CH 20 Hull Fundamentals 7 The D

Fundamentals of Futures and Options Markets, 7th Ed, Ch 20, Copyright John C.
Hull 2010
Value at Risk
Chapter 20
1
Objective of VaR
VaR provides a single number that
summarizes the total risk in a portfolio of
financial assets
It is easy to understand
It asks the simple question: How bad can
things get?
More specifically:
What loss level is such that we are X%
confident it will not be exceeded in N business
days?
2
Fundamentals of Futures and Options Markets, 7th Ed, Ch 20, Copyright John C. Hull 2010
VaR and Regulatory Capital
Example of VaR in practice:
Regulators base the capital they require
banks to keep on VaR
The market-risk capital is k times the 10-day
99% VaR where k is at least 3.0

3
VaR
VaR is the loss corresponding to the
(100 x) percentile of the distribution of
the change in the value of a portfolio over
N days
If x = 99, VaR is the first percentile of the
distribution
If x = 97, VaR is the third percentile of the
distribution
4
N-Day VaR
Although the theoretical VaR is based on N
number of days, practitioners usually set N=1
and then find longer periods as follows:
N-day VaR = 1-day VaR x
N
5
Methods for Calculating VaR
Historical Simulation Approach
Model Building Approach

6
Historical Simulation
Use past data as a guide as to what might
happen in the future.
1. Identify the market variables (assets) that affect
the value of the portfolio
2. Create a database of the daily movements in all
market variables over n days
3. Define each days change in the variables as
scenarios:
1. Scenario 1 = Day 1 percent change in each variable
2. Scenario 2 = Day 2 percent change in each variable
3. Etc.
7
4. The first scenario trial assumes that the
percentage changes in all market variables are
as on the first day. The second scenario trial
assumes that the percentage changes in all
market variables are as on the second day,
and so on. The portfolios value tomorrow is
calculated for each scenario trial
5. For each scenario trial, calculate the dollar
change in portfolio value
6. Rank the outcomes from highest loss to lowest
loss.

8
7. The X percentile of the distribution is the worst
(n*X) observations
8. The X estimate of VaR at the (1-X)% confidence
level is the loss that occurs at the (n*X)
th

observation
9. Update the VaR estimate using a rolling window
of n observations. i.e. Drop Day 1 and add Day
n+1. Drop Day 2 and add Day n+2. Etc.
If the last n days is a good representation of what
might happen in the future, then the manager is (1
X)% certain that a loss greater than VaR will not
occur.
9
Historical Simulation continued
Suppose we use 501 days of historical data
Let v
i
be the value of a market variable on day i
There are 500 simulation trials
The ith trial assumes that the value of the
market variable tomorrow is
1
500
i
i
v
v
v
10
11
Sample Data
Day Market Market Market
Variable 1 Variable 2 .. . Variable n
0 20.33 0.1132 .. . 65.37
1 20.78 0.1159 .. . 64.91
2 21.44 0.1162 .. . 65.02
3 20.97 0.1184 .. . 64.90
.. .. .. .. . ..
.. .. .. .. . ..
498 25.72 0.1312 .. . 62.22
499 25.75 0.1323 .. . 61.99
500 25.85 0.1343 .. . 62.1
12
Results for Sample Data
Scenario Market Market Market Portfolio
Number Variable 1 Variable 2 .. .. Variable n Value ($mill)
1 26.42 0.1375 .. .. 61.66 23.71
2 26.67 0.1346 .. .. 62.21 23.12
3 25.28 0.1368 .. .. 61.99 22.94
.. .. .. .. .. .. ..
.. .. .. .. .. .. ..
499 25.88 0.1354 .. .. 61.87 23.63
500 25.95 0.1363 .. .. 62.21 22.87
The Model-Building Approach
The main alternative to historical simulation is to
make assumptions about the probability
distributions of return on the market variables
and calculate the probability distribution of the
change in the value of the portfolio analytically
This is known as the model building approach or
the variance-covariance approach
13
Daily Volatilities
Since we are working with variances, we
need to define the appropriate volatility
estimate.
In option pricing we express volatility as
volatility per year
In VaR calculations we express volatility
as volatility per day

o
o
day
year
=
252
14
Daily Volatility continued
Strictly speaking we should define o
day
as
the standard deviation of the continuously
compounded return in one day
In practice we assume that it is the
standard deviation of the percentage
change in one day
15
Microsoft Example
One Asset Case
We have a position worth $10 million in
Microsoft shares
The volatility of Microsoft is 2% per day
(about 32% per year)
The standard deviation of the change in
the portfolio in 1 day is $10 Mill x 2% =
$200,000
We use N = 10 and X = 99
16
Microsoft Example
Over a one-day period, the expected
change in the value of Microsoft is zero.
(This is OK for short time periods)
We assume that the change in the value of
the portfolio is normally distributed
Review normal distributions using the
Table on p. 590-591.
17
Microsoft Example
We want to compute a 10-day 99% VaR.
Use N = 10 and X = 99
We need 0.01 in the left tail and 0.99 in the rest
of the distribution. How many standard
deviations is this?
Table Entries

.00 .01 .02 .03 .04
-2.3 .0107 .0104 .0102 .0099 .0096
18
Microsoft Example continued
The 1-day 99% VaR for the $10 million
position is

The 10-day 99% VaR is

Were 99% sure that we wont lose over
$1,473,621 over the next 10 days

19
AT&T Example
Find 10-day 99% VaR
Consider a position of $5 million in AT&T
The daily volatility of AT&T is 1% (approx
16% per year)
The Std. Dev. for 1 day is

The S.D per 10 days is

The VaR is

20
Portfolio
Now consider a portfolio consisting of both
Microsoft and AT&T
Suppose that the correlation between the
returns is 0.3
21
S.D. of Portfolio
A standard result in statistics states that

Where o
X
= 200,000, o
Y
= 50,000, and =
0.3.
The standard deviation of the change in
the portfolio value in one day is:

Y
X Y X Y X
o o + o + o = o
+
2
2 2
22
VaR for Portfolio
The 10-day 99% VaR for the portfolio is

The individual VaRs were:
Microsoft $1,473,621
AT&T $ 368,405
So, the benefits of diversification are
(1,473,621+368,405)1,622,657=$219,369
657 , 622 , 1 $ 33 . 2 = 10 x x $220,227
23
Diversification Benefits
The diversification benefit is due to a
correlation coefficient less than +1. This is
the same result that you were introduced
to in your investments classes. As
securities are added to a portfolio, the risk
declines.
24
Generalizing The Linear Model
We assume
The daily change in the value of a portfolio
is linearly related to the daily returns from
market variables
The returns from the market variables are
normally distributed
25
Markowitz Result for Variance of
Return on Portfolio
s instrument th and
th of returns between n correlatio is
portfolio in
instrument th on return of variance is
portfolio in instrument th of weight is
Return Portfolio of Variance
ij
2
i
j
i
i
i w
w w
i
n
i
n
j
j i j i ij
o
o o =
= = 1 1
26
VaR Result for Variance of
Portfolio Value (o
i
= w
i
P)

day per value portfolio the in change the of SD the is
return) daily of SD (i.e., instrument th of volatility daily the is
P
i
n
i
j i j i
j i
ij i i P
n
i
n
j
j i j i ij P
n
i
i i
i
x P
o
o
o o o o + o o = o
o o o o = o
A o = A

= <
= =
=
1
2 2 2
1 1
2
1
2
27
Microsoft and AT&T Example
o
p
2
=
1
2
o
1
2
+
2
2
o
2
2
+ 2[
1,2

1
2
o
1
o
2
]

The 10-day 99% VaR is:
28
Diversification Benefits
The diversification benefit is due to a
correlation coefficient less than +1. This is
the same result that you were introduced
to in your investments classes. As
securities are added to a portfolio, the risk
declines.
29
Handling Interest Rates
Bond Portfolios
We do not want to define every bond as a
different market variable
We therefore choose as assets zero-coupon
bonds with standard maturities: 1-month, 3
months, 1 year, 2 years, 5 years, 7 years, 10
years, and 30 years
Cash flows from instruments in the portfolio are
mapped to bonds with the standard maturities

30
Cash Flow Mapping
Assume a portfolio containing two bonds:
Bond #1 Bond #2
1.2 Yr Maturity 0.8 Yr Maturity
Semi-annual pmts Semi-annual pmts
Cash Flows at: Cash Flows at:
0.2 Yrs (2.4 months) 0.3 Yrs (3.6 months)
0.7 Yrs (8.4 months) 0.8 Yrs (9.6 months)
1.2 Yrs (14.4 months)

31
Cash Flow Mapping
Bond 1 CFs 1 Month Bond 2 CFs
2.4 months 3 Months 3.6 months
8.4 months 6 Months 9.6 months
14.4 months 1 Year
2 Years
5 Years
7 Years
10 Years
30 Years
32
Cash Flow Mapping
The values assigned to each standard
category are based upon interpolations of
the actual interest and principal payments.
For a change in market interest rates, the
resulting change in bond values is
computed for the 9 categories shown.
Only 9 standard deviations are required
with cash flow mapping.
33
Cash Flow Mapping
Correlation Coefficients
1 Mo 3 Mo 6 Mo 1 Yr 2 Yr 5 Yr 7 Yr 10 Yr 30 Yr
1 Mo
1m,1m

1m,3m
1m,6m
1m,1y
1m,2y
1m,5y
1m,7y
1m,10y
1m,30y
3 Mo
3m,3m
3m,6m
3m,1y
3m,2y
3m,5y
3m,7y
3m,10y
3m,30y
6 Mo
6m,6m
6m,1y
6m,2y
6m,5y
6m,7y
6m,10y
6m,30y
1 Yr
1y,1y
1y,2y
1y,5y
1y,7y
1y,10y
1y,30y
2 Yr
2y,2y
2y,5y
2y,7y
2y,10y
2y,30y
5 Yr
5y,5y
5y,7y
5y,10y
5y,30y
7 Yr
7y,7y
7y,10y
7y,30y
10 Yr
10y,10y
10y,30y
30 Yr
30y,30y
34
When Linear Model Can be Used
Portfolio of stocks
Portfolio of bonds
Cash flow mapping
Forward contract on foreign currency
Value as bonds and use cash flow mapping
Interest-rate swap
Value as bonds and use cash flow mapping
35
The Linear Model is Only an
Approximation for Options
T d
T
T r K S
d
T
T r K S
d where
d N S d N e K p
d N e K d N S c
rT
rT
o
o
o
o
o
=
+
=
+ +
=
=
=
1
0
2
0
1
1 0 2
2 1 0
) 2 /
2
( ) / ln(
) 2 /
2
( ) / ln(
) ( ) (
) ( ) (
36
The Greek Letters
Delta is the rate of change of the option value
with respect to the price
Gamma is the rate of change of delta with
respect to the price of the underlying asset
Theta is the rate of change of the option value
with respect to the passage of time
Vega is the rate of change of the option value
with respect to volatility
Rho is the rate of change of the option value
with respect to the interest rate

37
38
Delta ()
Option value function is not linear.

Option
price
A
B
Slope = A
Stock price
The Linear Model and Options

To calculate the VaR of an option portfolio
using the linear model, we need the standard
deviation of the price changes. We obtain
the o by utilizing the delta that is defined in
Chapter 17
Consider a portfolio of options dependent on
a single stock price, S.
Define and

S
P
A
A
= o
S
S
x
A
= A
39
Linear Model and Options
continued (equations 20.3 and 20.4)
As an approximation

For a portfolio of many underlying market
variables

where o
i
is the delta of the portfolio with
respect to the ith asset

x S S P A o = A o = A
A o = A
i
i i i
x S P
40
Example
Consider an investment in options on Microsoft and
AT&T.
MSFT AT&T
S = $120 $30
= 1,000 20,000
o = 2% 1%
= 0.3
As an approximation

where Ax
1
and Ax
2
are the percentage changes in
the two stock prices

2 1
000 , 20 30 000 , 1 120 x x P A + A = A
41
Example
Using the equation for variance and substituting o
1
for
Ax
1
and o
2
for Ax
2
, the variance of the portfolio (in
thousands) is:
o
2
port
= 120
2
(.02)
2
+ 600
2
(.01)
2
+ 2(.3)(120)(600)(.02)(01)
o
2
port
= 50.4 and o
port
= 7.09929574
Find the 5-day 95% VaR of the portfolio.
Determine the number of standard deviations using
the normal distribution table. Value of 0.05 falls
between 1.64 and 1.65. Using 1.65 results in:
VaR = 1.65 x 7.09929574 x = $26.193 (000)
5
42
But the distribution of the daily
return on an option is not normal
(See Figure 20.4, page 444)

Positive Gamma
Negative Gamma
43
Skewness

For options, the linear model is an approximation.
It fails to capture the skewness in the probability
distribution of the portfolio value.
As chapter 17 points out, the delta is computed at
a single point. Gamma provides a measure of
curvature and thus, the degree of skewness.
The VaR is dependent on the left tail of the
probability distribution. When skewness occurs,
there is an error in the VaR estimate.
44
Skewness
When Gamma is positive, the left tail of
the distribution is less heavy than normal,
and the VaR estimate is too high.
When Gamma is negative, the left tail of
the distribution is more heavy than normal,
and the VaR estimate is too low.
How is the error corrected? By using a
quadratic equation (rather than linear)
45
Quadratic Model
to correct skewness error

For a portfolio dependent on a single
stock price

where is the gamma of the portfolio.
This becomes

2
) (
2
1
S S P A + A o = A
2 2
) (
2
1
x S x S P A + A o = A
46
Use of Quadratic Model
Analytic results are not as readily available
Monte Carlo simulation can be used in
conjunction with the quadratic model (This
avoids the need to revalue the portfolio for
each simulation trial)
The quadratic model is also sometimes
used in conjunction with historical
simulation
47
Estimating Volatility for Model
Building Approach (equation 20.6)
Define o
n
as the volatility per day on day n, as
estimated at end of day n-1
Define S
i
as the value of market variable at end of
day i
Define u
i
= ln(S
i
/S
i-1
)
The usual estimate of volatility from m observations
is:

=

=

=
= o
m
i
i n
m
i
i n n
u
m
u
u u
m
1
1
2 2
1
) (
1
1
48
Simplifications
(equations 20.7 and 20.8)
Define u
i
as (S
i
S
i-1
)/S
i-1
Assume that the mean value of u
i
is zero
Replace m1 by m

This gives

o
n n i
i
m
m
u
2 2
1
1
=

=
49
Weighting Scheme
Instead of assigning equal weights to the
observations we can set

o o
o
n i n i
i
m
i
i
m
u
2 2
1
1
1
=
=
=
=
where
50
EWMA Model (equation 20.10)
In an exponentially weighted moving
average model, the weights assigned to
the u
2
decline exponentially as we move
back through time
This leads to

2
1
2
1
2
) 1 (

+ o = o
n n n
u
51
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 through time
The same method can be used to find
correlations between market variables
when computing VaR
52
Correlations
Define u
i
=(U
i
-U
i-1
)/U
i-1
and v
i
=(V
i
-V
i-1
)/V
i-1
Also
o
u,n
: daily vol of U calculated on day n-1
o
v,n
: daily vol of V calculated on day n-1
cov
n
: covariance calculated on day n-1
cov
n
=
n
o
u,n
o
v,n
where
n
is the correlation between U and V

53
Correlations continued
(equation 20.12)

Using the EWMA
cov
n
= cov
n-1
+(1-)u
n-1
v
n-1

RiskMetrics uses = 0.94 for daily
volatility forecasting

54
Comparison of Approaches
Advantages Disadvantages
Model
Building
Results produced
quickly
Accommodates
volatility estimating
approach like EWMA
Assumes multivariate
normal distributions
Gives poor results for
low delta portfolios
Historical
Simulation
Determines the joint
probability
distributions of the
variables
Avoids the need for
cash flow mapping
Computationally slow
Doesnt easily allow
volatility updating
schemes such as EWMA
55
Back-Testing

Tests how well VaR estimates would
have performed in the past
We could ask the question: How often
was the loss greater than the VaR
level

56
Stress Testing

This involves testing how well a
portfolio would perform under some
of the most extreme market moves
seen in the last 10 to 20 years
57

CH 20 Hull Fundamentals 7 The D

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

CH 20 Hull Fundamentals 7 The D

Hochgeladen von

Copyright:

Verfügbare Formate

Fundamentals of Futures and Options Markets, 7th Ed, Ch 20, Copyright John C.

Das könnte Ihnen auch gefallen