CHAPTER 15

Market Risk

McGraw-Hill/Irwin Copyright © 2011 by The McGraw-Hill Companies, Inc. All Rights Reserved.
 Overview
 This chapter discusses the nature of
market risk and appropriate measures
– RiskMetrics
– Historic or back simulation
– Monte Carlo simulation
– Links between market risk and capital
requirements

15-2
 Market Risk
 The trading portfolio: contains assets,
liabilities and derivative contracts that can
be quickly bought or sold on Financial
markets.

 Investment portfolio: contains assets,

liabilities, that are relatively illiquid and
held for longer holding periods.

10-3
 Market Risk
 Also, a differentiation should be made
between income from trading activities (i.e.
trading portfolio), and that from FI’s
traditional activities (i.e. investment
portfolio).

 So, we concern about earnings uncertainty

that results from fluctuations in trading
portfolio, therefore, trading portfolio is the
reason behind market risk.

10-4
 Market Risk
 Market risk: the risk related to the
uncertainty of an FI’s earnings on its
trading portfolio resulted from the
changes in market conditions such as
the price of an asset, interest rate …
etc.
 the risk related to the uncertainty
resulting from changes in market
prices
10-5
 Market Risk
 Market risk emphasizes the risks to FI that
actively trade assets and liabilities rather
than hold them for longer-term
(investment, funding or hedging purposes.
– Affected by other risks such as interest rate
risk and FX risk
– Can be measured over periods as short as
one day or as longer as a year.

10-6
 Market Risk Measurement
 FI, specifically, are concerned about
the fluctuation in value (value at risk)
of their trading account assets and
liabilities for periods as short as one
day; so-called daily earnings at risk,
DEAR.

 Usually measured in terms of dollar exposure

amount or as a relative amount against some
benchmark
10-7
 Market Risk Measurement
 Important in terms of:
– Management information (provides
management with information on the
risk exposure take by FI traders, which
will be compared with FI’s capital
resources).
– Setting limits ( will lead to the
establishment of economically logical
position limits per trader in each area of
trading).

10-8
Market Risk Measurement
– Resource allocation (risk/return tradeoff):
that is, it compares returns to market risk in
different trading areas, and then allow for
identification of areas with greatest potential
return per unit of risk into which more
capital and resources can be directed.
– Performance evaluation: considers the
return-risk ratio of traders, which allows to
identifies those with lower returns and risk
exposures and those with higher returns and
risk exposure.

10-9
Market Risk Measurement
– Regulation
 Provide regulators with information on how
to protect banks and financial system
against failure due to extreme market risk.

 Allowances for use of internal models to

calculate capital requirements

10-10
 Calculating Market Risk
Exposure
 Many market risk models are developed.
 Generally concerned with estimated
potential loss under adverse circumstances
 Three major approaches of measurement:
– JPM RiskMetrics (or variance/covariance
approach)
– Historic or Back Simulation
– Monte Carlo Simulation

10-11
 RiskMetrics Model
– The ultimate objective of the market risk
measurement can be seen in this
statement; “At close of business each
day tell me what the market risks are
cross all the business and locations”
– Or it can be seen in this “ I am X% sure
that the FI will not lose more than $VAR
in the next T days”
– FIs have an active trading positions in
fixed income securities, FX, derivatives,
proprietary assets…etc.
10-12
 RiskMetrics Model
– So FI’s managers need to know a single
dollar number to tell what FI’s market
risk exposure will be next day-
especially if those days turn out to be
extremely “bad” days.
– That is, FI concentrates on measuring
the market risk exposure at the close of
business day (on daily basis), because it
concerns with how to preserve equity if
market conditions move adversely next
day.
10-13
 RiskMetrics Model
– Market risk = Estimated potential
loss under adverse circumstances.
– Idea is to determine the daily earnings at
risk, which has three components
= dollar market value of position × price
sensitivity × potential adverse move in
yield or,
DEAR = dollar market value of position ×
price volatility.

10-14
 RiskMetrics Model
Where,
price volatility = price sensitivity of position
× potential adverse move in yield

10-15
 Market risk of Fixed-Income
Securities
 When FI hold Fixed-income securities in
its trading portfolio, then it want to know
potential exposure that it could face
should interest rate move against the FI.

 The amount of loss depends on the

bond’s price volatility.

10-16
 Market risk of Fixed-Income
Securities
 Daily price volatility can be stated as:
DPV = (MD) × (potential adverse daily
yield move)
where,
MD = D/(1+R).
MD = Modified duration
D = Macaulay duration

10-17
 Market risk of Fixed-Income
Securities
– Example: Suppose that we are long in 7-
year zero-coupon bonds, with face value
$1,631,483 and 7.243% yield.
– The market value of this position is
&1,000,000.
– The risk in this case will be result from
the change in the yield (i.e. an increase in
the interest rate).
– So, we need to define the probability of
“bad” yield changes; suppose that there
is only a 5% chance of the yield change.
10-18
 Market risk of Fixed-Income
Securities
– Based on the probability of outcome; The
point here is to determine the amount of
the adverse change in the yield.

– To do so we use the normal distribution

(of the past changes in yields of 7 year
zero coupon bond) to get an estimate of
the size of the adverse rate move through
building of the confidence intervals.

10-19
 Confidence Intervals
– If we assume that changes in the yield are
normally distributed, we can construct
confidence intervals around the projected
DEAR (other distributions can be
accommodated but normal is generally
sufficient)
– Assuming normality, 90% of the time the
disturbance (i.e. yield changes )will be within
±1.65 standard deviations of the mean
 (5% of the extreme values remain in each tail
of the distribution)
 10% of the area under the normal distribution
is found beyond ±1.65σ . 10-20
Confidence Intervals: Example
– Thus, the adverse move in the yield will be
±1.65 σ (the yield have fluctuated by more
than ±1.65 σ )
– Concern is that yields will rise more than
the expected deviation. If the standard
deviation is 10 basis points. this
corresponds to 16.5 basis points.

10-21
Adverse 7-Year Rate Move

10-22
Confidence Intervals: Example
 Yield on the bonds = 7.243%, so MD =
6.527 years
 Price volatility = (MD)  (Potential
adverse change in yield)
= (6.527)  (0.00165) = 1.077%
DEAR = Market value of position 
(Price volatility)
= ($1,000,000)  (.01077) =
$10,770 10-23
N

Confidence Intervals: Example

 To calculate the potential loss for more
than one day:
Market value at risk
(VARN) = DEAR × N
 Example:
For a five-day period,
VAR5 = $10,770 5×
= $24,082

10-24
 Foreign Exchange
 In the case of foreign exchange,
DEAR is computed in the same
fashion we employed for interest rate
risk
 DEAR = dollar value of position × FX
rate volatility, where the FX rate
volatility is taken as 1.65 FX

10-25
 Foreign Exchange
 Example: Suppose the FI had a 1.4 million
trading position in spot Euros.
 The risk exposure in this case, is that the
next day will be the bad day with respect
to the value of the euro against dollar.
 If the volatility or standard deviation of
daily changes in the spot exchange rate is
56.5 bp.
 Suppose that FI is interested in adverse
move; the probability of bad moves is 5%.
10-26
 Foreign Exchange
 So if historical exchange rats are normally
distributed, then the exchange rates must
change in the adverse direction (i.e.
volatility) by 1.65 X .

10-27
 Foreign Exchange
 First: find the dollar value of the position: FX
position X exchange rate ($ per unit of
foreign currency)
 Dollar value of position = 1.4 X 0.714286 = $
1m
 Second: DEAR = Dollar value position X FX
volatility
 DEAR= $1m X (1.65 x 56.5 bp) = $9,320

10-28
 Equities
 For equities, total risk = systematic risk +
unsystematic risk
 Systematic risk reflects the co-movements
of the stock return with market portfolio and
it is measured by beta
 Unsystematic risk is the risk related to the
firm itself.
 If the portfolio is well diversified, FI’s
portfolio follows stock market portfolio, then
the standard deviation of the portfolio will
be equal to the standard deviation of the
stock market index. 10-29
 Equities
 If the portfolio is well diversified, then
DEAR = dollar value of position × stock
market return volatility, where
market volatility taken as 1.65 m

 If not well diversified, a degree of error will

be built into the DEAR calculation
 That is, DEAR will not depend only on
systematic risk but also the effect of
unsystematic risk should be considered.
10-30
 Equities
 If FI hold $1 million trading position in
stocks, which reflects the U.S stock market,
assuming that, over the last year, the
deviation in daily returns (Standard
deviation) on the stock market index was
200 bp, and the adverse move in returns is
expected to occur 5% of the times. then
DEAR = $ value of position × stock market
return volatility
= $1,000,000 x1.65 x 0.02 =
$33,000
10-31
 Aggregating DEAR Estimates
 Cannot simply sum up individual DEARs, and
consider the total as aggregate risk of the entire
trading position. Why?
 Because these individual trading positions
represent a trading portfolio in which we cannot
ignore any degree of offsetting covariance or
correlation among different trading positions.
 More specifically, the adverse moves related to
each trading position may be negatively related.
 In this case, we apply the portfolio theory in
calculating the risk for the entire trading portfolio.
10-32
 Aggregating DEAR Estimates
 In order to aggregate the DEARs from
individual exposures we require the correlation
matrix.
 Three-asset case:
DEAR portfolio = [DEARa2 + DEARb2 +
DEARc2 + 2ab × DEARa × DEARb + 2ac ×
DEARa × DEARc + 2bc × DEARb × DEARc]1/2

10-33
DEAR: Large US Banks 2005 &
2008

10-34
 Historic or Back Simulation
 Risk Metrics model assume a symmetric
distribution for all asset returns.
 Because of this, new market risk model has
been developed; Historic or Back Simulation
model, which is:
- simple
- does not require the normality assumption.
- does not require the calculation of the
correlation or standard deviation of assets
return.
10-35
 Historic or Back Simulation
 Basic idea: Revalue portfolio based on
actual prices (returns) on the assets
that existed yesterday, the day before
that, etc. (usually previous 500 days)
 Then calculate 5% worst-case (25th
lowest value of 500 days) outcomes
 That is, only 5% of the time, the value
of the portfolio would fall below this
number (i.e. 25th lowest value).
10-36
Estimation of VAR: Example
 Convert today’s FX positions into dollar
equivalents at today’s FX rates
 Measure sensitivity of each position
– Calculate its delta
 Measure risk
– Actual percentage changes in FX rates for
each of past 500 days
 Rank days by risk from worst to best
10-37
 Historic or Back Simulation
 Advantages:
– Simplicity
– Does not need correlations or standard
deviations of individual asset returns
– Does not require normal distribution of
returns (which is a critical assumption
for RiskMetrics)
– Directly provides a worst case value,
while Risk Metrics does not.
10-38
 Weaknesses
 Disadvantage: 500 observations is not
very many from a statistical standpoint
 Increasing number of observations by
going back further in time is not
desirable
 Risk manager could either weight
recent observations more heavily and
go further back or use Monte Carlo
Simulation approach.
10-39
 Monte Carlo Simulation
 To overcome problem of limited number of
observations, additional observations can
be generated.
 This can be done by Monte Carlo
Simulation: is a tool for considering portfolio
valuation under all possible combinations of
factors that determine a security’s value.
 This model generated random market
values drawn from the multivariate normal
distributions representing each variable.

10-40
 Monte Carlo Simulation
 Employs historic covariance matrix
and random number generator to
obtain the additional observations
– Objective is to replicate the distribution of
observed outcomes with synthetic data
 These additional observations reflect
the probability with which they have
occurred in recent historic time
periods.

10-41