Ch13 Model Building Approaches

Market Risk VaR: ModelBuilding Approach
Chapter 13
Market Risk VaR: Model Building Approach

13:1 The Basic Methodology
13:2 Generalization
13:3 Correlation and Covariance Matrices
13:4 Handling the Interest Rate
13:5 Applications of the Linear Model
13:6 Linear Model and Options
13:7 Quadratic Model
13:8 Monte Carlo Simulation
13:9 Non-normal Distribution
13:10 Model Building vs. Historical Simulation
2
The Model-Building Approach

The main alternative to historical simulation is to
make assumptions about the probability
distributions of the returns on the market
variables
This is known as the model building approach (or
sometimes the variance-covariance approach).
Assume a joint distribution of changes in market
variables and using historical data to estimate
model parameters.
Risk Management and Financial Institutions 3e, Chapter 15, Copyright John C. Hull 2012
The Methodology
The
portfolio consists of one security:

(Microsoft stock)
We have a position worth $10 million in
Microsoft shares
The volatility of Microsoft is 2% per day
(about 32% per year)
We use N=10 and X=99
20,$10632,45
Microsoft Example continued

The
standard deviation of the change in

the portfolio in 1 day is $200,000
The standard deviation of the change in
10 days is
Microsoft Example continued

We
assume that the expected change in

the value of the portfolio is zero
We assume that the change in the value
of the portfolio is normally distributed
Since N(2.33)=0.01, the VaR is
508,$,412.35$8,1640
AT&T Example
Consider
a position of $5 million in AT&T

The daily volatility of AT&T is 1% (approx
16% per year)
The SD per 10 days is
The
VaR is
Portfolio
Now
consider a portfolio consisting of both

Microsoft and AT&T
Suppose that the correlation between the
returns is 0.3
2
2
XYXYXY
S.D. of Portfolio
A
standard result in statistics states that
this case X = 200,000 andY = 50,000

and = 0.3. The standard deviation of the
change in the portfolio value in one day is
therefore 220,227
In
20,7102.3$1,6257
VaR for Portfolio

The
The
10-day 99% VaR for the portfolio is
benefits of diversification are

(1,473,621+368,405)1,622,657=$219,369
Less than perfect correlation leads to
some of the risk being diversified away.
10
13:2 Generalization:
Daily changes in the value of a portfolio equal the total daily

changes in the values of individual assets:
n
P i xi
i 1
If daily changes of the values of individual assets are normally

distributed then daily changes in the value of the portfolio are
normally distributed.
The variance of the daily changes of portfolio value is given by:
n
i2 i2 2 ij i j i j
2
P
i 1
i j
i is the daily volatility of the ith asset (i.e., SD of daily returns)

P is the SD of the change in the portfolio value per day
i =wi P is amount invested in ith asset
11
13:3 Correlation and Covariance Matrices

Instead of working with correlations and volatilities analysts
use variances and co variances matrix.
n
P2 cov ij i j
i 1 j 1
var1
cov 21
C cov 31

cov
n1
cov1n
cov 2 n
cov 3n
cov12
cov13
var2
cov 32
cov 23
var3
cov n 2
cov n 3

varn
12
Example Involving Four Investments

Correlations and Covariance matrices are carried out
using the samer weights
The variance of the portfolio is calculated using the
following formula:
cov ij i j
2
P
i 1 j 1
Then VaR of the portfolio is estimated

Exponentially weighted average method with certain
value could be used following the same procedure.
13
Volatilities and Correlations

Volatilities (% per day)
DJIA
FTSE
CAC
Nikkei
Equal Weights
1.11
1.42
1.40
1.38
EWMA
2.19
3.21
3.09
1.59
Correlations
1
0.489 0.496 0.062
0.489
1
0.918 0.201
0.496 0.918
1
0.211
0.062 0.201 0.211

1
Equal weights
1
0.611 0.629 0.113
0.611
1
0.971 0.409
0.629 0.971
1
0.342
0.113 0.409 0.342

1
EWMA
14
13:4 Handling Interest Rates
Model building approach (Variance Covariance Approach) suits

portfolios consisting of short and long positions in fixed income
securities.
A separate market variable for every single bond price is defined.
For simplification, a Duration Approach is followed: Linear relation
between P and y (but assumes parallel shifts)
For VaR calculations the cash flows from instruments in the portfolio
are mapped into cash flows occurring on the standard maturities,
choosing zero-coupon bonds with standard maturities (about 10
different maturities) as market variables; cash-flow mapping
Using the standard volatilities and correlations, the variance of the
portfolio and its VaR is estimated.
15
Handling Interest Rates: Cash Flow

Mapping
Zero-coupon bond prices with standard maturities (1mm, 3mm,

6mm, 1yr, 2yr, 5yr, 7yr, 10yr, 30yr) are chosen as market variables.
Suppose that the 5yr rate is 6% and the 7yr rate is 7% and we will
receive a cash flow of $10,000 in 6.5 years.
The volatilities per day of the 5yr and 7yr bonds are 0.50% and
0.58% respectively
16
Example continued
We
1.06,7056.56,540
interpolate between the 5yr rate of 6%

and the 7yr rate of 7% to get a 6.5yr rate
of 6.75%
The PV of the $10,000 cash flow is
17
Example continued
We
interpolate between the 0.5% volatility

for the 5yr bond price and the 0.58%
volatility for the 7yr bond price to get
0.56% as the volatility for the 6.5yr bond
We allocate of the PV to the 5yr bond
and (1- ) of the PV to the 7yr bond
18
0.56220.582(1)220.650.8(1)
Example continued
Suppose
that the correlation between

movement in the 5yr and 7yr bond prices
is 0.6
To match variances
This
gives =0.074
19
6,540.92764$,08546
Example continued
The value of 6,540 received in 6.5 years
in 5 years and
in 7 years.
This cash flow mapping preserves value and variance

Using volatilities and correlations for standard maturities
VaR is estimated with all exposures at these
maturities being estimated
20
13: 5 Applications of The Linear Model

In all of the previous analysis we assume the change in portfolio value
is linearly dependent in the percentage change in the value of the
underlying asset/s
Portfolio of stocks
Portfolio of bonds
Forward contract on foreign currency
Interest-rate swap
21

S
x
13:6 The Linear Model and Options
Consider a portfolio of options dependent

on a single stock price, S. Define delta of
the portfolio:
P
S
and
22
PP
SSixix
Linear Model and Options

continued
As an approximation
Similarly when there are many underlying

market variables
where i is the delta of the portfolio with

respect to the ith asset
23
Example
Consider an investment in options on Microsoft and

AT&T. Suppose the stock prices are 120 and 30
respectively and the deltas of the portfolio with respect to
the two stock prices are 1,000 and 20,000 respectively
As an approximation
P 120 1,000x1 30 20,000x2
where x1 and x2 are the percentage changes in the two

stock prices
If the daily volatilities of Microsoft and AT&T are 2% and 1%
respectively, the correlation coefficient is .3. the SD of
Pbe 7.1$ and the 99% five-day VaR is 37$.
will
24
But the Distribution of the Daily

Return on an Option is not Normal
The linear model fails to capture skewness
in the probability distribution of the
portfolio value.
25
Impact of Gamma
Delta: The rate of change in the portfolio value relative to underlying market
variable.
Gamma: The rate of change Delta with respect to market variable; measures
the curvature of the relationship between portfolio value and the underlying
asset.
Positive Gamma
Negative Gamma
26
Translation of Asset Price Change

to Price Change for Long Call
(Figure 15.2, page 337)
Long Call
Asset Price
27
Translation of Asset Price Change

to Price Change for Short Call
(Figure 15.3, page 338)
Asset Price
Short
Call
28
S
(x22
S)2x
13:7 Quadratic Model
For a portfolio dependent on a single asset price it is

approximately true that
so that
Moments are
E (P ) 0.5S 2 2
E (P 2 ) S 2 2 2 0.75S 4 2 4
E (P 3 ) 4.5S 4 2 4 1.875S 6 3 6
29
Quadratic Model continued

When
there are a small number of underlying

market variable moments can be calculated
analytically from the delta/gamma
approximation
The Cornish Fisher expansion can then be
used to convert moments to fractiles
However when the number of market
variables becomes large this is no longer
feasible
30
Cornish Fisher Expansion
31
Cornish Fisher Expansion
32
Example: Cornish Fisher Expansion
33
13: 8 Monte Carlo Simulation

To calculate VaR using MC simulation we
Value portfolio today
Sample once from the multivariate
distributions of the xi
the xi to determine market
variables at end of one day
Revalue the portfolio at the end of day
Use
34
Monte Carlo Simulation continued

P
Repeat many times to build up a
probability distribution for P
VaR is the appropriate fractile of the
distribution times square root of N
For example, with 1,000 trial the 1
percentile is the 10th worst case.
Calculate
35
Speeding up Calculations with the

Partial Simulation Approach
Use
the approximate delta/gamma

relationship between P and the xi to
calculate the change in value of the
portfolio
This is also a way of speeding up
computations in the historical simulation
approach
36
13:9 Non-Normal Distribution

Assumption in Monte Carlo
In
a Monte Carlo simulation we can

assume non-normal distributions for the xi
(e.g., a multivariate t-distribution)
Can also use a Gaussian or other copula
model in conjunction with empirical
distributions
37
13: 10 Model Building vs

Historical Simulation
Model building approach can be used for
investment portfolios where there are no
derivatives, but it does not usually work
when for portfolios where
There are derivatives
Positions are close to delta neutral
38
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
39

Ch13 Model Building Approaches

Uploaded by

Document Information

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

Ch13 Model Building Approaches

Uploaded by

Copyright:

Available Formats

Market Risk VaR: ModelBuilding Approach

Market Risk VaR: Model Building Approach

The Model-Building Approach

portfolio consists of one security:

Microsoft Example continued

standard deviation of the change in

Microsoft Example continued

assume that the expected change in

a position of $5 million in AT&T

consider a portfolio consisting of both

standard result in statistics states that

this case X = 200,000 andY = 50,000

VaR for Portfolio

10-day 99% VaR for the portfolio is

benefits of diversification are

Daily changes in the value of a portfolio equal the total daily

If daily changes of the values of individual assets are normally

i is the daily volatility of the ith asset (i.e., SD of daily returns)

13:3 Correlation and Covariance Matrices

Example Involving Four Investments

Then VaR of the portfolio is estimated

Volatilities and Correlations

0.062 0.201 0.211

0.113 0.409 0.342

13:4 Handling Interest Rates

Model building approach (Variance Covariance Approach) suits

Handling Interest Rates: Cash Flow

Zero-coupon bond prices with standard maturities (1mm, 3mm,

interpolate between the 5yr rate of 6%

interpolate between the 0.5% volatility

that the correlation between

The value of 6,540 received in 6.5 years

This cash flow mapping preserves value and variance

13: 5 Applications of The Linear Model

13:6 The Linear Model and Options

Consider a portfolio of options dependent

Linear Model and Options

Similarly when there are many underlying

where i is the delta of the portfolio with

Consider an investment in options on Microsoft and

P 120 1,000x1 30 20,000x2

where x1 and x2 are the percentage changes in the two

But the Distribution of the Daily

Translation of Asset Price Change

Translation of Asset Price Change

13:7 Quadratic Model

For a portfolio dependent on a single asset price it is

Quadratic Model continued

there are a small number of underlying

Cornish Fisher Expansion

Cornish Fisher Expansion

Example: Cornish Fisher Expansion

13: 8 Monte Carlo Simulation

Monte Carlo Simulation continued

Speeding up Calculations with the

the approximate delta/gamma

13:9 Non-Normal Distribution

a Monte Carlo simulation we can

13: 10 Model Building vs

The Linear Model

You might also like