Historical Simulation and

Extreme Value Theory

Week 4 Lecture
Presenter: Dr. David Liu

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 1
 Historical Simulation

⚫ Collect data on the daily movements in all

market variables (known as risk factors)
⚫ The first simulation trial assumes that the
percentage changes in all risk factors are as on
the first day
⚫ The second simulation trial assumes that the
percentage changes in all risk factors are as on
the second day
⚫ and so on

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 2
 Percentage Changes vs. Actual
Changes
⚫ For variables such as interest rates, credit
spreads, and (perhaps) volatilities we
typically match actual changes rather than
percentage changes

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 3
 Historical Simulation continued
⚫ Suppose we use n days of historical data with
today being day n
⚫ Let vi be the value of a variable on day i

⚫ The ith trial assumes that the value of the

market variable tomorrow (i.e., on day n+1) is
vi
vn if percentage changes are matched
vi −1
vn + vi − vi −1 if actual changes are matched

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 4
 Example: Portfolio on Sept 25,
2008 (Table 13.1, page 295)

Index Amount Invested ($000s)

DJIA 4,000
FTSE 100 3,000
CAC 40 1,000
Nikkei 225 2,000
Total 10,000

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 5
 U.S. Dollar Equivalent of Stock
Indices (Table 13.2, page 295)
Day Date DJIA FTSE CAC 40 Nikkei
0 Aug 7, 2006 11,219.38 11,131.84 6,373.89 131.77
1 Aug 8, 2006 11,173.59 11,096.28 6,378.16 134.38
2 Aug 9, 2006 11,076.18 11,185.35 6,474.04 135.94
3 Aug 10, 2006 11,124.37 11,016.71 6,357.49 135.44
…. …… …… ….. …… ……
499 Sep 24, 2008 10,825.17 9,438.58 6,033.93 114.26
500 Sep 25, 2008 11,022.06 9,599.90 6,200.40 112.82

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 6
 Scenarios (Table 13.3, page 296)
11,173.59
= 11,022.06 
11,219.38

Scenario DJIA FTSE CAC Nikkei Portfolio Loss

Number Value
1 10,977.08 9,569.23 6,204.55 115.05 10,014.334 -14.334
2 10,925.97 9,676.96 6,293.60 114.13 10,027.481 -27,481
3 11,070.01 9,455.16 6,088.77 112.40 9,946.736 53,264
…. …… ….. ….. ….. ….. …..
499 10,831.43 9,383.49 6,051.94 113.85 9,857.465 142.535
500 11,222.53 9,763.97 6,371.45 111.40 10,126.439 -126.439

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 7
 Losses (Table 13.4, page 297)
Scenario Number Loss ($000s)
494 477.841
We estimate
339 345.435 VaR as the fifth
349 282.204 worst loss and
329 277.041 ES as the
average of
487 253.385 losses less
227 217.974 than VaR.
131 205.256

One-day 99% VaR=$253,385

One-day 99% ES is
(477,841+345,435+282,204+277,041)/4=$345,630
Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 8
 Stressed VaR and Stressed ES
⚫ Instead of basing calculations on the
movements in market variables over the
last n days, we can base calculations on
movements during a period in the past that
would have been particularly bad for the
current portfolio
⚫ This produces measures known as
“stressed VaR” and “stressed ES”
Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 9
 Accuracy of VaR (page 299-301)

Suppose that x is the qth percentile of the loss

distribution when it is estimated from n observations.
The standard error of x is
1 q (1 − q )
f ( x) n
where f(x) is an estimate of the probability density
function of the loss at the qth percentile calculated by
assuming a probability distribution for the loss.

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 10
 Example 13.1 (page 300)
⚫ We estimate the 1-percentile from 500 observations as
$25 million
⚫ We estimate f(x) by approximating the actual empirical
distribution with a normal distribution mean zero and
standard deviation $10 million
⚫ The 1-percentile of the approximating distribution is
NORMINV(0.99,0,10) = 23.26 and the value of f(x) is
NORMDIST(23.26,0,10,FALSE)=0.0027
⚫ The estimate of the standard error is therefore
1 0.01 0.99
 = 1.67
0.0027 500

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 11
 Weighting of Observations
⚫ Let weights assigned to observations
decline exponentially as we go back in
time. The weight given to Scenario i is
𝜆𝑛−𝑖 1−𝜆
(n is the number of scenarios)
1−𝜆𝑛
⚫ Rank observations from worst to best
⚫ Starting at worst observation sum weights
until the required percentile is reached
(e.g. 99% VaR, the sum >=0.01)
Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 12
 Application to 4-Index Portfolio
l=0.995 (Table 13.5, page 302)
Scenario Loss Weight Cumulative
Number ($000s) Weight
494 477.841 0.00528 0.00528
339 345.435 0.00243 0.00771
349 282.204 0.00255 0.01027
329 277.041 0.00231 0.01258
487 253.385 0.00510 0.01768
227 217.974 0.00139 0.01906

One-day 99% VaR=$282,204

One day 99% ES is

[0.00528×477,841+0.00243×345,435+(0.01−0.00528 − 0.00243)×282,204]/0.01 =
$400,914

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 13
 Volatility Scaling
⚫ Use a volatility updating scheme to monitor
volatilities of all market variables (EWMA)
⚫ If the current volatility for a market variable is b
times the volatility on Day i, multiply the actual
change observed on day i by b
⚫ Value of market variable under ith scenario
becomes
vi −1 + (vi − vi −1 ) n +1 / i b
vn
vi −1

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 14
 Volatilities (% per Day) Estimated for
Next Day in 4-Index Example (Table 13.6, page
304)

Day Date DJIA FTSE CAC 40 Nikkei

0 Aug 7, 2006 1.11 1.42 1.40 1.38
1 Aug 8, 2006 1.08 1.38 1.36 1.43
2 Aug 9, 2006 1.07 1.35 1.36 1.41
3 Aug 10, 2006 1.04 1.36 1.39 1.37
…. …… …… …… …… ……
499 Sep 24, 2008 2.21 3.28 3.11 1.61
500 Sep 25, 2008 2.19 3.21 3.09 1.59

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 15
 Volatility Adjusted Losses (Table 13.7,
page 304)

Scenario Number Loss ($000s)

131 1,082.969
494 715.512
227 687.720
98 661.221
329 602.968
339 546.540
74 492.764

One-day 99% VaR = $602,968

One-day 99% ES = (1,082,969+715,512+687,720+661,221)/4 =

$786,855

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 16
 Volatility Scaling for the Portfolio
⚫ Monitor variance of simulated losses on the
portfolio using EWMA
⚫ If current standard deviation of losses is bP times
the standard deviation of simulated losses on
Day i, multiply ith loss given by the standard
approach by bP
⚫ This approach gives VaR and ES as $627,916
and $777,545, respectively

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 17
The Bootstrap Method to
Determine Confidence Intervals
⚫ Suppose there are 500 daily changes
⚫ Calculate a 95% confidence interval for
VaR by sampling 500,000 times with
replacement from daily changes to obtain
1000 sets of changes over 500 days
⚫ Calcuate VaR for each set and calculate a
confidence interval
⚫ This is known as the bootstrap method
Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 18
 Computational Issues
⚫ To avoid revaluing a complete portfolio many
times a delta/gamma approximation is
sometimes used
⚫ When a derivative depend on only one
underlying variable, S
1
P  S +  (S ) 2
2
⚫ With many variables
n n n
1
P   i Si +   ij Si S j
i =1 i =1 j =1 2
Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 19
 Extreme Value Theory (EVT) (page
307-313)

⚫ Extreme value theory can be used to investigate the

properties of the right tail of the empirical distribution of
a variable.
⚫ We first choose a level u somewhat in the right tail of the
distribution
⚫ We then use Gnedenko’s result which shows that for a
wide class of distributions as u increases the probability
distribution that v lies between u and u+y conditional that
it is greater than u tends to a generalized Pareto
distribution

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 20
 Generalized Pareto Distribution

⚫ This has two parameters x (the shape

parameter) and b (the scale parameter)
⚫ The generalized Pareto cumulative distribution
G(y) is
−1 / x
 x 
1 − 1 + y 
 b 

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 21
 Maximum Likelihood Estimator
(Equation 13.7, page 309)

⚫ The observations are sorted in

descending order. Suppose that
there are nu observations vi greater
than u
⚫ We choose x and b to maximize
− 1 / x −1
nu 1  x(v i − u )  
 ln  b  1 + b
 
i =1    

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 22
 Equivalence to the Power Law
Our estimator for the cumulative probability that the
variable v is greater than x is
−1 / x
nu  x −u
1 + x 
n  b 

Setting u = b x we see that this correspond s to the power law

-
Prob( v  x ) = Kx

where
−1 / x
nu  x  1
K = 
    =
n b x

Extreme value theory therefore explains why the power law

holds so widely

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 23
 Estimating VaR and ES Using
Extreme Value Theory (page 310)
⚫ Setting the probability that v > x equal to
1−q we obtain an estimate of VaR with a
confidence level of q

b  n 
−x

VaR = u +  (1 − q) − 1
x  nu  

⚫ Also
VaR + b − xu
ES =
1− x

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 24
 Using Maximum Likelihood for 4-
Index Example, u=160 (Table 13.10, page
311)

 1  x(v − u )  −1/ x−1 

Scenario Loss ($000s) Rank ln  1 + i
 
 b  b  

494 477.841 1 -8.97

339 345.435 2 -7.47
349 282.204 3 -6.51
329 277.041 4 -6.42
487 253.385 5 -5.99

304 160.778 22 -3.71

Total -108.37

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018
25
Choice of u
⚫ A thumb of rule is that u should be
approximately equal to the 95th
percentile of the empirical distribution.
⚫ In the search for the optimal values of
ᵦ and ξ, both variables should be
constrained to be positive.

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 26
 End of Week 4 Lecture PPT

Risk Management and Financial Institutions 5e, Chapter 13, Copyright © John C. Hull 2018 27