Models of Volatility

Smiles I
Chapter 8: ADVANCED
OPTION PRICING MODEL
2
Put-Call Parity Arguments
Put-call parity p +S
0
e
-qT
= c +K e
r T
holds regardless of the assumptions
made about the stock price distribution
It follows that
p
mkt
-p
bs
=c
mkt
-c
bs
3
Implied Volatilities
When p
bs
=p
mkt
, it must be true that c
bs
=c
mkt
It follows that the implied volatility
calculated from a European call option
should be the same as that calculated
from a European put option when both
have the same strike price and maturity
The same is approximately true of
American options
4
Volatility Smile
A volatility smile shows the variation of
the implied volatility with the strike price
The volatility smile should be the same
whether calculated from call options or
put options
5
The Volatility Smile for Foreign
Currency Options
Implied
Volatility
Strike
Price
6
Implied Distribution for Foreign
Currency Options
Both tails are heavier than the lognormal
distribution
It is also more peaked than the
lognormal distribution
7
The Volatility Smile for Equity
Options
Implied
Volatility
Strike
Price
8
Implied Distribution for Equity
Options
The left tail is heavier and the right tail is
less heavy than the lognormal
distribution
9
Other Volatility Smiles?
What is the volatility smile if
True distribution has a less heavy left tail
and heavier right tail
True distribution has both a less heavy left
tail and a less heavy right tail
10
Possible Causes of Volatility
Smile
Asset price exhibiting jumps rather than
continuous change
Volatility for asset price being stochastic
(One reason for a stochastic volatility in
the case of equities is the relationship
between volatility and leverage)
11
Volatility Term Structure
In addition to calculating a volatility smile,
traders also calculate a volatility term
structure
This shows the variation of implied
volatility with the time to maturity of the
option
12
Volatility Term Structure
The volatility term structure tends to be
downward sloping when volatility is high
and upward sloping when it is low
13
Example of a Volatility Surface
Strike Price
0.90 0.95 1.00 1.05 1.10
1 mnth 14.2 13.0 12.0 13.1 14.5
3 mnth 14.0 13.0 12.0 13.1 14.2
6 mnth 14.1 13.3 12.5 13.4 14.3
1 year 14.7 14.0 13.5 14.0 14.8
2 year 15.0 14.4 14.0 14.5 15.1
5 year 14.8 14.6 14.4 14.7 15.0
14
Three Alternatives to
Geometric Brownian Motion
Constant elasticity of variance
(CEV) by Cox and Ross 1976
Mixed Jump diffusion by Merton
1976
Variance Gamma by Madan, Carr
and Chang 1998
15
CEV Model
When = 1 the model is Black-Scholes
case
When > 1 volatility rises as stock price
rises
When < 1 volatility falls as stock price
rises
European option can be valued
analytically in terms of the cumulative
non-central chi-square distribution
Implied volatility o
imp
= oS
-1.
dz S Sdt q r dS

o + = ) (
16
CEV Models Implied Volatilities
o
imp
K/S
< 1
> 1
17
Distributional Implications
When < 1 probability distribution similar
to that observed for equities with heavy left
tail and less heavy right tail.
When > 1 probability distribution similar
to that sometimes observed for options on
futures with heavy right tail and a less
heavy left tail.
18
CEV European Call and Put
When 0 < < 1,
When > 1,
) , 2 , ( )] , , ( 1 [
) , , ( )] , 2 , ( 1 [
2
0
2
2 2
0
c b a e S a b c Ke p
a b c Ke c b a e S c
qT rT
rT qT
+ =
+ =




) , , ( )] , 2 , ( 1 [
) , 2 , ( )] , , ( 1 [
2
0
2
2 2
0
a b c e S c b a Ke p
c b a Ke a b c e S c
qT rT
rT qT
=
=




19
CEV European Call and Put
(Contd)
Where
and
2
(z,k,v) = noncentral chi-square distribution
v = noncentrality parameter
k = degrees of freedom less than z.
CEV model useful for valuing exotic equity models.
Parameters of the model chosen to fit the prices of plain
vanilla options by minimizing the sum of the squared
differences between model prices and market prices.
T
Se
c b
T
K
a
T q r
2 2
) 1 ( 2 ) (
2 2
) 1 ( 2
) 1 (
) (
,
1
1
,
) 1 ( o o

=

20
Mixed Jump Diffusion Model
Merton produced a pricing formula when the stock
price follows a diffusion process overlaid with random
jumps
dp is the Poisson random jump
k is the expected size of the jump
P dt is the probability that a jump occurs in the next
interval of length dt
dz is a Wiener process
dp and dz are independent
q dividend yields
dp dz dt k q S dS + + = o P Q ) ( /
21
Mertons European Options
An important case is where the logarithm of the size of
the percentage jump is normal. Merton shows the
European price
P=P(1+k)
f
n
= Black-Scholes option price with dividend rate q.
In f
n
, substitute o
2
+(ns
2
)/T for Variance rate and
r-Pk+nK/T for risk-free rate where s = standard deviation
of the normal distribution of the logarithm of the size of
percentage jump and K = ln(1+k)
n
n
n T
f
n
T e
V

g
=

=
0
'
!
) ' (P
P
22
Jumps and the Smile
Gives rise to heavier left and right tails
than Black-Scholes can be used for
pricing currency options
Jumps have a big effect on the implied
volatility of short term options
They have a much smaller effect on the
implied volatility of long term options
23
The Variance-Gamma Model
A popular pure jump model
g is change over time T in a variable that follows a
gamma process. This is a process where small jumps
occur frequently and there are occasional large
jumps
The probability density for g is of gamma
distribution
+(.) = gamma function
) / (
) , / ; (
/
/ 1 /
v T v
e g
v v T g
v T
v g v T
+
=

o
24
Understanding the Variance-
Gamma Model
g defines the rate at which information
arrives during time T (g is sometimes
referred to as measuring economic time)
If g is large the the change in ln S has a
relatively large mean and variance
If g is small relatively little information
arrives and the change in ln S has a
relatively small mean and variance
25
The Variance-Gamma Model
(Contd)
Conditional on g, ln S
T
is normal.
The conditional mean = lnS
0
+(r-q)T+[T+U g
The conditional standard deviation =
Its variance proportional to g
[ = (1/v) ln(1-Uv-o
2
v/2)
There are 3 parameters
v, the variance rate of the gamma process
o
2
, the average variance rate of ln S per unit time
U, a parameter defining skewness
g o
26
Skewness
When U=0, lnS
T
is symmetric
When U<0, lnS
T
is negatively skewed (as
for equities)
When U>0, lnS
T
is positively skewed
27
Volatility Smile
The Variance-gamma model tends to
produce U-shaped volatility smile.
Not necessarily symmetrical
Very pronounced for short maturities
Dies away for long maturities
Can be fitted to either equity or foreign
currency plain vanilla option prices.
28
VG Option Pricing
Madan et al (1998) offers a semi-analytic
European option valuation formula using
the modified bessel function of the second
kind and the degenerate hypergeometric
function of two variables.
We can also calculate by Monte Carlo
simulation.
29
VG Option Pricing (Contd)
The value g is generated by inverse gamma distribution
function, i.e., g=Gamma
-1
(rn
1
;T/v, v) with random number
rn
1
from uniform distribution of zero mean and unit variance
The stock price S
T
is given by
The value x is generated by inverse standard normal
distribution function, i.e., x=N
-1
(rn
2
)
The European call price is
Therefore,
] ) [(
0
x g g T T q r
T
e S S
o U [ + + +
=
)] 0 , [max( ) , ; (
0
K S E e T K S c
T
rT
=

n
i S K
e p
n
K i S
e c
n
i
T
rT
n
i
T
rT

=

=
1 1
] 0 ), ( max[
;
] 0 , ) ( max[