Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.1

Volatility

Chapter 9

Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.2

Definition of Volatility

The volatility σ of a variable is the standard

deviation of its return per unit of time (one year,

one day) with the return being expressed with

continuous compounding

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.3

0

0

0

0

0

)ln(

)ln(:

:)ln(

S

SS

S

S i.e.

change percentage the to close is return

compoundedly continuous the small,is Twhen

S

S of devation standardtheT

TtimeinearnedreturntotalS

S

TT

T

T

Definition of Volatility

Example:

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.4

$2.080.041650 : weekonein

pricestock in the movedeviation standardoneA

%16.452/130 :elyapproximat

weekonein pricestock in change percentage of std.

yearper %30

50$0

S

Some points concerning volatility

The variance rate is the square of volatility

Standard deviation of returns increase with

the square root of time, the variance of this

return increases linearly with time

Normally days when markets are closed

are ignored in volatility calculations (252

days per year)

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.5

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.6

Implied Volatilities

Implied volatilities are the volatilities

implied from option prices

Of the variables needed to price an option

the one that cannot be observed directly is

volatility

We can therefore imply volatilities from

market prices and vice versa

Options, Futures, and Other Derivatives, 7th Edition, Copyright © John C. Hull 2008 7

Estimating Volatility from

Historical Data

1. Take observations S0, S1, . . . , Sn at intervals of t years

2. Calculate the continuously compounded return in each interval as:

3. Calculate the standard deviation, s , of the ui´s

4. The historical volatility estimate is:

uS

Si

i

i

ln1

t

sˆ

Are Daily Percentage Changes in

Financial Variables Normal?

The Black-Scholes-Merton Model

assumes that asset prices change

continuously and have constant volatility.

This means that the return in a short

period of time Δt always has a normal

distribution with a standard deviation of

.

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.8

t

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.9

Daily movements of 12 different

exchange rates over ten years

Real World (%) Normal Model (%)

>1 SD 25.04 31.73

>2SD 5.27 4.55

>3SD 1.34 0.27

>4SD 0.29 0.01

>5SD 0.08 0.00

>6SD 0.03 0.00

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.10

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.11

An Alternative to Normal

Distributions: The Power Law

The value v of the variable satisfies

Prob(v > x) = Kx-a

For x large, where K and α are constants

This seems to fit the behavior of the returns

on many market variables better than the

normal distribution

The Power Law

Example:

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.12

... 0.0019,30)Pr(v

0.0062520)Pr(v

estimate can weand ,50:Then

05.0)10Pr(:nObservatio

3 law withpower a Suppose

K

v

a

The Power Law

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.13

holds lawpower ether thecan test w we

,lnagainst x)](vPr[ln plottingby So

lnlnx)](vPr[ln

have We

x

xK-α

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.14

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.15

Monitoring Daily Volatility

Define n as the volatility per day between

day n-1 and day n, as estimated at end of day

n-1

Define Si as the value of market variable at

end of day i

Define ui= ln(Si/Si-1)

n n ii

m

n ii

m

mu u

um

u

2 2

1

1

1

1

1

( )

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.16

Simplifications Usually Made in

Risk Management

Define ui as (Si-Si-1)/Si-1

Assume that the mean value of ui is zero

Replace m-1 by m

This gives

n n ii

m

mu2 2

1

1

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.17

Weighting Scheme

Instead of assigning equal weights to the

observations we can set

a

a

n i n ii

m

ii

m

u2 2

1

1

1

where

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.18

ARCH(m) Model

In an ARCH(m) model we also assign

some weight to the long-run variance rate,

VL:

a

a

m

i

i

m

i iniLn uV

1

1

22

1

where

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.19

EWMA Model

In an exponentially weighted moving

average (EWMA) model, the weights

assigned to the u2 decline exponentially as

we move back through time, i.e.

where λ is between 0 and 1

This leads to2

1

2

1

2 )1( nnn u

ii aa 1

EWMA Model

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.20

12

1

2

22

1

12

2

2

22

2

2

1

2

2

1

2

1

2

2

1

)1(for

as same theisequation thism, largefor

)1(

thatsee we way,in this continuing

))(1(

get to

)1(

in for Substitute

i

iin

m

i

in

mn

m

in

m

i

i

n

nnnn

nnn

n

u

u

uu

u

aa

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.21

Attractions of EWMA

Relatively little data needs to be stored

We need only remember the current estimate of the variance rate and the most recent observation on the market variable

Tracks volatility changes

RiskMetrics uses = 0.94 for daily volatility forecasting

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.22

The GARCH (1,1) Model

In GARCH (1,1) we assign some weight to

the long-run average variance rate

Since weights must sum to 1

a b 1

2

1

2

1

2

ba nnLn uV

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.23

GARCH (1,1) continued

Setting w V the GARCH (1,1) model

reads as

and

w

ba

w

1LV

2

1

2

1

2

baw nnn u

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.24

Example

Suppose

The long-run variance rate is 0.0002 so

that the long-run volatility per day is 1.4%

0002.0

01.086.013.011

86.013.0000002.0 2

1

2

1

2

LL

nnn

VV

u

w

ba

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.25

Example continued

Suppose that the current estimate of the

volatility is 1.6% per day and the most

recent percentage change in the market

variable is 1%.

The new variance rate is

The new volatility is 1.53% per day

0 000002 013 0 0001 086 0 000256 0 00023336. . . . . .

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.28

Maximum Likelihood Methods

In maximum likelihood methods we

choose parameters that maximize the

likelihood of the observations occurring

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.29

Example 1

We observe that a certain event happens one time in ten trials. What is our estimate of the proportion of the time, p, that it happens?

The probability of the outcome is

We maximize this to obtain a maximum likelihood estimate: p=0.1

9)1( pp

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.30

Example 2

Estimate the variance of observations from a

normal distribution with mean zero

Maximize:

or:

This gives:

1

2 2

1

2

1

2

1

2

1

v

u

v

vu

v

vn

u

i

i

n

i

i

n

i

i

n

exp

ln( )

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.31

Application to GARCH (1,1)

We choose parameters that maximize

ln( )vu

vi

i

ii

n 2

1

Application to GARCH (1,1),

Example

The next table analyzes data on the

Japanese yen exchange rate between

January 6, 1988, and August 15, 1997.

The numbers are based on trial estimates

of the three parameters: ω, α, β.

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.32

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.33

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.34

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.35

Variance Targeting

One way of implementing GARCH(1,1)

that increases stability is by using variance

targeting

We set the long-run average volatility

equal to the sample variance

Only two other parameters then have to be

estimated

How Good is the Model?

The assumption underlying a GARCH

model is that volatility changes with the

passage of time, during some periods

high, during others relatively low. In other

words, when is high, there is a tendency

for to be high and vice versa.

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.36

2

iu

,..., 2

2

2

1 ii uu

How Good is the Model?

So assuming that the exhibit

autocorrelation. If a GARCH model is

working well, the autocorrelation for the

variables should be very small.

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.37

2

iu

22 / iiu

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.38

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.39

How Good is the Model?

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.40

Forecasting Future Volatility

A few lines of algebra shows that

fleeingmean :1

revertingmean :1

)()(][ 22

ba

ba

ba Ln

k

Lkn VVE

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.41

Volatility Term Structures

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.42

]V)[V(aT

eVV(t)dt

T

]V)[V(eVV(t)

βα

], aE[σV(t)

L

T aT

L

L

at

L

tn

011

T andday between to

day per rate varianceaverage The

0

then1

ln

Define

0

2

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.43

Forecasting Future Volatility

The volatility per year for an option lasting

T days is

L

aT

L VVaT

eVT )0(

1252)(

Risk Management and Financial Institutions, Chapter 5, Copyright © John C. Hull 2006 5.44

Volatility Term Structures

The GARCH (1,1) model allows us to

predict volatility term structures changes

When (0) changes by (0), GARCH

(1,1) predicts that (T) changes by

)}252

)0((

1{252(T)

because ),0()(

)0(1

22

L

aT

L

aT

VaT

eV

TaT

e