You are on page 1of 31

Chapter 23

Volatility

Introduction
Implied volatility
Volatility estimation
Volatility and variance swaps
Option pricing under stochastic volatility

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-2

Implied Volatility
The volatility is unobservable
One can use historical stock return data to
calculate stock return volatilitysample
standard deviation
One can also use the observed option price
and the Black-Scholes model to back out the
volatilityimplied volatility (IV)
IV is the volatility implied by the option price
observed in the market
Copyright 2006 Pearson Addison-Wesley. All rights
reserved.

23-3

Implied Volatility (contd)


Volatility skew
Volatility smirk
Volatility smile
Implied volatilities are not constant
across strike prices and over time

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-4

Volatility Index VIX


Introduced by Professor Bob Whaley at
Duke University in 1993
It provides investors with market
estimates of expected volatility
It is computed by using near-term S&P
100 index options

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-5

Measurement and Behavior of Volatility


Stock price process
dS t / S t ( )dt ( S t , X t , t )dZ

: continuously compounded expected return


: continuously compounded divided yield
(st, xt, t): instantaneous volatility

If we observe a series of stock price every h


periods, we can compute continuously
compounded return

t h ln(St h / St )

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-6

Historical Volatility
We observe n continuously compounded
stock return over a period of length T
and h=T/n

1 1
(n 1)
h
n

i 1

2
i

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-7

Historical Volatility (contd)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-8

Time Varying Volatility: ARCH Model


The ARCH(m) model
Autoregressive conditional
heteroskedasticity model
ln( St / St h ) ( 0.5 2 )h t
qt E ( t / t 1 )
2

qt a0 ai t2i
i 1

0 0,

0,

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-9

Time Varying Volatility:


ARCH Model (contd)
Intuition

ARCH model suggests that the level


of variance depends on recent past
level of variance

Empirical regularity

Volatility is highly persistent


High volatility tends to be followed by
high volatility

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-10

The Garch(m,n) Model


The GARCH model

Generalized ARCH model


m

qt a0 a
i 1

where a0 0,

ai 0,

2
i t i

b j qt j
j 1

i 1,L , m,

b j 0,

j 1,L , n

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-11

The Garch(m,n) Model (contd)


Intuition

Volatility at a point in time depends on


recent volatility and recent squared returns

Special case

GARCH (1,1) model

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-12

Estimation of ARCH, GARCH Model


Maximum likelihood estimation
Volatility forecasts

Conditional expectation of volatility

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-13

Variance Swaps
A forward contract that pays the difference
between a forward price F0,(v2) and some
measure of the realized stock price variance,
v2, over a period of time multiplied by a
notional amount

payoff =

F0,T (v ) N

N: the notional amount of the contract

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-14

Variance Swaps (contd)


Measurement issues

How frequently the return is measured

Whether returns are continuously compounded


or arithmetic

Whether the variance is measured by subtracting


the mean or by simply squaring the returns

The period of time over which variance is measured

How to handle days on which trading does


not occur

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-15

Extension of the Black-Scholes Model


Three extensions

The Merton jump diffusion model

The constant-elasticity of variance model

The stochastic volatility model

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-16

Mertons Jump Diffusion Model


The impact of jump on option prices
Example

$40,

$40,

30%,

T-t=0.25 year

=0.5% probability per year

Call price=$2.81, put price=$2.02

8%

Without jump

Call price=$2.78, put price=$1.99

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-17

Mertons Jump Diffusion Model (contd)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-18

Mertons Jump Diffusion Model (contd)


Implied volatility computed using the
Black-Scholes model when option prices
are computed using the jump model

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-19

Mertons Jump Diffusion Model (contd)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-20

Mertons Jump Diffusion Model (contd)


If prices of options properly account for
the jump
Yet we use the Black-Scholes model to
back out option implied volatility
Then out-of-the money puts have higher
implied volatility than at-the-money ones
In-the-money calls have higher implied
volatility than at-the-money ones
Copyright 2006 Pearson Addison-Wesley. All rights
reserved.

23-21

Constant Elasticity of Variance Model


Cox (1975) proposed the constant elasticity of variance
(CEV) model
Volatility varies with the level of the stock price

dS ( ) Sd t S

/2

dZ

The instantaneous standard deviation of the stock


return
( 2) / 2

( )

If <2 volatility decreases with the stock price


If >2 volatility increases with the stock price
If =2 the CEV model reduces to the lognormal process

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-22

The CEV Call Price


For the case <2, the CEV call price is

Se

2
2
rT
1 Q(2 y,2 2 ,2 x) Ke Q (2 x, 2 ,2 y )

2( )
2
(2 )( ( )( 2 ) 1)

x ks

( r )( 2 )T

y kK 2
Q(a,b,c) denotes the noncentral Chi-squared distribution function will
b degrees of freedom and noncentrality parameter c, evaluated at a
Copyright 2006 Pearson Addison-Wesley. All rights
reserved.

23-23

The CEV Call Price (contd)


For >2 the CEV call price is

Se

2
2
rT
1 Q(2 x, 2 ,2 y ) Ke Q(2 y,2 2 ,2 x)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-24

Implied Volatility in the CEV Model


When <2, the CEV model generates a
Black-Scholes implied volatility skew
Implied volatility decreases with the
option strike price

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-25

The Heston Stochastic Volatility Model


The model allows volatility to vary
stochastically but still to be correlated with the
stock price

dS ( ) Sdt v(t ) S dZ1


dv(t ) k (v v(t ))dt v v(t ) dZ 2
when

v(t )

E dZ dZ dt

is the volatility

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-26

The Heston Stochastic


Volatility Model (contd)
Assuming v(t), the instantaneous stock return variance
follows an It process, the multivariate Black-Scholes
partial differential equation is
1
1 2
2
v(t ) S VSS + v(t )Vvv + v(t ) v SVSv
2
2
+ (r ) SVS + [v v(t )] v(t ) v Vv + Vt = rV

This equation has an integral solution that can be


solved numerically

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-27

The Heston Stochastic


Volatility Model (contd)
Hestons stochastic volatility model offers a
closed-form solution for option prices
Empirical test of Hestons model
Bakshi, Cao and Chen (1997, Journal
of Finance)

They find that the feature of stochastic volatility can


lead to 80% reduction in Black-Scholes model pricing
error
The stochastic volatility is of first order importance in
comparison to the jump feature

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-28

The Heston Stochastic


Volatility Model (contd)
Implied volatility computed using the
Black-Scholes model when option prices
are computed using the Heston model

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-29

The Heston Stochastic


Volatility Model (contd)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-30

The Heston Stochastic


Volatility Model (contd)

Copyright 2006 Pearson Addison-Wesley. All rights


reserved.

23-31

You might also like