Chap 10 Market Risk

Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 115

Chapter 10

Market Risk

McGraw-Hill/Irwin © 2008 The McGraw-Hill Companies, Inc., All Rights Reserved.


10-2
Overview

 This chapter discusses the nature of


market risk and appropriate measures
 Dollar exposure
 Risk Metrics
 Historic or back simulation
 Monte Carlo simulation
 Links between market risk and capital
requirements
10-3
Market Risk
 Market risk can be defined as the risk related to the
uncertainty of an FI’s earnings on its trading portfolio
caused by changes, and particularly extreme changes,
in market conditions such as the price of an asset,
interest rates, market volatility, and market liquidity.

 Thus, risks such as interest rate risk, credit risk,


liquidity risk and foreign exchange risk affect market
risk.
 So important is market risk in determining the viability
of an FI that since 1998, U.S. regulators have included
market risk in determining the required level of capital
an FI must hold
10-4
Trading Risks

 Trading exposes banks to risks


 1995 Barings Bank
 1996 Sumitomo Corp. lost $2.6 billion in
commodity futures trading
 AllFirst/ Allied Irish $691 million loss
 Allfirst eventually sold to Buffalo based M&T
Bank due to dissatisfaction among stockholders
of Allied Irish
10-5
Trading Book
10-6
Market Risk
 Market risk is the uncertainty resulting
from changes in market prices .

 Affected by other risks such as interest rate


risk and FX risk
 It can be measured over periods as short as
one day.
 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
 Setting limits
 Resource allocation (risk/return tradeoff)
 Performance evaluation
 Regulation
 BIS and Fed regulate market risk via capital
requirements leading to potential for overpricing
of risks
 Allowances for use of internal models to
calculate capital requirements
10-8
Calculating Market Risk Exposure

 Generally concerned with estimated


potential loss under adverse
circumstances.
 Three major approaches of measurement
 JPM Risk Metrics (or variance/covariance
approach)
 Historic or Back Simulation
 Monte Carlo Simulation
10-9
Market Risk Measurement
 The ultimate objective of market risk measurement
models can best be seen from the following question
from an FI manager: “I am X% sure that the FI will not
lose more than $VAR in the next T days.”

 In a nutshell, the FI manager wants a single dollar


number that tells him the FI’s market risk exposure
over the next days— especially if those days turn out
to be extremely “bad” days.
10-10
JP Morgan RiskMetrics Model
Market Risk = Estimated potential loss under
adverse circumstances
Or,
Market Risk = Daily earnings at Risk (DEAR)
10-11
JP Morgan RiskMetrics Model
10-12
JP Morgan risk Metric Model
 We concentrate on how the RiskMetrics model
calculates daily earnings at risk in three trading
areas—
1) fixed income,
2) foreign exchange (FX)
3) Equities

then on how it estimates the aggregate risk of the


entire trading portfolio to meet an FI manager’s
objective of a single aggregate dollar exposure
measure across the whole bank on a given day.
10-13
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 will be within 1.65 standard
deviations of the mean.
 (5% of the extreme values greater than +1.65
standard deviations {95% Confidence limit} and
5% of the extreme values less than -1.65 standard
deviations)
10-14
Adverse 7-Year Rate Move
10-15
Confidence Intervals: Example
 Suppose that we are long in 7-year $ 1 million zero-
coupon bonds with 7.243% YTM and we define “bad”
yield changes such that there is only 5% chance of the
yield change being exceeded in either direction.

 Assuming normality, 90% of the time yield changes


will be within 1.65 standard deviations of the mean.

 If the standard deviation is 10 basis points, this makes


potential adverse change in price equal to 16.5 basis
points (1.65*SD = 1.65*10bp = 16.5 bp).

 Concern is that yields will rise. Probability of yield


increases greater than 16.5 basis points is 5%.
10-16
Confidence Intervals: Example
Price volatility
= (MD)  (Potential adverse change in
yield)
=
DEAR
= Market value of position  (Price
volatility)
=
10-17
Confidence Intervals: Example
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
N 10-18
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 ????
N 10-19
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
10-20
Confidence Interval - Example
 From statistics, we know that (the middle) 98 percent
of the area under the normal distribution is to be found
within +/- 2.33 standard deviations from the mean and
2 percent of the area under the normal distribution is
found beyond +/- 2.33 sd (1 percent under each tail or
99% confidence limit for one tail, +2.33sd and -2.33 sd
, respectively).

 For 95 percent of the area under the normal


distribution (2.5 percent under each tail or 97.5%
confidence limit for one tail), we use +/-1.96sd,

 For 90 percent of the area (5 percent under each tail or


95% confidence limit for one tail) we use +/- 1.65sd.
10-21
Confidence Interval - Example
 Suppose that during the last year the mean change in
daily yields on seven-year zero-coupon bonds was 0
percent, while the standard deviation was 10 basis
points (or 0.001).
 Thus, 2.33sd is 23.3 basis points (bp).
 In other words, over the last year, daily yields on
seven-year, zero-coupon bonds have fluctuated (either
positively or negatively) by more than 23.3 bp 2
percent of the time.
 Adverse moves in yields are those that decrease the
value of the security (i.e., the yield increases). These
occurred 1 percent of the time, or 1 in 100 days
10-22
Confidence Intervals: Example
Price volatility
= (MD)  (Potential adverse change in
yield)
=
DEAR
= Market value of position  (Price
volatility)
=
10-23
Confidence Intervals: Example
Price volatility
= (MD)  (Potential adverse change in yield)
= (6.527)  (0.00233) = 1.521%

DEAR
= Market value of position  (Price volatility)
= ($1,000,000)  (.01521) = $15,210

That is, the potential daily loss in earnings on the


$1 million position is $15,210 if the 1 bad day in
100 occurs tomorrow.
10-24
Confidence Intervals: Example
10-25
Risk Metric Models - Practice Q1
10-26
Risk Metric Models - Practice Q1

(a) MD = D ÷ (1 + R)
= 5 ÷ (1.07)
= 4.6729 years
10-27
Risk Metric Models - Practice Q1

Potential adverse move in yield at 1 percent

= 2.33
= 2.33 x 0.0012
= .002796
10-28
Risk Metric Models - Practice Q1

Price volatility
= MD x potential adverse move in yield
= 4.6729 x .002796
= 0.01306 or 1.306percent
10-29
Risk Metric Models - Practice Q1

DEAR
= ($ value of position) x (price volatility)
= $1,000,000 x 0.01306
= $13,060
10-30
Risk Metric Models - Practice Q2

The DEAR for a bank is $8,500. What is the VAR


for a 10-day period? A 20-day period? Why is the
VAR for a 20-day period not twice as much as
that for a 10-day period?
10-31
Risk Metric Models - Practice Q2

 For the 10-day period:


VAR = 8,500 x [10]½
= 8,500 x 3.1623
= $26,879.36
10-32
Risk Metric Models - Practice Q2

 For the 20-day period:


VAR = 8,500 x [20]½
= 8,500 x 4.4721
= $38,013.16

 The reason that VAR20  (2 x VAR10) is because


[20]½  (2 x [10]½).
 The interpretation is that the daily effects of an
adverse event become less as time moves
farther away from the event.
10-33
Risk Metric Models - Practice Q3
10-34
Risk Metric Models - Practice Q3
(a) If yield changes are normally distributed, 99
percent of the area of a normal distribution
will be 2.33 standard deviations (2.33) from
the mean for a one-tailed distribution.

 In this example, it means 2.33 x 15 = 34.95 bp.

 Thus, the maximum adverse yield change


expected for this zero-coupon bond is an
increase of 34.95 basis points, or 0.3495
percent, in interest rates.
10-35
Risk Metric Models - Practice Q3
 If a 95 percent confidence limit is required,
then 95 percent of the area will be 1.65
standard deviations (1.65) from the mean.

 Thus, the maximum adverse yield change


expected for this zero-coupon bond is an
increase of (1.65 x 15 =) 29.40 basis points, or
0.294 percent, in interest rates.
10-36
Risk Metric Models - Practice Q4
10-37
Risk Metric Models - Practice Q4

MD = D/(1 + R)
= 15/(1.095)
= 13.6986.
10-38
Risk Metric Models - Practice Q4

Price volatility
= (MD) x (potential adverse move in yield)

= (13.6986) x (.0025)
= 0.03425 or 3.425 percent.
10-39
Risk Metric Models - Practice Q4
Daily earnings at risk (DEAR)
= ($ Value of position) x (Price volatility)

Dollar value of position


= $400m./(1 + 0.095)15
= $102,529,300.

Therefore,

DEAR = $102,5293,500 x 0.03425


= $3,511,279
10-40
Risk Metric Models - Practice Q4
The potential adverse move in yields

(PAMY) = confidence limit value x standard


deviation value.

 Therefore,

25 basis points = 2.33 x ,


 = .0025/2.33
= .001073 or 10.73 bp.
10-41
Risk Metric Models - Practice Q5
Bank Two has a portfolio of bonds with a
market value of $200 million. The bonds have
an estimated price volatility of 0.95 percent.
What are the DEAR and the 10-day VAR for
these bonds?
10-42
Risk Metric Models - Practice Q5

Daily earnings at risk (DEAR)


= ($ Value of position) x (Price volatility)
= $200 million x .0095
= $1,900,000
10-43
Risk Metric Models - Practice Q5

Value at risk (VAR) = DEAR x N


= $1,900,000 x 10
= $1,900,000 x 3.1623
= $6,008,328
10-44
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
10-45
Foreign Exchange – Example
 Suppose the FI had a :800,000 trading position in spot
euros at the close of business on a particular day.
 The FI wants to calculate the daily earnings at risk
from this position (i.e., the risk exposure on this
position should the next day be a bad day in the FX
markets with respect to the value of the euro against
the dollar).
 The first step is to calculate the dollar value of the
position:
Dollar equivalent value of position
= (FX position) * ($ per unit of foreign currency)
10-46
Foreign Exchange – Example

Suppose for simplicity that the exchange rate is:

Euro0.8000/$1 or $1.25/Euro : at the daily close;

then:
Dollar value of position :
= 800,000 * $1.25
= $ 1 million
10-47
Foreign Exchange – Example
 Suppose that, looking back at the :/$ exchange rate
over the past year, we find that the volatility, or
standard deviation, of daily percentage changes in the
spot exchange rate was 56.5 bp.
 However, suppose that the FI is interested in adverse
moves—that is, bad moves that will not occur more
than 1 percent of the time, or 1 day in every 100.

 FX Volatility = ???
10-48
Foreign Exchange – Example

FX Volatility = 2.33 * 56.5bp


= 131.645 bp

DEAR
= Dollar Value of position * FX Volatlity
= ???
10-49
Foreign Exchange – Example

FX Volatility = 2.33 * 56.5bp


= 131.645 bp

DEAR
= Dollar Value of position * FX Volatlity
= ($1 million ) * 0.0131645
= $13,164.5
10-50
Foreign Exchange – Practice Q6
10-51
Foreign Exchange – Practice Q6

FX position of € = 20m x 0.40 = $8 million


FX position of £ = 25m x 1.28 = $32 million

FX volatility € = 2.33 x 65bp = 151.45bp,


or 1.0545%

FX volatility £ = 2.33 x 45bp = 104.85bp,


or 1.0485%
10-52
Foreign Exchange – Practice Q6
DEAR = ($ Value of position) x (Price volatility)

DEAR of € = $8m x .015145 = $12,1160


DEAR of £ = $32m x .010485 = $33,5520

VAR of € = $12,1160 x 10


= ???
VAR of £ = $33,5520 x 10
= ???
10-53
Foreign Exchange – Practice Q6
DEAR = ($ Value of position) x (Price volatility)

DEAR of € = $8m x .015145 = $12,1160


DEAR of £ = $32m x .010485 = $33,5520

VAR of € = $12,1160 x 10


= $ 12,1160 x 3.1623
= $383,141
VAR of £ = $33,5520 x 10
= $33,5520 x 3.1623
= $106,1007
10-54
Foreign Exchange – Practice Q7
10-55
Foreign Exchange – Practice Q7
VAR of Yen = DEAR x 10
350,000 = DEAR x 3.1623

DEAR Yen = 350,000/3.1623


= 110,678.9362
10-56
Foreign Exchange – Practice Q7
DEAR = ($ Value of position) x (Price volatility)

Price Volatility Yen = 2.33 * 0.0075


= 0.017475

DEAR of YEN = $ Value of Position * .017475


110,678.94 = $ Value of position * 0.01745

$ Value of position = 110,678.94 /0.01745


= $6,342,632.66

Yen Value of Position = 80 Yen/$ * 6,342,632.66


= 50,741,0613.2 yen
10-57
Foreign Exchange – Practice Q7

VAR of SF = DEAR x 10


500,000 = DEAR x 3.1623

DEAR SF = 500,000/3.1623
= 158,112.77
10-58
Foreign Exchange – Practice Q7
DEAR = ($ Value of position) x (Price volatility)

Price Volatility SF = 2.33 * 0.0050


= 0.01165

DEAR of SF = $ Value of Position * .01165


158,112.77 = $ Value of position * 0.01165

$ Value of position = 158,112.77 /0.01165


= $13,571,911.59

SF Value of Position = 0.96 SF/$ * 13,571,911.59


= 13,029,035 SF
10-59
Equities

For equities,
Total risk
= Systematic risk + Unsystematic risk

If the portfolio is well diversified then


DEAR = dollar value of position ×
stock market return volatility
where
the market return volatility is taken as
1.65 M.
10-60
Equities - example

Suppose FI holds $1 million trading position in


stocks that reflect a US stock market index.
Beta = 1. If over the last year SD of market index
was 200 bp then calculate DEAR with 99 %
confidence interval.
10-61
Equities - example

DEAR
= $ Market value of position * Stock market
return volatility
= $1million * (2.33 * 0.0200)
= $1million * (0.0466)
= $46,600
10-62
Risk Metric Models - Practice Q8
Bank of Alaska’s stock portfolio has a market
value of $10 million. The beta of the portfolio
approximates the market portfolio, whose
standard deviation (m) has been estimated at
1.5 percent. What is the 5-day VAR of this
portfolio, using adverse rate changes in the
99th percentile?
10-63
Risk Metric Models - Practice Q8

DEAR = ($ Value of portfolio) x (2.33 x m )


= $10m x (2.33 x .015)
= $10m x .03495
= $349,500
10-64
Risk Metric Models - Practice Q8

VAR = $349,500 x 5
= $349,500 x 2.2361
= $781,506
10-65
Aggregating DEAR Estimates

 Cannot simply sum up individual DEARs.


 In order to aggregate the DEARs from
individual exposures we require the
correlation matrix.
 Three-asset case:

DEAR portfolio = [DEARa2 + DEARb2 +


DEARc2 + 2rab × DEARa × DEARb + 2rac ×
DEARa × DEARc + 2rbc × DEARb ×
DEARc]1/2
10-66
Aggregating DEAR Estimates - Example

 DEAR zero coupon Bond = 15,210


 DEAR Foreign Currency = 13,164
 DEAR Stock index = 46,600
 Corelation bond & FX = -0.2
 Correlation bond & stock = 0.4
 Correlation FX & Stock = 0.1
10-67
Aggregating DEAR Estimates - Example
10-68
Aggregating DEAR Estimates – Practice Q9
10-69
Aggregating DEAR Estimates – Practice Q9

2 0.5
( DEARL )  ( DEARFX )  ( DEAREQ )
2 2

 
  (2 r L , FX x DEARL x DEARFX ) 
portfolio   
 (2 r L , EQ x DEARL x DEAREQ )
 
  (2 r FX , EQ x DEARFX x DEAREQ ) 
 

0.5
$300,700  $274,000  $126,700  2(0.3)($ 300,700)($ 274,000) 
2 2 2
 
 2(0.7)($ 300,700)($ 126,700)  2(0.0)($ 274,000)($ 126,700) 

 $284,322,626,000  $533,219
0. 5
10-70
Aggregating DEAR Estimates – Practice Q10
10-71
Aggregating DEAR Estimates – Practice Q10

0.5
( DEARS )  ( DEARFX )  ( DEARB )
2 2 2

 
  (2 r S , FX x DEARS x DEARFX ) 
portfolio   
 (2 r S , B x DEARS x DEARB )
 
  (2 r FX , B x DEARFX x DEARB ) 

0.5
$300,000 2  $200,000 2  $250,000 2  2(0.1)($ 300,000)($ 200,000) 
 
  2 ( 0.75)($ 300, 000 )($ 250, 000)  2( 0.20)($ 200, 000 )($ 250,000 ) 

 $313,000,000,000  $559,464
0.5
10-72
Aggregating DEAR Estimates – Practice Q10

rtfolio (correlatio ncoefficients  1) 


0.5
$300,000 2  $200,000 2  $250,000 2  2(1.0)($300,000)($200,000) 
 
  2(1.0)($ 300,000)($ 250,000)  2(1.0)($ 200,000)($ 250,000) 

 $562,500,000,000  $750,000
0.5

The DEAR for a portfolio with perfect correlation would be


$750,000.

Therefore, the risk reduction is


$750,000 - $559,464 = $190,536.
10-73

Historic or Back Simulation

 Advantages:
 Simplicity
 Does not require normal distribution of
returns (which is a critical assumption for
Risk Metrics)
 Does not need correlations or standard
deviations of individual asset returns.
10-74
Historic or Back Simulation
 Basic idea: Revalue portfolio based on
actual prices (returns) on the assets that
existed yesterday, the day before, etc.
(usually previous 500 days).
 Then calculate 5% worst-case (25th
lowest value of 500 days) outcomes.
 Only 5% of the outcomes were lower.
10-75
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.
10-76
Weaknesses
 Disadvantage: 500 observations is not
very many from statistical standpoint.
 Increasing number of observations by
going back further in time is not
desirable.
 Could weight recent observations more
heavily and go further back.
10-77
Historic or Back Simulation - Example
10-78
Historic or Back Simulation - Example
10-79
Historic or Back Simulation - Example
10-80
Historic or Back Simulation – Practice Q11
10-81
Historic or Back Simulation – Practice Q11
10-82
Historic or Back Simulation – Practice Q11

Part a. )

Japanese Yen:
¥300,000,000/¥80.13 = $3,743,916.14

Swiss Francs:
Swf10,000,000/Swf0.9540 = $10,482,180

Part B)

Delta measures the change in the dollar value of each FX


position if the foreign currency depreciates by 1 percent against
the dollar.
10-83
Historic or Back Simulation – Practice Q11
c.

Japanese Yen:
1.01 x current exchange rate
= 1.01 x ¥80.13 = ¥80.9313/$

Revalued position in $s
= ¥300,000,000/80.9313
= $3,706,847.66

Delta of $ position to Yen


= $3,706,847.66 – 3,743,916.14
= -37,068.78
10-84
Historic or Back Simulation – Practice Q11
c.
Swiss Francs:
1.01 x current exchange rate = 1.01 x Swf0.9540
= Swf0.96354/$

Revalued position in $s
= Swf10,000,000/0.96354
= $10,378,396

Delta of $ position to Swf


= $10,378,396 - $10,482,180
= -$103,784
10-85
Historic or Back Simulation – Practice Q11

2/4 2/3 2/2 2/1 1/29 1/28


80.13 80.84 80.14 83.05 84.35 84.32
Change -0.71 0.7 -2.91 -1.3 0.03
% -.71/ 0.7/ -2.91/ -1.3/ 0.03/
Change 80.84 80.14 83.05 84.35 84.32
-0.8783 0.8735 -3.5039 -1.5412 0.0356
10-86
Historic or Back Simulation – Practice Q11

Japanese yen
Day Delta %Change Risk
2/4 -37,068 -0.8783
2/3 -37,068 0.8735
2/2 -37,068 -3.5039
2/1 -37,068 -1.5412
1/29 -37,068 0.0356
10-87
Historic or Back Simulation – Practice Q11

Japanese yen
Day Delta %Change Risk
2/4 -37,068 -0.8783 32,557
2/3 -37,068 0.8735 -32,379
2/2 -37,068 -3.5039 129,882
2/1 -37,068 -1.5412 57,129
1/29 -37,068 0.0356 1320
10-88
Historic or Back Simulation – Practice Q11

2/4 2/3 2/2 2/1 1/29 1/28


0.9540 0.9575
0.9533 0.9617 0.9557 0.9523
Change -0.0035 0.0042
-0.0084 0.006 0.0034
% -.0035/ 0.0042/
- 0.006/ 0.0034/
Change 0.9575 0.9533
0.0084/ 0.9557 0.9523
0.9617
-0.3655 0.4406 -0.8735 0.6278 0.3570
10-89
Historic or Back Simulation – Practice Q11

Swiss Franc
Day Delta %Change Risk
2/4 -103,784 -0.3655
2/3 -103,784 0.4406
2/2 -103,784 -0.8735
2/1 -103,784 0.6278
1/29 -103,784 0.3570
10-90
Historic or Back Simulation – Practice Q11

Swiss Franc
Day Delta %Change Risk
2/4 -103,784 -0.3655 37,933
2/3 -103,784 0.4406 -45,727
2/2 -103,784 -0.8735 90,655
2/1 -103,784 0.6278 -65,155
1/29 -103,784 0.3570 -37,050
10-91
Historic or Back Simulation – Practice Q11

Total Risk
Day Yen SF Total
2/4 32,557 37,933 70,550
2/3 -32,379 -45,727 -78,106
2/2 129,882 90,655 220,537
2/1 57,129 -65,155 -8026
1/29 1320 -37,050 -35,730
10-92
Historic or Back Simulation – Practice Q11

Part f)

The appropriate procedure would be to repeat the process


illustrated in part (e) above for all 500 days.

These 500 days would be ranked on the basis of total risk from
the worst-case to the best-case.

The 1st percentile from the absolute worst-case situation would


be day 5 in the ranking.
10-93
Historic or Back Simulation – Practice Q11

Part g)

Management would expect with a confidence level of 99


percent that the total risk on February 5 would be no
worse than the total risk value for the 5th worst day in the
previous 500 days. This value represents the VAR for the
portfolio.
10-94
Historic or Back Simulation – Practice Q11

Part h)

The analysis can be upgraded at the end of the each day.


The values for delta may change for each of the assets in
the analysis. As such, the value for VAR may also
change.
10-95
Monte Carlo Simulation
 To overcome problem of limited number
of observations, synthesize additional
observations.
 Perhaps 10,000 real and synthetic
observations.
 Employ historic covariance matrix and
random number generator to synthesize
observations.
 Objective is to replicate the distribution of
observed outcomes with synthetic data.
10-96
Expected Shortfall
 As mentioned earlier, a criticism of VAR is that it tells
the FI manager the level of possible losses that might
occur with a given confidence level—that is, the 99th
percentile—assuming a normally shaped return
distribution.

 Expected shortfall (ES), also referred to as conditional


VAR and expected tail loss, tells us the average of the
losses in the tail of the distribution beyond the 99th
percentile—that is, if 1 in every 100 days there is a
loss, ES tells us the average of those 1 in 100 day
losses.
10-97
Expected Shortfall - Example
10-98
Expected Shortfall - Example
 Thus, while the VAR is identical for both securities, the
ES finds that security B has the potential to subject the
FI to much greater losses than security A.

 Specifically, if tomorrow is a bad day, VAR finds that


there is a 1 percent probability that the FI’s losses will
exceed $920 million on either security.
10-99
Expected Shortfall - Example
ESa = -920 million

ESb = (0.25 *(-920m) + (0.75*(-1704m))


= -1508 million

 If tomorrow is a bad day, ES finds that there is a 1


percent probability that the FI’s losses will exceed
$920 million if security A is in its trading portfolio, but
losses will exceed $1,508 million if security B is in its
trading portfolio.
 ES scales up the risk factor to account for fat tails in
the probability distribution, using 2.665 for a 1 percent
confidence level (and 2.063 for a 5 percent confidence
 level).
10-100
Expected Shortfall - Example
 FI has 1 million euro in its trading portfolio on the
close of business on a particular day. Suppose also
that looking back at the daily percentage changes in
the Euro/$ exchange rate over the past year, we find
that the volatility, or standard deviation, of daily
percentage changes in the spot exchange rate was
44.3 bp. However, the FI is interested in adverse
moves—bad moves that will not occur more than 1
percent of the time, or 1 day in every 100. Exchange
Rate is $1.2527/euro.
10-101
Expected Shortfall - Example
Dollar value of Position = 1.2527 *1
= $1,252,700

VAR:
FX Volatility = 2.33 * 44.3bp
= 103.219 bp

VAR =1,252,700 * 0.0103219


= $12,930
10-102
Expected Shortfall - Example
Dollar value of Position = 1.2527 *1
= $1,252,700

ES
FX Volatility = 2.665 *44.3bp
= 118.0595 bp

ES = 1,252,700 * 0.01180595
= $14,798
10-103
Expected Shortfall – Practice Q12
10-104
Expected Shortfall – Practice Q12
ESa = -740 million

ESb = (0.40 *(-740m)) + (0.60*(-1393m))


= -1131.8 million
10-105
Expected Shortfall – Practice Q13
10-106
Expected Shortfall – Practice Q13
Dollar value of Position = 1.5625 *5
= $7,812,500

VAR:
FX Volatility = 2.33 * 58.5bp
= 136.305 bp

VAR =7,812,500 * 0.0136305


= $106,488
10-107
Expected Shortfall – Practice Q13
Dollar value of Position = 1.5625 *5
= $7,812,500

ES:
FX Volatility = 2.665 * 58.5bp
= 155.90 bp

VAR =7,812,500 * 0.015590


= $121,796.88
10-108
Expected Shortfall – Practice Q14
10-109
Risk Metric Models - Practice Q14

DEAR = ($ Value of portfolio) x (2.33 x m )


= $250m x (2.33 x .0225)
= $250m x .052425
= $13,106,250

VAR = $13,106,250 x 5
= $13,106,250 x 2.2361
= $29,306,885.63
10-110
Risk Metric Models - Practice Q14

ES = ($ Value of portfolio) x (2.33 x m )


= $250m x (2.665 x .0225)
= $250m x .060
= $15,000,000

ES (5-day) = $15,000,000 x 5
= $15,000,000 x 2.2361
= $33,541,500
10-111
Regulatory Models

 BIS (including Federal Reserve)


approach:
 Market risk may be calculated using
standard BIS model.
 Specific risk charge.
 General market risk charge.
 Offsets.
 Subject to regulatory permission, large
banks may be allowed to use their internal
models as the basis for determining capital
requirements.
10-112
BIS Model
 Specific risk charge:
 Risk weights × absolute dollar values of long and
short positions
 General market risk charge:
 reflect modified durations  expected interest
rate shocks for each maturity
 Vertical offsets:
 Adjust for basis risk
 Horizontal offsets within/between time zones
10-113
Web Resources
For information on the BIS framework, visit:
Bank for International Settlement www.bis.org
Federal Reserve Bank
www.federalreserve.gov
Large Banks: BIS versus 10-114

RiskMetrics
 In calculating DEAR, adverse change in rates
defined as 99th percentile (rather than 95th
under RiskMetrics)
 Minimum holding period is 10 days (means that
RiskMetrics’ daily DEAR multiplied by 10 )*.
 Capital charge will be higher of:
 Previous day’s VAR (or DEAR  10 )
 Average Daily VAR over previous 60 days times a
multiplication factor  3.

*Proposal to change to minimum period of 5 days under


Basel II, end of 2006.
10-115
Pertinent Websites
American Banker www.americanbanker.com
Bank of America www.bankofamerica.com
Bank for International Settlements
www.bis.org
Federal Reserve www.federalreserve.gov
J.P.Morgan/Chase www.jpmorganchase.com
RiskMetrics www.riskmetrics.com

You might also like