to the stock. That
theory leads to the following relationships (expressed in a simplified form):

Equity + Put Option = Debt + Call Option

Equity - Debt = Call Option - Put Option

Finally, this relationship can be rearranged to show the equivalence with the forward contract
 16

equation described earlier in the previous paragraphs:

Equity - Debt = Forward Contract

40. Further examples of these equations are illustrated (again, in simplified form) in Appendix I.
Although application of these equivalent relationships may be complex in practice (depending on the
transaction costs and the particular characteristics of actual securities) the fundamental principle they
demonstrate is simple: each type of financial product, whether classified as debt, equity, forward
contracts, options or combinations thereof, can be structured through positions taken in other types of
instruments. Moreover, in spite of the more complex form required for actual market transactions,
the relationships between forwards, options and physical securities would still hold when real world
costs and characteristics of the instruments are incorporated.

41. This put-call parity framework is central to the pricing methodology for swaps and other
derivatives. As discussed further below, the ability to "replicate" one instrument type through a
combination of others presents a significant challenge to any tax system that bases the income tax
treatment of financial products on traditional classifications.

E. Securities Loans and Repurchase Agreements

42. In a securities lending transaction, the owner of securities (the "lender") transfers them to
another party (the "borrower") subject to a contractual agreement that identical securities will be
returned upon the lender's demand or within a specified time. In a typical transaction, the borrower
seeks the temporary ownership of the securities for use in a separate transaction (for example, to meet
the borrower's obligations in a short sale to another party). Credit risk is generally controlled by the
transfer of collateral subject to daily valuation and margin requirements. At the end of the
arrangement, the borrower purchases and returns to the lender equivalent securities plus a lending fee
in exchange for the return of the collateral. Significantly, during the term of the loan, the borrower
agrees to make "substitute" or "in lieu of" payments that return to the lender the amount of income
that it would have received with respect to the security if the loan had not taken place.

43. In a repurchase agreement, the owner of securities sells them to a counterparty and
simultaneously agrees to repurchase equivalent securities at a later date from the counterparty at a
specified price. As in the case of securities lending, the exchange of a "purchase price" for the
temporary transfer of securities of equivalent value reduces the credit risk for both parties, and the
counterparty agreement to make substitute payments with respect to the income that would have
accrued on the securities during the term of the repurchase arrangement puts the original owner back
in the same position as if ownership had not been transferred. Moreover, the "repurchase price"
normally includes an interest component measuring the benefit of the use of the cash during the
agreement term. If the counterparty receives the right to dispose of the securities in the interim, the
transaction closely resembles the securities loan described above. Where the counterparty does not
have such a right, then the transaction more closely resembles a collateralised financing, with the
"repurchased" securities serving merely as collateral against a loan of cash.
 17

44. Both types of transactions challenge the traditional tax concepts of ownership. The original
owner of the securities generally transfers the securities themselves, but retains through these
contractual arrangements the "benefits and burdens" of economic ownership. A series of loans or
repurchase arrangements can pass such economic attributes of ownership through several
counterparties, and yet only one entity will hold the right to convey legal title to the underlying
securities themselves.

45. In the absence of effective rules addressing these issues, securities lending and repurchase
agreements present significant opportunities for abuse. Obvious examples of the type of transaction
which would not take place at all but for the asymmetries of tax treatment of the participants are
"dividend stripping" and "bond washing". These strategies involve the short-term "sale" of an asset to
an entity which can safely maximise the advantages of receiving income without a tax penalty and the
subsequent return of the asset to the original owner after the dividend/interest date at a price which
reflects at least part of the value of the tax advantage gained by the linked sale and repurchase.

46. Similarly, certain trading strategies using securities loans can terminate a taxpayer's economic
interest in owning the security while possibly deferring the tax costs of an immediate disposition. For
example, a taxpayer who holds specific securities could enter into a short sale of equivalent securities,
but meet the short sale obligations with securities borrowed in a securities loan. The disposition for
tax purposes could be deferred in this case until the borrower closes the securities loan by transferring
the original securities to the counterparty in the securities lending transaction. The borrower thus
enjoys the full sales proceeds from the short sale, and has eliminated the risks of continued ownership
in the securities, but potentially defers recognition of tax on the disposition. This deferral would not
be possible, however, in those countries that recognise the disposition upon the closing of the short
sale.

III. Income Tax Issues and Problems Presented by Innovative Financial Instruments

A. In General

47. Innovative financial instruments raise a number of issues and problems for the development and
administration of an income tax system. They not only present new challenges with respect to the
traditional tax issues of character, amount, timing and source; in addition, they raise wholly new
questions regarding the most basic classifications on which traditional tax rules are based. Moreover,
unique issues are presented by the need to eliminate artificial tax barriers on the legitimate use of risk
management tools, while at the same time reducing potential tax arbitrage opportunities.

48. In some cases, governments have developed specific rules tailored to meet one or more of these
issues. In other cases, the policy options discussed in part V below could be used to address a
combination of these problems. For certain financial transactions, it may not be possible to address
each of these issues in a complete and satisfactory manner, and a trade-off must be accepted. Before
addressing the relevant goals and policy options, however, this section provides a general overview of
the type of issues and problems presented by innovative financial transactions.
 18

B. Traditional Tax Issues

1. Character of Income

49. Many tax systems distinguish between transactions that are entered into in the ordinary course
of a business, which will give rise to "trading" or "ordinary" income or revenue, and those that are
entered into with an investment motive, which will give rise to "capital" treatment. In some cases, the
taxation of capital gain will be benefited by lower rates, deferral or exemption. In contrast, the use of
capital losses to offset ordinary income, as well as income or gains realised in other taxable periods
may be limited.

50. Innovative financial transactions present the traditional issues of the relationship between the
taxpayer and the financial instrument in question to determine the character of any income or loss that
may arise. Character issues can become increasingly important in this context, however, due to the
use of financial instruments as risk management tools. A critical question in this regard is whether
the purpose for which a taxpayer entered into a transaction should affect the treatment and the
character of the income arising from the transaction. In particular, as discussed below, both the
taxpayer and the government have an interest in ensuring that character is preserved between financial
instruments used for hedging purposes and the underlying transactions to which they relate.

51. A similar issue is whether the characterisation of income and loss from financial products
should depend on the economic characteristics of particular instruments, or alternatively, whether all
instruments should be characterised the same, in order to avoid tax arbitrage opportunities. Systems
of income taxation must take into account the ease with which financial products can be tailored to fit
within any particular category while maintaining the same economic characteristics.

2. Timing and Amount of Income

52. In general, the amount and timing of items of income and loss are determined either as the
income and loss accrues over time (the "accruals basis") or at the time the contractual right to
payment becomes fixed (the "realisation" or "cash flow" basis). For example, the interest income and
deduction attributable to a discount bond would be recognised over the term of the instrument under
the accruals basis, but would be deferred until redemption or sale of the bond under the cash flow
basis.
53. Innovative financial instruments present a number of issues in reconciling these two methods
and determining when each should apply in particular contexts. Inadequate timing rules can be found
in several different circumstances. One is that tax may be payable on income that does not represent
the true economic gain because largely offsetting cash flows that will be outlaid in a later income year
are not taken into account. Alternatively, tax relief may be provided in relation to losses that
misrepresent the taxpayer's overall economic position. Another is where, in relation to instruments
such as deep discount securities, not only is payment of tax deferred by the holder of the instrument,
but the issuer is granted tax relief on an accruals basis. Tax deferral, which in itself can result in tax
arbitrage, is compounded by the differential tax timing treatment as between the holder and the issuer.
 19

54. Forward contracts and options may be accounted for on a cash flow basis; since the returns are
considered relatively uncertain, recognition for income tax purposes in such cases is delayed until the
contractual rights become fixed or the instrument is disposed of or terminated. Similarly, since
notional principal contracts are generally constructed as a series of forwards and/or options, they may
be subjected to tax only as the payment rights become fixed. As a result the implicit time value factor
of the "carrying costs" built into the pricing of these derivative instruments is not recognised, and
income tax effects are deferred until the rights to such payments become fixed. In this situation, tax
arbitrage opportunities may arise where the combination of a derivative and the underlying asset
produces relatively certain returns. For example, where a taxpayer borrows funds to purchase an
asset, and also enters into a forward contract to sell that asset in the future, the transactions may result
in current interest deductions in respect of the borrowing, while the offsetting "gain" on the
subsequent sale (i.e. the difference between the current market price paid for the asset and the higher
forward sale price for the asset) may be treated as a deferred capital gain.

55. If significant payments called for by a derivative instrument are made up-front, "back loaded,"
or required to be paid at irregular intervals, the agreement could contain an implicit or "embedded"
loan. Similarly, a "deep in the money" option may actually constitute a loan of the premium amount
which is relatively certain to be returned when the option is exercised. This can present a substantial
timing issue for tax systems that recognise interest on an accruals basis, but apply cash basis
accounting to derivative instruments.

56. Timing issues are also presented by the need to determine when an instrument has been
disposed of. For example, modification or extension of a contractual arrangement may or may not be
sufficient to trigger taxation. Similarly, tax rules must address the circumstances under which
entering into offsetting contracts will be treated as a termination for tax purposes.

57. Innovative financial transactions present other unique challenges with respect to when a
transaction should be treated as closed, or when a security should be treated as transferred or disposed
of. Securities loans and repurchase agreements are generally not treated as realisation events, since
the transferor is not viewed as significantly altering its economic ownership rights. Where the
taxpayer continues to hold property that is equivalent to the "underlying assets," however, a securities
borrowing to meet a short sale obligation can transfer away the economic burdens and benefits of
ownership. Similarly, an equity swap can be used to alter the economic risks of continuing to own the
underlying stock. These transactions present the issue of whether timing rules should follow
traditional concepts of ownership for tax purposes, or yield to economic attributes.

58. As discussed below, adoption of either hedging or mark-to-market regimes within a tax system
will alter the above timing rules in order to meet particular policy objectives. This variation from the
normal rules with respect to timing and amount of recognised income and loss highlights the need to
carefully delineate the application of these alternative regimes to prevent the retroactive selection of
the most favourable tax result by the taxpayer.
 20

3. Source of Income

59. Source rules are critical for purposes of determining whether a particular jurisdiction is
permitted to impose withholding taxes on specific cross-border payments, and whether a jurisdiction
is required to provide double taxation relief through credits or exemptions on income received by its
residents. Different types of innovative transactions have raised new challenges to the ways source
rules apply.

60. In general, tax authorities have not imposed withholding taxes on income from derivative
financial instruments. This reflects a general recognition that such gross basis taxation would be
inappropriate. Since the offsetting contractual obligations generally result in net payments flowing in
one or both directions, cross-border withholding on a gross basis would be impractical to administer
and would destroy the economic relationship underlying the contract. Moreover, given the mobility
of these contracts, withholding taxes would not be an effective means of raising revenue. There is
variation, however, in how specific tax systems reach this result. Some adopt source rules that
exclude derivative instrument income from gross basis withholding, while others simply classify the
transactions as not being within the withholding tax rules.

61. Securities lending transactions and repurchase agreements present unique issues with respect to
the taxation at source of substitute or "in lieu of " payments. Since these transactions "replace" the
economic benefits of the income on the transferred securities during the term of the agreement, the
question is presented whether the source and withholding rules should apply to these payments in the
same manner as the income they replace. For example, a "look-through" approach would reduce the
potential for abusive dividend washing schemes by sourcing the substitute payments for withholding
and credit purposes in the same manner as the underlying income.

4. Treaty Classification

62. The tax issues presented by each the above traditional issues are further modified by the
application of treaty provisions. The initial application and type of treaty benefits may be determined
based on the classification of a particular payment as "dividend", "interest", or "other" income.
Definitions of the types of payments that receive particular treaty benefits are often provided within
the treaty itself. In the absence of specific definitions, the domestic law of the country applying the
treaty generally determines the treaty classification of the payments at issue. Moreover, the character
of a particular payment as "ordinary" or "capital", as well as its timing and source, can affect whether
the income in question is afforded treaty benefits.
63. Each of the above distinctions are subject to challenges presented by innovative financial
products. The tailored reproduction and repackaging of specific economic attributes afforded by
derivative transactions may allow the parties to select whether the payments fit within the parameters
of specific treaty provisions. In particular, the treatment of income from derivative transactions as
"other income" (that is potentially exempt from withholding tax) may allow taxpayers to obtain tax
exemption under treaties for payments that bear the same or similar economic attributes of income
that would otherwise be subject to tax.
 21

C. Instrument Classification

1. Debt/Equity/Derivative Distinctions

64. Because tax systems have traditionally relied on categories of transactions, it frequently will
be necessary to determine what category of transaction an instrument falls under before addressing
any other tax issues. Under traditional tax systems, there are a number of fundamental differences
between the tax treatment of instruments depending on whether they are classified as debt or equity.
Some of these distinctions include the following:

! interest is deductible but dividends are not

! interest is taxable but some dividends get concessional treatment, including
qualification for favoured treatment under integration/imputation systems
! interest may accrue for tax purposes, but dividends generally are taxed on realisation
! there is greater likelihood that disposal of debt will be on revenue account and disposal
of equity on capital account
! interest withholding tax and dividend withholding tax rates differ, with interest
generally being subject to lower rates or exemption

65. In general, the effect of these differences favour the classification of instruments as debt,
although this presumption is less clear where a corporate integration system applies.

66. The expanded use of derivative instruments has introduced a new category of transactions, and
substantially increases these classification problems. As discussed above, derivative instruments may
be taxed for some purposes like debt or equity or a combination of the two. In particular, recognition
of income and loss for derivative contracts may be on the cash flow basis, and the payments are
generally exempt from withholding taxes. In contrast, these disparities could be reduced by adopting
a uniform system in certain contexts regardless of the instrument type (for example, as discussed
below, by applying a mark-to-market system for all instruments held by financial dealers).

67. These distinctions lead to significant issues of how to place particular transactions within one
of the debt/equity/derivative classifications. There is difficulty in classifying financial instruments
because:

! there are many different forms of instruments

! some have complex structures
! at the margin, the dividing line between primary instruments (debt and equity) and
their derivatives (futures, options, swaps, forward rate agreements, etc.) is not always
clear.

68. One way of explaining the existence of the distinction, without making the distinction much
clearer, is by reference to the degree of certainty attached to the direction, timing and amount of cash
flows. Debt is traditionally described as providing an investor with cash flow rights that are certain as
far as the amount invested is concerned. Thus, an example of cash flow rights that are relatively
 22

certain include those found in zero coupon bonds, debentures and floating rate notes. Of course, the
cash flow rights in respect of the excess above the amount invested can vary -- they can be fixed, or
can be contingent on an external interest rate variable.
69. Examples of cash flow rights whose value and direction is relatively uncertain include arm's
length cash settled forward transactions and typical interest rate swaps (where payment on the floating
leg is netted against payment on fixed leg). In an option, the value on issue of the holder's right to a
future cash flow is typically uncertain unless it is quite deep in the money. In these cases, the
payments are generally wholly contingent on external variables.

70. In contrast, dividends (payable at the discretion of non-controlled directors) and ordinary
shares are examples where there is no right to a cash flow as a result of issue of the transaction.

71. The development of innovative financial products, however, shows that each of these
elements can be structured with varying levels of risks. Thus the risk/return distinction on which
these classifications are based is more in the nature of a continuum. Moreover, as discussed in the
next section, the introduction of new combinations that replicate instruments from one classification
using building blocks from another one, brings into question the very foundation of these distinctions.

2. Synthetics and Hybrids

72. Perhaps the most fundamental challenge to tax systems from the use of innovative financial
instruments is their ability to replicate the economic properties of other instruments. A combination
of financial products that closely reproduces the economic attributes of an existing instrument (but
that may be classified differently for tax or regulatory purposes) is generally referred to as a
"synthetic" form of the existing instrument. Similar issues are presented by transactions that combine
only selected attributes of different instruments to create new "hybrid" forms of financial transactions.

73. To the extent such strategies can be implemented (given applicable transactions costs and
counterparty credit risk), they present the possibility that any distinction for tax purposes between
traditional classifications will be fully elective for taxpayers. In short, instruments can be designed to
replicate any desired set of economic attributes, but tax attributes could be independently crafted into
the instrument to obtain the most advantageous results.

74. These potential advantages would be particularly clear where the parties to the synthetic or
hybrid instrument were not subject to offsetting tax constraints. For example, if one of the parties
was a tax-exempt entity or had losses that could fully absorb current income, it would be easier to
implement a transaction that shifted tax benefits to the other party. Similarly, where the rules of two
jurisdictions classify the transaction differently (or offer different forms of tax advantage),
cross-border instruments can be structured to maximise tax advantages with respect to both systems.

D. Breakdown of Ownership Concepts

75. Determining ownership for tax purposes is a critical threshold issue for many purposes.
Income or gains (either capital or ordinary) will generally be recognised only upon a transfer of
 23

ownership. In cross-border transactions, the imposition of withholding tax on income will be based
on the identity and residence of the "beneficial owner" of the income in question. Moreover, owners
of certain securities may qualify for benefits, either domestically or pursuant to tax treaties, such as
tax-exempt or tax-favoured income, credits and deductions.

76. Common law systems traditionally identify the owner of an asset for tax purposes as the
person that bears the economic burdens and benefits of ownership. In the case of financial assets, the
significant factors in this regard are (i) which party has the right to dispose of the asset; and (ii) which
party bears the risk of profit or loss with respect to the asset. The rule can be more complicated in
some civil law jurisdictions where it is possible to create and sell partial interest in property (such as a
usufruct).

77. Innovative financial transactions present significant challenges to these traditional views of
tax ownership. Derivative instruments offer the ability to receive or to transfer the benefits or
burdens of economic ownership with respect to a specified "underlying" security or other asset,
regardless of the counterparties' legal ownership of the asset. In short, these instruments are capable
of separating the attributes of risk and reward from legal ownership.

78. If tax systems respond to this challenge by treating the economic indicators of risk and reward
as being equivalent to ownership for tax purposes, the result would be to accept the existence of
multiple owners for a single piece of property. This could lead to further arbitrage opportunities in
any case where the income from such property were tax-favoured. For example, in a securities
lending transaction involving the transfer of tax-exempt bonds, the substitute payments returned to
the lender are the economic equivalent of the underlying interest flows; but if each lender in a chain
of securities loans is classified as the "owner" of the bond for tax purposes, then the tax-exempt
interest benefits would be multiplied. For this reason, domestic tax rules generally clarify that only a
single recipient of tax-favoured income is permitted to obtain the beneficial treatment, and other
parties that may receive "equivalent" payments (or payments measured by reference to such amounts)
do not receive the same tax benefits.
79. Accordingly, any attempt to address this breakdown of traditional "ownership" concepts will
require a balanced approach to prevent abuses that can occur under either broad or narrow definitions
of when a taxpayer is recognised as "owning" an asset.

E. Matching Positions for Effective Risk Management

80. The marketplace developments in the last twenty years, described in section (I) above,
demonstrate a growing need for effective risk management. There has been a corresponding
acceptance among tax authorities that, to the extent possible, tax rules should not impose a barrier to
effective hedging strategies. This development, however, leads to the principle that tax rules can, and
in some circumstances should, treat taxpayers differently with respect to the same transaction
depending on the purpose of the transaction. In particular, to hedge an underlying asset or liability
with one or more financial instruments, the offsetting tax attributes of the hedging instruments must
be "matched" with that underlying position.
 24

81. On the other hand, where certain positions are offsetting, the failure to take such economic
effects into account presents opportunities for abuse, allowing taxpayers to choose the most
advantageous application of character, timing and source rules.

82. As discussed in Part V below, one general approach to these problems is to permit taxpayers to
"match" offsetting positions through elective hedging regimes. In addition, it may be appropriate in
some circumstance for the tax system itself to impose a linkage between offsetting positions in order
to more clearly apply specific tax attributes to related transactions. Yet any approach that "integrates"
discrete transactions, whether through an elective or mandatory system, presents a number of
significant and difficult issues:

! What type of activities should the regime apply to? For example, should a voluntary
system permit hedging of "investment" activities, or should it be limited to "business" or
"inventory" activities? How should such distinctions be defined and administered?

! Should the regime apply to net hedges of an aggregate risk of interest rate, price changes
and/or currency risks, or should the linkage require identification of separate transactions?
If net hedges are permitted, how will the rules establish that the "overall risks" of the
taxpayer have been effectively reduced through the hedge?

! May a transaction qualify as a hedging transaction if it hedges an anticipated risk? If so,

how will the system address cases where the future risk does not materialise?

! How should the system treat the linked transactions if they are not each entered into and
disposed of at the same time?

! Where the integration system is elective, will the taxpayer be required to identify
transactions as hedges in order to treat them as such? If so, when and how? Must the
identification be made on the basis of individual transactions or by identifying particular
classes of transactions? Will there be penalties for misidentification to prevent abuses of
the system?

83. An alternative to elective hedging regimes is to define the tax treatment of specific derivative
instruments in a uniform manner based on how the instruments are most often used for hedging
purposes. As discussed in Part V, this more generalised approach may help to produce similar results
as elective hedging without presenting the problems described above.

F. Tax Arbitrage

84. Tax arbitrage involves taking advantage of differences in the tax treatment, either of persons
undertaking transactions or of the transactions themselves, where these differences arise either as a
result of the asymmetries within or between tax systems or as a consequence of the specific tax
position of the persons themselves. In many cases, tax arbitrage transactions are entered into merely
 25

to obtain tax benefits, and have no significant business or financial effects apart from such benefits.
Accordingly, the term broadly refers to any type of abusive transaction that may be available with
respect to each of the categories of issues described above in this section. If the tax system does not
include adequate rules to prevent such abuse, taxpayers can construct arbitrage transactions to
maximise the tax benefits of character, timing, source, ownership, or hedging rules.

85. It is clear that innovative financial transactions have expanded the opportunities for tax
arbitrage. Particular concerns are presented by synthetic and hybrid instruments that carry tax
attributes of one classification of investment, but provide the economic risks and returns of a different
classification.

86. Arguably, tax authorities should be concerned where transactions are entered into, not for
genuine commercial reasons, but solely to obtain tax arbitrage benefits available either within an
individual system or through cross-border transactions between tax systems. Tax authorities will also
be concerned by potential distortion effects where cross-border movement of funds is designed by the
parties to share tax benefits from asymmetries between jurisdictions.

IV. General Tax Policy Goals

87. Since innovative financial transactions are essential risk management tools of modern business,
systems of taxation must strive to reduce or eliminate obstacles to their effective use. On the other
hand, tax authorities have a strong interest in reducing the opportunities for tax arbitrage presented by
these transactions. It is perhaps a moot point as to whether rules of taxation with respect to such
transactions should facilitate market efficiency and financial innovation rather than narrowing the
scope for tax avoidance. Both overall goals must be taken into account.

88. A tax system should raise revenue as neutrally, simply and equitably as possible. The selection
of rules for imposing tax in this context will depend on the weight given to each of the following
policy goals:

89. Neutrality and Equity. Tax neutrality fundamentally requires that transactions with the same
economic substance attract the same tax treatment. Non-neutrality is evident when tax laws are based
on the legal form rather than the economic substance of the transaction. This is particularly so for
financial transactions, which can be easily structured in different ways, without altering their non-tax
economic effects. Asymmetries and mismatch have the potential to cause significant threats to the
revenue and inefficiencies in the way finance decisions are made. A more neutral tax law in this area
therefore both promotes the efficiency of the financial markets and protects the revenue base.
Neutrality also promotes the related goal of equity, by ensuring that tax burdens are imposed based on
the parties that realise the economic benefits of the income at issue. In the context of financial
products, an equitable system of taxation will select the appropriate party as the "owner" of particular
items of income, and will ensure that the measure and timing of the tax reflects the underlying
economic results of each transaction as closely as possible. On the other hand, equitable application
of the rules also is required to limit the ability of parties to avoid or defer tax through arbitrage and
other strategies.
 26

90. Simplicity and Lower Compliance Costs. The goal of simplicity should take into account the
ease or difficulty in administering the law as well the taxpayers' costs in complying with it.
Compliance by taxpayers is affected by the degree of certainty in the law and the clarity of the
concepts. Cost of compliance is also affected by the amount and type of information that taxpayers
must be keep or produce to meet tax obligations, as well as the degree of difficulty for tax authorities
to verify taxpayers' claims.

91. Certainty. When entering into transactions, taxpayers have an interest in knowing the tax
results in advance. Economic efficiency can be harmed if transactions are avoided or financial
products are not developed because of the uncertainty of the tax treatment. Uncertainty can also
facilitate avoidance or deferral through transactions which take best advantage of different
interpretations of the rules, allowing for timing mismatches or differential tax treatments between
transactions that are economically similar but different in strict legal form. In addition, uncertainty
about the application of tax laws can lead to greater levels of controversy between taxpayers and
revenue authorities.

92. Robustness. Particularly in the case of financial products that evolve at an increasingly rapid
rate, tax rules must be both broad and flexible in order to address new transactions and variations of
existing transactions. To the extent the rules are crafted to reflect the general economic structure of
the transactions at issue, they are more likely to impose an appropriate level of income tax on the
parties with respect to new transactions as well. The ideal is a set of broad rules that are sufficiently
flexible to reflect market activity and facilitate innovation, yet resistant to easy manipulation for
taxation advantage.

93. These varying goals inevitably require trade-offs. For example, "bright-line" tests are simple
and administerable, but, because they are arbitrary, may not serve the goals of neutrality and
robustness. Similarly, tests based on a broad reliance on the facts and circumstances provide
flexibility and are harder to avoid, but may not provide sufficient certainty.

94. Although each country's approach to the taxation of innovative financial transactions will be
determined by many factors, the usefulness of those approaches can be evaluated by considering how
well they meet each of the above goals. As discussed in the following section, the tax policy options
currently used or considered for addressing the taxation of innovative financial transactions illustrate
the difficult issues and inevitable compromises that must be accepted in this area.

V. Policy Options

A. General Approaches vs. Specific Rules

95. In recent years, it has become increasingly clear that if the tax treatment of financial products
depends on the promulgation of specific rules, it will be difficult if not impossible for tax authorities
to meet the policy goals described above. In particular, the inherent time lag between development of
new products and introduction of targeted rules both permits abuses of the system to go unchecked,
 27

and denies taxpayers the certainty they require for legitimate transactions.

96. In light of such difficulties, it may be necessary to consider more general approaches other than
the proliferation of discrete rules to deal with specific new products. Jurisdictions with flexible
taxation systems, or those with well-developed anti-abuse rules, may be at an advantage as compared
to systems in which the government is less able to attack abusive transactions on the basis of
substance-over-form or similar doctrines. In some countries, it may be possible to reach consensus on
appropriate treatment of new transactions simply by encouraging taxpayers to enter into a dialogue
with governmental experts.

97. Attempts to prevent "tax arbitrage" between jurisdictions through harmonisation of countries=
approaches to the taxation of innovative financial transactions will have to take into account not only
the differences in legislation in each country, but differing approaches to the administration of the tax
law. Again, development of international norms in this context, either through model treaty
provisions or bilateral negotiations, will require greater acceptance of general policy approaches
rather than adoption of narrow rules targeted at specific instruments. In particular, opportunities for
tax arbitrage may be reduced, and the speed with which rules are promulgated can be increased,
through more widespread international agreement on a theoretical basis for the taxation of financial
products and administrative co-ordination on these issues amongst tax authorities.

98. The remaining sections of this part examine some of the policy options that have been applied
by governments and discussed by commentators in addressing the challenges of innovative financial
transactions. Each approach carries its own problems and shortcomings, and none provides a
comprehensive solution to the issues and policy goals described above. In short, the problems
presented by new financial instruments often relate to underlying weaknesses of the traditional
systems of taxation that have simply been more clearly exposed by innovative transactions. For
example, arbitrary differences in character or timing under traditional tax systems may result in
transactions with similar or identical economic characteristics being subjected to significantly
different tax treatment. It is hoped, however, that further consideration of such general policy
approaches will lead to both a re-examination of those fundamental weaknesses, as well as assist tax
systems in developing more effective and equitable measures of addressing these problems.

B. Reliance on Financial Accounting Rules

99. Some countries follow accounting standards and principles closely in determining the amount
and timing of income subject to tax. In those cases, the accounting and tax treatment of innovative
financial products present one unified set of issues, and their resolution must balance the concerns of
both tax policy and accounting standards. One advantage of this approach is that it simplifies
compliance for both tax authorities and taxpayers. Also some countries consider that their accounting
standards reflect "real world" practices and have thus kept pace with the developments in financial
innovation. A further advantage can be that where tax laws follow accounting standards and
principles they may more easily cope with the development of new financial instruments.

100. In countries where taxation rules are largely independent of financial accounting rules, some
 28

have advocated legislative change that would allow taxation of innovative financial transactions to be
determined on the same basis as for financial reporting purposes. The advantages of doing so would
be as described above. The disadvantages for such countries may be:

! Accounting standards for financial arrangements may not be well settled in such
countries.

! The application of general accounting principles in particular circumstances may fail to

provide sufficiently objective, consistent and verifiable results necessary for an
administerable and robust tax system. In particular, accounting principles do not
necessarily reduce the risk that similar instruments would be given different tax
treatments on the basis of disparate interpretations of accounting rules.

! Financial accounts are designed to provide information that bears on users' needs in
relation to economic decision making and accountability. There may be a variety of users
(e.g. shareholders, creditors, regulators), which can affect what information is provided
and how it is provided. This has often been reflected in a principle of conservatism for
financial and regulatory reporting purposes which is not necessarily a fundamental
principle of tax accounting. In contrast, tax laws require a single, statutorily derived
figure to determine liability to tax.

! Accounting standards have in some circumstances also failed to keep pace with
innovative financial transactions or lagged behind tax rules. For example, the
development of new instruments has transformed conventional balance sheet analysis as
instruments are unrecognised in the balance sheet or transactions go "off balance sheet".
While some progress has been made by accounting standard-setting bodies to improve
disclosure in this area, the accounting guidance is still not specific enough to eliminate
the problems described above.

101. Differences in accounting principles between countries may lead to different tax treatment of
transactions even where tax rules follow those principles. But there will also be different tax
treatments where countries operate specific tax rules with regard to particular transactions.

102. Financial accounting principles and standards may provide valuable guidance for some
countries in formulating their rules for taxation of financial transactions. The financial accounting
principles of consistency, substance over form and recognising revenues and costs in the period in
which they relate reflect sound taxation policy. Moreover, any reduction in compliance costs from
harmonising tax and financial accounting records would be a desirable outcome. Therefore, to the
extent specific, consistent and objective rules can be developed as appropriate accounting standards
for innovative financial transactions, the same solutions might be adopted for income tax purposes as
well.

C. Bifurcation
 29

1. In General

103. Where a financial right or obligation is taxed differently according to whether it is issued as a
separate instrument or is embedded in another instrument, it is arguable that there is tax
non-neutrality. In effect, the combination of different financial instruments into the one composite
transaction obscures the essential features of the separate components. Therefore, one approach
considered by tax authorities is to "bifurcate" (or more accurately, disaggregate) each of the
component parts into independent economic units, and to subject each part to appropriate tax rules
separately.

104. An example that is often cited in this regard is a convertible bond that carries the right of
conversion into the stock of the issuer or a party related to the issuer. Economically, this instrument
consists of a debt instrument issued at a discount and carrying below market rate interest coupons,
coupled with an equity option. The fusing of the two components may obscure the discount element,
which arguably should be treated for tax purposes as additional interest. Non-neutrality occurs where
the discount would have accrued if the debt component had been issued separately.

105. The different tax treatment of like instruments is a fundamental driving force for bifurcation.
By breaking the transaction down into its component parts, bifurcation can reveal time value of
money elements of transactions that are not being accrued. This leaves the exposed elements to be
taxed in a consistent manner. Two simple examples of this approach are provided in Appendix II.

106. The theory for bifurcation in exposing the various components of financial instruments is that,
if issued separately, each would be taxed differently. This rationale points to the main type of
financial instrument for which bifurcation could be considered. Where the timing, character, or other
tax attributes would be different for each of the underlying components, it would arguably be
appropriate to tax each element separately under its respective rules. For example, in a system that
taxed debt on an accruals basis, while options and forwards are taxed on a cash flow basis, the
question whether or not to bifurcate would be raised in an instrument that has both debt and equity
features, or both debt and option or forward features.

2. Problems with Bifurcation

107. Although bifurcation arguably promotes consistency in tax systems that differentiate between
debt, equity and derivatives, it raises its own questions and difficulties. Essentially, the arguments
against bifurcation are twofold:

! the economic substance argument in favour of bifurcation is illusory

! the benefits of bifurcation are outweighed by the costs.

108. Since all financial transactions can be replicated by aggregating or disaggregating other
financial transactions, it is arguably pointless to disaggregate compound instruments into their
component parts. In short, each of the component parts in themselves can be replicated by other
transactions. Given that there are no fundamental individual particles and no unique financial
 30

transactions, any system of bifurcation will be arbitrary.

109. The argument taken to the extreme, though, belies the fact of tax differentiation. If the different
tax treatment of debt, equity and derivative is to continue, notwithstanding the murkiness of the
boundaries, the question is what effort should be made to preserve those boundaries. Also, the
categorisation or classification of transactions that are not close to the boundaries may be relatively
straightforward.

110. Other arguments against bifurcation include:

! The non-accrual of the discount in a compound instrument approximates the deduction

the holder should get for the option premium. However, it is by no means clear that such
an outcome for the holder will always be consistent with the economic substance of the
transaction.

! Bifurcation ignores a synergism between the option (or other non-debt) and debt portions
of the compound instrument. The argument is that bifurcation may wrongly value the
separate portions. The question is whether any distortion in this regard is greater than the
distortion of not bifurcating.

! Bifurcating a convertible security and taxing the discount on an accruals basis assumes
that there will be redemption, i.e., no conversion. It is argued that there is an assumption
that the holder will obtain back the face value amount through redemption, whereas in
fact the holder may exercise the option to convert into equity and therefore not obtain
back the face value. On the other hand, it can be argued that even if the holder does
convert, it will be obtaining the value of the face amount through the equity it obtains
(and can sell for value). Otherwise, the holder would not convert; thus in fact the holder
is guaranteed from the outset a minimum of the face value.

3. Practical Issues

111. It is sometimes argued that, particularly for holders, the bifurcation process of determining the
net present value of the debt component is too complex to justify a departure from financial
accounting. Given that the instruments themselves are reasonably sophisticated and are effectively
issued at a discount, there is an issue of whether bifurcation calculations are any more complex than
deep discount legislation that might apply to the instruments if the discounted security portion was
acquired separately.

112. Another practical difficulty is determining the yield on a comparable noncontingent debt
obligation issued by a comparable issuer. This will not always be easy and will often involve
judgement. One possibility may be to use a benchmark rate of the government borrowing rate plus a
risk premium, calculated by reference to certain factors such as the issuer's credit rating.

4. Alternative Methodology: Estimating the Yield

 31

113. The approach described above views the contingent payments under the contingent component
(such as the embedded option) as being quite different from the noncontingent payments under the
debt component. An alternative approach for some contingent securities is to view the difference as
one of degree, not requiring separate tax rules for the two components, but requiring instead that
estimates be made of the payment in each income period under the security. This can be used to
determine an estimated yield for accrual purposes. The estimate would be revised each income
period.

114. The estimates need to be reasonable and it may be necessary to provide rules for different
situations, e.g., where the unknown payments are based on the change in an index or variable, as
distinct from the amount of an index or variable. Estimates can be problematic where there is no
market index, variable, value etc.

D. Development of Elective Hedging Systems

1. The Demand for Risk Management

115. Broadly defined, "hedging" is any action taken to reduce or eliminate risk. An entity may enter
into a transaction to hedge its exposure to changes in interest rates, exchange rates or to prices in any
underlying asset or liability. A simple but common example of hedging is the use of a
fixed-to-floating interest rate swap to reduce the risk of adverse interest rate movements. Appendix
III provides additional illustrations of the use of hedging contracts to reduce business and investment
risks.

116. As previously discussed, a driving force behind financial innovation is the increasing
sophistication of (and demand for) risk management. Entities are better able to isolate the risk that
they wish to reduce or eliminate. Aggregation and disaggregation of financial instruments provides
the means for more precise and effective matching between the risk and the risk management
instrument.

117. The flexibility accorded by financial innovation facilitates the separation of sourcing of funds
from the management of the relevant exposures, in order to seek cheaper finance or higher yields.
Swap driven funding has sometimes been suggested as an example of this, the idea of which is that
the comparative advantage that two parties have in different markets is arbitraged through a swap and
captured in the lower effective funding cost or higher yield.

2. The Need for Effective Tax Rules

118. Where the income tax system fails to "match" the relevant tax attributes of the hedging
instrument to the related transaction, the hedge may be ineffective on an after-tax basis. The
mismatch can take different forms, depending on the tax treatment of the hedge and hedged positions.
 32

An example of a mismatch would be where the loss on the hedge is of a capital nature, while the
correlated gain on the hedged position is of an income or revenue nature. If capital losses can only be
offset against capital gains, the asymmetry is to the disadvantage of the taxpayer.

119. Another example would be where a gain on a hedge is taxed on a market value basis, and the
loss on the hedged position is deductible on a realisation basis (and both hedge and hedged position
straddle two income years). The taxing of the unrealised market value gain may result in cash flow
problems for the taxpayer. Moreover, the gain may in fact fail to eventuate in subsequent income
years.

120. Tax mismatches can cause undesirable fluctuations in taxable income; if the potential
fluctuations are very severe, companies may be discouraged from undertaking particular types of
hedge activity.

3. Characteristics of Elective Hedging Systems

121. Hedging rules have been adopted by tax systems in response to this inconsistency, and seek to
match the timing and character of the gain or loss on the hedging transaction with that of the hedged
item. Thus a hedging system often involves changing the character of the hedging instruments and
the deferral of gains and losses on hedging instruments from the period in which they occur - and
would otherwise be recognised - to a later period when the related gains and losses on the items being
hedged are recognised.

122. One of the difficult issues for an elective hedging system is ensuring that it applies only to
legitimate and genuine cases of risk reduction. As an initial matter, the tax rules must determine
whether the hedging system will apply only to, for example, inventory purchases, to investments
related to business activities, or to any financial investments.

123. In addition, tax authorities may require the taxpayer to maintain books and records identifying
hedging transactions (generally on a contemporaneous basis) and containing whatever more specific
identification is needed to verify the application of the taxpayer's tax accounting method. The onus of
correctly identifying hedging transactions should lie on the taxpayer, and the audit trail from hedging
transaction to hedged items should be made clear. Consideration should also be given to the
consequences of misidentification of a hedging transaction, in order to encourage correct
identification. Measures such as these assist in dealing with the inherent difficulty in determining the
subjective intent of the taxpayer in entering into a transaction; in effect they make the assessment
more objective than it would otherwise be.

124. Some form of test is also required to ensure that the hedging instrument is in fact used for risk
reduction purposes. Thus some degree of correlation must be documented between the risks inherent
in the underlying asset or liability, and the offsetting nature of the hedging instrument. In this regard,
it is sometimes argued that account must be taken of the overall "enterprise risk" of the taxpayer, with
consideration of all of the outstanding positions currently held. That is because a risk in respect of an
individual transaction may be offset by an equal and opposite exposure from another transaction of
 33

the enterprise. To attempt to hedge either of these separate offsetting positions would increase, not
decrease, the overall risk of the enterprise.

125. On the other hand, in the case of a large decentralised taxpayer, identification and matching of
all outstanding positions in order to correlate offsetting risks may prove to be an onerous requirement.
Moreover, such rules could require auditors to understand the sophisticated hedging strategies,
including portfolio hedging, employed by companies. However, it can be seen that record keeping
would be the key to acceptance of hedge rules. While it may impose further compliance burdens on
taxpayers, this is a necessary consequence of purpose-based rules.

126. Timing issues are presented where the hedging instruments or the underlying position are not
entered into and/or terminated at the same time. Clearly, the hedging system should not apply during
periods in which both positions are not in place; but transition questions must be considered by tax
systems that permit hedging. One general approach for the early disposition of one of the positions is
to mark to market the unrealised gain or loss on the position that is not disposed of, thus permitting an
effective matching of the gain or loss on both items at the time the offsetting position ceases to be in
place.

4. Problems of Scope -- Portfolio and Anticipatory Hedging

127. Rather than hedging a single, underlying transaction with a hedging instrument, some entities
seek to hedge the net exposure of a portfolio of assets, liabilities and transactions. By reducing the
number of hedge transactions, transaction costs are reduced. Also, aggregating cash flows and using a
portfolio technique such as duration analysis (the time weighted average of future cash flows) to
determine the net exposure, enables non-generic derivatives which are difficult to match on a
one-to-one basis, to be hedged more effectively. These hedging strategies, which traditionally have
been employed by financial institutions, have now been adopted by many multinational business
operations.

128. Nevertheless, some tax authorities have been reluctant to recognise portfolio hedging for tax
purposes (although it should be noted that a country that requires an enterprise risk reduction test
effectively requires a portfolio analysis to be performed). Net or portfolio hedge treatment raises the
question about the administrative ability to track the hedging instrument to the various hedged items.
This is particularly the case where the net exposure is managed or adjusted frequently.

129. One reason for this reluctance to expand the scope of hedging is to ensure that the hedging
instrument is entered into for the reason that is claimed, namely to hedge the hedged item --
one-to-one identification between the hedging instrument and the hedged item makes it much easier
to demonstrate that the hedging instrument reduces the risk of loss on the hedged item. A further
advantage of linking the hedging instrument to a specific hedged item is that it provides a basis for
spreading the gain or loss on the hedging instrument. Given that there are many hedged items in
portfolio hedging, presumably with different maturities, it is difficult to determine the period over
which the gain or loss on the hedging instrument should be spread.

130. Where portfolio hedging is undertaken, the artificiality of identification of specific underlying
 34

transactions means that safeguards such as contemporaneous designation are especially important.
Transparency of the trail from the internal decision maker (e.g., the asset/liability management
committee responsible for the hedge program) to the dealer, together with a demonstration of how any
given transaction seeks to reduce net exposure, should be contemplated. The point of the latter
requirement would not be to track each underlying transaction back from the hedging instrument, but
to obtain assurance that the hedging instrument followed an analysis of the net exposure comprising
the underlying transactions.

131. Difficult issues are also presented by the extension of tax accounting systems to permit hedging
of future anticipated risks. On the one hand, there is little commercial distinction between exposure
that an entity is subject to as a result of a transaction that it is contractually committed to, and one that
it is all but contractually committed to. For example, projections of future inventory purchases may
be a relatively certain obligation of the business, even in the absence of binding contracts. On the
other hand, reluctance to extend hedging rules for tax purposes to anticipated transactions probably
stems from recognition of the opportunity presented for deferral where the entity has the power to
decide whether or not to enter into the future underlying transaction.

5. Partial vs. Full Integration

132. In general, hedging rules for tax purposes perform a "partial integration" function, by simply
linking and matching specified tax attributes (such as timing and character) among selected
instruments. Alternatively, a more comprehensive hedging system would permit "full integration" of
the offsetting positions. Where full integration applies, the taxpayer would be allowed to amalgamate
or combine the cash flows of the hedged instruments into one synthetic transaction for tax purposes.

Example

A company uses 6 month forward foreign exchange contracts on a rolling basis to hedge the
foreign exchange exposure of holding shares in a foreign company. Assume the shares are on capital
account for tax purposes and are taxed on a realisation basis. Hedging rules reflecting a "partial
integration" approach could provide that any gain or loss on the hedge contracts are on capital account
or taxed on a realisation basis or both. A "full integration" approach would be to adjust the cost base
of the shares by the gain or loss on the hedge contracts. This would have the effect of deferring past
realisation the gain or loss on all hedge contracts terminated before the underlying shares are disposed
of.

E. Non-Elective Integration

1. Use of Integration to Prevent Tax Avoidance

133. The use of "integration" techniques as a mandatory rule in specified circumstances has also
been considered as an approach for dealing with the problems presented by certain innovative
financial transactions. This is because the taxation of an offsetting hedge on a non-hedge realisation
basis can in some cases lead to abuses that are detrimental to the revenue.
 35

Example

A company X invests in a long term foreign currency denominated zero coupon bond. The
foreign currency is strong relative to its domestic currency. To compensate for the anticipated
appreciation, the instrument has an interest rate yield that is lower than equivalent instruments
denominated in the domestic currency. Assume that the discount on the instrument is taxed on a
constant yield to maturity accruals basis, using the spot rate when the bond is acquired. If the foreign
currency appreciates against the domestic currency to the extent expected, there will be an exchange
gain on maturity of the bond. Assume the gain is taxed on a realisation basis. By comparison with a
domestic currency denominated bond, the foreign currency bond will be taxed at a lower effective tax
rate. This is because the accrued discount is lower and the compensating gain is deferred. If,
however, the company removes the exchange exposure with a hedge that is also taxed on a realisation
basis -- both the gain and offsetting loss will be deferred until realisation. Alternatively, the taxpayer
could borrow in a weak currency thereby accelerating losses (assuming the interest or discount was
deductible on an accrual basis). In domestic currency terms, these gains and losses can be locked in
by foreign currency hedges. Thus, despite the foreign currency bond and hedge being equivalent to a
domestic bond, some gains can be deferred or losses accelerated where one of the transactions is
taxed on a realisation basis.

134. By being able to choose the form of a transaction, and avoiding hedge accounting, taxpayers
can generally take advantage of a realisation basis of taxation. In contrast, however, if the hedge and
the underlying transaction were required to be integrated by the tax system, systematic accruals
taxation could be applied to the single, synthetic transaction. This is possible by amalgamating the
cash flows of the linked positions. Integration techniques thus have a potential revenue protection
role in preventing deferral of the taxation of locked-in gains and acceleration of deductions for
locked-in losses.

135. Integration or hedge logic can also be used to defeat tax-motivated straddles. Straddles of this
type involve a taxpayer holding offsetting profit and loss positions in, for example, futures or swaps
that reduce the taxpayer's risk of loss. If the taxpayer is taxed on the transactions on a realisation
basis, it can select which transaction to realise, or when to make a payment, in order to obtain a tax
benefit. For example, it can create a taxable loss just prior to the end of the income year, in the
knowledge that the loss can be economically offset by realisation or payment/receipt at the beginning
of the next income year. By linking the offsetting positions, and treating them as one transaction, the
so-called 'timing option' is rendered ineffective for tax purposes.

2. Problems with Mandatory Integration

136. The use of integration as an anti-avoidance measure necessitates identification of the linked
offsetting positions. Administratively, this may be very difficult to do. The procedural issues
described above in the context of elective hedging rules are also present in designing a mandatory
integration system; moreover, the problems are intensified since the rules would apply without regard
to the taxpayer's co-operation in identifying offsetting positions. Generally, an integration regime
requires the taxpayer's purpose to be determined, imposing potentially difficult practical
 36

administrative issues for the tax system to enforce in a mandatory system.

3. Non-Elective Integration through Uniform Characterisation Rules

137. Some countries achieve non-elective integration by establishing uniform rules that treat certain
types of derivative instruments for tax purposes in a manner that is consistent with the normal
treatment of the investments that they are most likely to hedge. For example, it might be determined
that derivatives relating to interest and currency contracts would normally be used to hedge
transactions which themselves produce ordinary income or loss. In contrast, it might be assumed that
other derivatives, such as those related to the value of equity, are generally used to hedge transactions
that give rise to capital gain or loss.

138. If the tax rules assign a corresponding character to classes of derivative instruments on a
non-elective uniform basis, the rules could provide a general matching between the offsetting
positions in a hedging transaction without presenting the difficult issues described above for elective
systems. Thus rather than applying to identified discrete transactions, this approach offers certainty to
the hedging party through uniform rules for all instruments in a particular category. The value of this
approach, however, will depend on whether uniform rules can assign an appropriate character to
specified classes of derivative instruments. Countries disagree on the extent to which this is possible
and therefore the extent to which character mismatches can be avoided.

F. Are Bifurcation and Integration Compatible?

139. Bifurcation and integration appear to be contrasting processes; bifurcation divides transactions
into small components; integration combines transactions into a composite whole. Yet they derive
from the same logic; to establish what, in substance, is happening in terms of the
debt/equity/derivative trichotomy. Accordingly, they can logically operate together in appropriate
circumstances.

140. For example, in the case of a convertible security, bifurcation would break the security into a
debt component and an equity option component. The debt component could be taxed on a yield to
maturity basis. The treatment of the equity option could then depend on whether it satisfied
integration/hedging tests. If it did, it could be taxed in accordance with the underlying transaction.

G. Mark-to-Market Valuation Systems

141. Many of the classification problems described above lose significance when all transactions
are marked to market. Under a mark-to-market approach, each transaction is treated as having been
disposed of at market value at year end, and reacquired at that value at the beginning of the
subsequent year. The gain or loss in each income year is the increase or decrease in the market value,
adjusted for amounts paid or received during the year.
 37

142. A mark-to-market approach takes the decision about when to dispose of a transaction out of
the taxpayer's hands, preventing any tax motivated "timing option". It also obviates the need for
bifurcation and hedging as all gains and losses are treated in the same way, with gains being offset by
losses of the period.

143. A mark-to-market system as a taxing methodology has attractions for traders whose business
activity is dealing in financial transactions, turning them over frequently. This is particularly the case
where the trader has a "matched book", where the market value of the portfolio of assets and
liabilities broadly offset each other, and the extent of mismatch is the taking of a deliberate position to
make a profit on market movements. Often the financial intermediaries holding these positions will
mark such books to market for financial accounting purposes.

144. Difficulties and concerns in relation to a mark-to-market approach include:

! valuing illiquid transactions, where there is not a deep and liquid market for that
transaction

! the potential for deductions being allowed for doubtful debts

! the fact that the system taxes unrealised gains (and allows deductions for unrealised
losses).

145. The last point is a problem where there is no "matched book" and financial transactions are
not turned over frequently, leading to possible cash flow difficulties in having to pay tax on unrealised
gains. The problem would be exacerbated where the gains in fact do not eventuate (unless
ameliorated by, for example, carry back of losses). Allowing deductions for unrealised losses in the
reverse situation may also produce a problem for the revenue.

146. On the other hand, where the taxpayer actually uses a mark-to-market basis for purposes of
accounting, the significance of the above practical problems may be diminished. In such cases, the
assets and liabilities will already be valued for each accounting period, so that the tax rules do not
increase such burdens.

147. Moreover, those accounting or tax systems that adopt mark-to-market valuation often limit
them to particular sectors where they are most well-suited, such as for measuring taxable income for
financial intermediaries. Some countries believe that where the taxpayer actually trades in the
positions of the type at issue, the problems of valuing illiquid transactions and the cash-flow problems
of paying tax on unrealised gains may be largely theoretical.

148. Thus a mark-to-market approach offers a solution to the difficulties associated with taxing
debt, equity and derivatives differently, since income is measured in a consistent manner for each type
of instrument. As such it potentially offers an approach for reducing, if not eliminating, arbitrage
between the economic results and tax treatment. Moreover, the practical issues presented by
 38

valuation and application to illiquid assets may be less significant in cases where the taxpayer is a
financial intermediary that trades in the position or otherwise applies a mark-to-market system for
purposes of its accounting method.

H. Anti-Abuse Measures

149. The potential for tax arbitrage exists in a wide variety of the transactions described above. In
general, jurisdictions have attempted to respond to discovery of such arrangements on a piece-meal
basis. Sometimes the response has taken the form of an attack on the transactions under a
"substance-over-form" or "sham" approach. In other cases specific legislation has been introduced to
remove or minimise the perceived tax arbitrage opportunity. Such responses are, inevitably, reactions
to what has become common practice. Although they may put a stop to a particular abuse, this is
generally simply replaced by another which the legislation does not address or which the tax
authorities take more time to uncover. Even those authorities which have a
"substance-over-form"/"sham" approach available to them are dependent on finding out the facts of a
case before the challenge can be effective.

150. Such reactive and piece-meal responses by individual jurisdictions to the abuse of tax
asymmetries either within a system or between systems may ultimately be found to be the only
approach to the problem which is feasible and, in the case of off-market pricing, no other approach is
likely to be possible. Consequently, tax authorities face a never-ending cycle of discovery;
legislation/litigation; taxpayer reaction; and back to discovery. Before concluding that nothing else
can be done, however, it is at least worth considering what other options might be available to tax
authorities to address tax arbitrage issues.

151. Of course, tax arbitrage opportunities could only be comprehensively addressed by a broad
change in the tax system, such as applying mark-to-market valuation to tax all transfers of cash or
value at the same time and at the same rate without allowing any form of relief or off-set and
irrespective of the character of the payer, recipient or payment. At least in the foreseeable future, this
is unlikely to be an acceptable response in any single tax system; let alone in them all.

152. Similarly, although harmonisation of tax systems world-wide would considerably reduce the
scope for asymmetries which give rise to cross-border arbitrage, there seems little real prospect of this
being achievable.

153. Realistically, all that countries can do in considering their own individual tax systems is to
review the tax asymmetries to which that system gives rise and consider whether there is a strong
enough reason for retaining those features of the system in the face of such arbitrage opportunities.
Given the range of different tax systems, it would not be feasible to conduct a similar rigorous
examination of all cross-border asymmetries. What could be done, however, is for two countries
engaged in negotiating a tax treaty to consider the scope for tax arbitrage to which that treaty might
itself give rise when the asymmetries between their respective tax systems are borne in mind.

154. So long as asymmetries of tax treatment exist within and between tax systems there will be
 39

scope for tax arbitrage arrangements between unrelated (as well as related) persons. Governments
will therefore need to continue to police their system and react to examples of abuse when these are
recognised. They can also consider taking more positive steps to restrict tax arbitrage opportunities to
the extent that these involve or rely on obtaining treaty benefits.

VI. Summary

155. Innovative financial transactions have become an essential tool of modern business and
investment. Through their ability to separate, replicate, and recombine economic attributes into new
forms of transactions, financial instruments enable the shifting and redistribution of risks. Indeed, the
growing sophistication of these transactions has been driven by the demands from business and
financial markets for more effective risk management. These developments, however, raise serious
challenges to income tax systems, involving issues of the character and source of income, the timing
of income and deductions, treaty classification of payment streams, and identification of the owner of
financial instruments. Moreover, the greater flexibility afforded to taxpayers by such transactions has
expanded opportunities for tax arbitrage. Tax systems that have responded to these concerns have
adopted a variety of different approaches, including: (i) reliance on financial accounting rules; (ii)
disaggregation of transactions into discrete components; (iii) integration of offsetting positions; (iv)
taxation of changes in value through "mark-to-market" systems; and (v) targeted anti-abuse rules.
Each of these methods has merit in certain circumstances, but none provides a comprehensive
solution to the tax issues presented. Thus governments and tax authorities should consider all of the
above, either individually or in combination, in developing new rules in this area; suitable approaches
must balance the various tax policy objectives at stake, while addressing the risk management needs
of taxpayers. Ultimately, the innovation and expanded use of financial instruments test the viability
of the current frameworks of income tax systems and international tax agreements, and will continue
to demand effective and appropriate responses.
 40

APPENDIX I: EXAMPLES OF REPLICATION

Example 1 -- Put-Call Parity

The "put-call parity" theorem states a fundamental relationship between debt, stock, and options
to purchase or sell such stock. In simplified form, the theory can be illustrated as follows. Assume
that Investor 1 holds a zero coupon bond (that matures on Date Y) with a face amount of $X plus a
call option to purchase a share of Stock on Date Y at an exercise price of $X. Investor 2 actually
purchases a share of Stock at its current value, and also buys a put option which allows him to sell
the Stock on Date Y at an exercise price of $X. Put-call parity states that the two investors have
economically equivalent positions. In the form of an equation:

Investor 1's Position = Investor 2's Position

Debt ($X) +Call Option ($X) = Stock + Put Option ($X)

This simplified form ignores the offsetting effects of dividends paid on the stock, as well as
transaction costs. Nevertheless, the fundamental relationship between these instruments can be
illustrated by considering the alternatives for each investor as of Date Y. Investor 1 holds Debt that
will return $X on Date Y; Investor 1 will then have the choice of keeping $X in cash or using the
cash to exercise the Call, thus acquiring Stock. Clearly, Investor 1 will exercise the call only if Stock
is worth more than $X as of Date Y. Thus, the investment participates in any appreciation of Stock,
but is protected from a drop in its value below $X. Similarly, Investor 2 holds the Stock and the
right to "put" the stock to another party for at least $X; thus Investor 2 will also participate in any
appreciation of the stock above $X, but due to the put rights, the investment will not drop below $X.
Accordingly, the two positions are equivalent.

As illustrated in the following examples, these economic equivalencies can be used to

"replicate" synthetic forms of each of the other financial instruments.

Example 2 -- Synthetic Debt

The above relationships can be used to construct a synthetic form of a debt instrument, based on
the following equation:

Debt = Stock + Put Option - Call Option

To illustrate, assume an investor bought a non-dividend paying share for $80. The investor also
(i) buys a put option to sell the share for $100 at a specified date; and (ii) sells a call option to buy the
share for $100 at that date. The effect of combining these instruments is equivalent to the investor
purchasing a zero coupon bond with a redemption price of $100. If the value of the share drops
below $100 on the exercise date, the investor will use the purchased stock to exercise the put and
receive $100 (and the holder of the call option will let the call lapse). In contrast, if the value of the
share is above $100, the holder of the call will exercise the option and pay the investor $100 for the
 41

share (and the investor will let the put option lapse). In each case, investor is guaranteed a return of
exactly $100, and will not participate in the appreciation or depreciation of the purchased shares.

The same relationships can be employed to produce "synthetic" forms of the option contracts
and the stock itself.

Example 3 -- Synthetic Forward Contract

In section II(D), above, several formulas were provided for the relationship between forward
contracts and other instruments. These equations can be illustrated by the following examples:

(1) Forward Contract = Equity - Debt

Assume an investor borrows $90 and promises to repay the principal plus $10 interest one year
later. The investor then uses the $90 to purchase a share of stock. The leveraged purchase of stock
operates in an equivalent manner as entering into a forward contract to purchase the stock in one year
for $100. In both cases, the investor expends none of his own cash upfront, but is subject to a
requirement to pay $100 in one year; moreover, both positions carry the risks and returns inherent in
the share of stock.

(2) Forward Purchase Contract = Call Option - Put Option

A forward contract constitutes both the right and the obligation to purchase the subject of the
contract at the forward price on the specified date. These two contractual elements can be replicated
through separate option contracts.

For example, assume an investor (i) buys a European call option for a share in X Co that has an
exercise price of $100 and is exercisable on Date Y, and (ii) sells a European put option also with an
exercise price of $100 exercisable at the same date. The investor will have effectively entered into a
forward contract to purchase the share on Date Y for $100. If the value of the share is more than
$100, the investor will exercise the call and purchase the share for $100. Alternatively, if the value of
the share is less than $100, the investor will be obligated under the put to purchase the share for $100.

(3) Forward Sale Contract = Put Option - Call Option

The reverse relationship from the previous illustration is the case in a forward contract to sell
shares. In this case, a forward sales contract can be replicated if the investor (i) sells a European call
option for a share in X Co that has an exercise price of $100 and is exercisable on Date Y, and (ii)
buys a European put option also with an exercise price of $100 exercisable at the same date.
 42

APPENDIX II: EXAMPLES OF BIFURCATION

Example 1

X Co, a copper producer, issues a structured note for $10m, consideration for which is a promise
to pay $10m in three years, plus $10m multiplied by any increase in the price of copper over that
period.

The investor has two types of right: a right to the $10m, which is not contingent on any external
variable (other than the issuer's ability to pay), and a right to a further payment that is contingent on
the price of copper.

The structured note can therefore be seen as comprised of two instruments:

i. a zero coupon instrument with an issue price equal to the net present value of the promise
to pay $10m in three years; and

ii. a cash settled option on the price of copper.

In effect, the exercise price with respect to the option element is the present price of copper.

Under a bifurcation approach, each of the elements would be valued and subject to income tax
separately. Determining the issue price of the zero coupon instrument can be done in one of two
ways. First, the face value or maturity amount can be discounted at the rate applicable to similar
securities, i.e. those issued under the same terms by that issuer. If such a security did not exist, a
security issued under the same terms at that time by an issuer with the same credit rating could be
used.

Alternatively, the value of the option could be derived by reference to the quoted price of traded
options with the same terms. The value of the zero coupon instrument, determined residually, would
be the difference between the face value of the instrument and the value of the option. If the option
does not have a traded option equivalent, an options pricing model may provide a reasonable
approximation of the value. In this situation, however, it may be easier to calculate the issue price of
the zero coupon instrument first and the value of the option residually.

Assume an annual yield on three year zero coupon instruments issued at that time by issuers of
similar credit ratings was 7%, compounded quarterly. The issue price of the zero coupon instrument
would be $8,120,579. Calculated residually, the issue price of the option would be $1,879,421. Since
this amount represents the discount (or interest) on the borrowing, it may be accrued as taxable
income and allowable deduction for the holder and issuer respectively. The option may be taxed
under the rules applicable to options (which are generally taxed on a realisation basis) or, under tax
hedge rules, consistently with an underlying transaction.

Example 2
 43

Y Co. issues a five year convertible security for $10m. The security pays an annual coupon of
6% per annum, and is convertible - at the investor's option - into a specified number of shares of Y
Co. That is, in five years the investor has the choice of receiving $10m or converting the principal
into the given number of Y Co shares.

Assume that the yield on a similar non-convertible security is 7.5% per annum, compounded
annually. Under a bifurcation approach, the issue price of the note would be $9,393,117. Determined
residually, the value of the equity option (at issue date) would be $606,883.

The effect of bifurcation is to treat the convertible security as two instruments:

! a security with a face value of $10m, issued at a discount of $606,883 and payment
annual coupons at the rate of 6%
! an equity option, with a premium of $606,883 and a strike price of $10m.

Note that bifurcation exposes the discount $606,883 in the convertible security, The
following table compares the taxing of only the nominal interest coupons with taxing of the discount
and interest on a compounding accruals basis:

Period Not bifurcated Bifurcated

0
1 600,000 704,484
2 600,000 712,320
3 600,000 720,744
4 600,000 729,800
5 600,000 739,535
3,000,000 3,606,883

Thus the total amount of discount and interest under the debt component of the security
exceeds the interest under the non-bifurcated security. Note that the taxing of the discount gain is
deferred for a non-bifurcated security. Symmetrical treatment for the issuer and holder would
suggest that under bifurcation the issuer would be able to accrue the discount as a deduction over
the term of the security.

APPENDIX III: HEDGING EXAMPLES

Example 1 -- Fixed to Floating Rate Swap

Assume that X Co has a floating rate debt liability. If X Co wants to hedge its exposure to
rising interest rates, it could enter into an interest rate swap under which it agrees with a
counterparty (Y Co) to pay a fixed cash flow calculated by reference to the principal amount of its
debt in consideration of Y Co agreeing to pay amounts equivalent to X Co's floating interest rate
 44

obligations less a margin.

X Co has effectively changed its risk profile from floating interest rates to fixed rate interest
rates. X Co's interest rate under the floating rate borrowing is LIBOR + 0.75%. In entering into the
swap, X Co pays 6% fixed and receives LIBOR. X Co's net cost after the hedge is 6.75% fixed:
(LIBOR + 0.75%) - LIBOR + 6%.

Example 2 -- "Overhedging" an Underlying Position

Assume that X Co in Example 1 borrows $20m, but that the swap has a notional principal of
$50m. X Co may be hedged in respect of the floating rate interest payments on its borrowing but
(assuming that it has no other offsetting position) has an exposure to floating interest rates in
respect of the excess of the notional principal amount of the swap over the actual loan principal
(i.e., $30m). In effect, X Co is speculating that interest rates will rise. There would also be a
speculative element if the notional principal of the swap was the same as the actual loan principal,
but the swap term was, say, 5 years and the borrowing was for, say, 2 years.

This example illustrates the practical difficulties in making clear distinctions, except in
straightforward situations, between hedging and other purposes.

Example 3 -- "Natural Hedge" of Investment in Foreign Subsidiary

An Australian company (A Co) pays £100m to acquire the shares in a United Kingdom
company whose profits are denominated in sterling. A Co borrows £100m on a long term basis. If
sterling depreciates against the Australian dollar, the value of the UK subsidiary in Australian dollar
terms decreases. However, the loss due to exchange rate movements is offset by the gain made by
having to use less Australian dollars to buy the sterling required to repay the borrowing.

In this example, it could be said that there are two purposes in borrowing in sterling.
Probably the dominant reason is to raise funds. However, a second and important reason is to
hedge against depreciation of sterling. By contrast, if A Co has entered into an Australian
dollar/sterling currency swap with a principal of £100m, and the company has no other sterling
exposure, the dominant purpose could be said to be one of hedging.