Hull RMFI4 e CH 10

Volatility
Chapter 10
Risk Management and Financial Institutions 4e, Chapter 10, Copyright John C. Hull 2015
Definition of Volatility
Suppose that Si is the value of a variable on

day i. The volatility per day is the standard
deviation of ln(Si /Si-1)
Normally days when markets are closed are

ignored in volatility calculations (see Business
Snapshot 10.1, page 207)
The volatility per year is 252 times the daily
volatility
Variance rate is the square of volatility
Implied Volatilities
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
VIX Index: A Measure of the Implied

Volatility of the S&P 500 (Figure 10.1, page
208)
Are Daily Changes in Exchange Rates

Normally Distributed? Table 10.1, page 209
Real World (%)
>1 SD
>2SD
>3SD
>4SD
>5SD
>6SD
25.04
5.27
1.34
0.29
0.08
0.03
Normal Model (%)
31.73
4.55
0.27
0.01
0.00
0.00
Heavy Tails
Daily exchange rate changes are not normally

distributed
The distribution has heavier tails than the normal

distribution
It is more peaked than the normal distribution
This means that small changes and large changes

are more likely than the normal distribution would
suggest
Many market variables have this property, known
as excess kurtosis
Normal and Heavy-Tailed

Distribution
Alternatives to Normal Distributions:

The Power Law (See page 211)
Prob(v > x) = Kx-
This seems to fit the behavior of the
returns on many market variables better
than the normal distribution
Log-Log Test for Exchange Rate Data (v

is number of standard deviations which
the exchange rate moves)
m
1
2
2
(
u
u
)
n
n
i
m
1
m
ui1n
Standard Approach to Estimating

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)
10
1
m
u
m
22
ni1ni
Simplifications Usually Made in

Risk Management
Define
ui as (SiSi-1)/Si-1
Assume
that the mean value of ui is zero
Replace
m-1 by m
This gives
11
w
h
r
e
1
m
22
n
i
n
i
i
1
m
ii
1
Weighting Scheme
Instead of assigning equal weights to the

observations we can set
12
w
r
h
e
1
m
2
2
n
L
i
n
i
i
1
m
ii
1
ARCH(m) Model
In an ARCH(m) model we also assign

some weight to the long-run variance rate,
VL:
13
(1)u
2n2n1n21
EWMA Model
In
(page 216)
an exponentially weighted moving

average model, the weights assigned to
the u2 decline exponentially as we move
back through time
This leads to
14
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
= 0.94 has been found to be a good
choice across a wide range of market
variables
15
V
u
2nLn212n1
GARCH (1,1), page 218
In GARCH (1,1) we assign some weight to

the long-run average variance rate
Since weights must sum to 1
16
2nV
2
2
n
1
n
L
1
GARCH (1,1) continued
Setting VL the GARCH (1,1) model is
and
17

0...2013u086
2nnn
2121
Example
Suppose
The
long-run variance rate is 0.0002 so

that the long-run volatility per day is 1.4%
18
0..2013.086.2560.236
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
19
p
q
22
2
nii
i1j
n
j1n
j
GARCH (p,q)
20
Other Models
Many
other GARCH models have been

proposed
For example, we can design a GARCH
models so that the weight given to ui2
depends on whether ui is positive or
negative
21
Maximum Likelihood Methods

In
maximum likelihood methods we

choose parameters that maximize the
likelihood of the observations occurring
22
p(1)
Example 1 (page 220)

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
23
Example 2 (page 220-221)

Estimate the variance of observations
from a normal distribution with mean zero
ui2
1

Maximize :
exp
2v
i 1
2v
n
ui2
Same as maximizing : ln( v)
v
i 1
1 n 2
Maximum value when v ui
n i 1
n
24
1
u
r
e
x
p
2
v
2
v
ln
(v)
u
2
m
i
iim
1
i
2
i
i
Application to EWMA and

GARCH
We choose parameters that maximize
Options, Futures, and Other Derivatives, 9th Edition, Copyright John C. Hull 2014
25

ln(v)
u
2
m
i
i
1i
S&P 500 Excel Application
Start with trial values of parameters ( for EWMA

and , , and for GARCH(1,1)
Update variances
Calculate
Use solver to search for values of parameters that

maximize this objective function
For efficient operation of Solver: set up spreadsheet
so that ensure that search is over parameters that
are of same order of magnitude and test alternative
starting conditions
26
S&P 500 Excel Application (Table 10.4)
vi =i2
ln(vi ) ui2 /vi
0.004759
0.00004531
9.5022
0.006606
0.00004447
9.0393
0.00016327
8.6209
Date
Day
Si
18-Jul-2005
1221.13
19-Jul-2005
1229.35
0.006731
20-Jul-2005
1235.20
21-Jul-2005
1227.04
..
13-Aug-2010
1279
1079.25
Total
ui=(SiSi-1)/Si-1
..
0.004024
10,228.2349
27
The S&P 500 (Figure 10.4)
28
The GARCH Estimate of

Volatility of the S&P 500 (Figure 10.5, page
530)
29
Variance Targeting
One
way of implementing GARCH(1,1)

that increases stability is by using variance
targeting
The long-run average variance equal to
the sample variance
Only two other parameters then have to
be estimated
30
How Good is the Model?

The
Ljung-Box statistic tests for

autocorrelation
We compare the autocorrelation of the
ui2 with the autocorrelation of the ui2/i2
31
E
[]V
()(V)
2nkm
k
2
L
n
L
m
1
1
2
E
k
k
0
Forecasting Future Volatility

(equation 10.14, page 532)
A few lines of algebra shows that
The variance rate for an option expiring on

day m is
32
Forecasting Future Volatility

continued (equation 10.15, page 534)
Define
a ln
The estimated volatility per annum for an option lasting T days is
1 e aT
252 VL
V
(0)
L
aT
33
S&P Example
1
a ln
0.006511
0.083394 0.910116
Option Life
(days)
10
30
50
100
500
Est. Volatility
(% per annum)
27.36
27.10
26.87
26.35
24.32
34
Volatility Term Structures
GARCH (1,1) suggests that, when calculating vega,

we should shift the long maturity volatilities less than
the short maturity volatilities
When instantaneous volatility changes by (0),
volatility for T-day option changes by
1 e aT (0)
(0)
aT (T )
35
Results for S&P 500 (Table 10.7)

When
instantaneous volatility changes by
1%
Option Life
(days)
10
30
50
100
500
Volatility
increase (%)
0.97
0.92
0.87
0.77
0.33
36

Hull RMFI4 e CH 10

Uploaded by

Copyright:

Available Formats

Hull RMFI4 e CH 10

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hull RMFI4 e CH 10

Uploaded by

Copyright:

Available Formats

Volatility

Suppose that Si is the value of a variable on

Normally days when markets are closed are

the variables needed to price an option

VIX Index: A Measure of the Implied

Are Daily Changes in Exchange Rates

Normal Model (%)

Daily exchange rate changes are not normally

The distribution has heavier tails than the normal

This means that small changes and large changes

Normal and Heavy-Tailed

Alternatives to Normal Distributions:

Log-Log Test for Exchange Rate Data (v

Standard Approach to Estimating

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

Simplifications Usually Made in

that the mean value of ui is zero

Instead of assigning equal weights to the

In an ARCH(m) model we also assign

an exponentially weighted moving

little data needs to be stored

GARCH (1,1), page 218

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

Since weights must sum to 1

GARCH (1,1) continued

Setting VL the GARCH (1,1) model is

long-run variance rate is 0.0002 so

that the current estimate of the

other GARCH models have been

Maximum Likelihood Methods

maximum likelihood methods we

Example 1 (page 220)

observe that a certain event happens

maximize this to obtain a maximum

Example 2 (page 220-221)

Application to EWMA and

We choose parameters that maximize

S&P 500 Excel Application

Start with trial values of parameters ( for EWMA

Use solver to search for values of parameters that

S&P 500 Excel Application (Table 10.4)

ln(vi ) ui2 /vi

The S&P 500 (Figure 10.4)

The GARCH Estimate of

way of implementing GARCH(1,1)

How Good is the Model?

Ljung-Box statistic tests for

Forecasting Future Volatility

A few lines of algebra shows that

The variance rate for an option expiring on

Forecasting Future Volatility

The estimated volatility per annum for an option lasting T days is

Volatility Term Structures

GARCH (1,1) suggests that, when calculating vega,

Results for S&P 500 (Table 10.7)

instantaneous volatility changes by

You might also like