Extreme Value Theory

EVT
Extreme value theory can be used to investigate the properties of the
right tail of the empirical distribution of a variable x. (If we interested
in the left tail we consider the variable x.)
We first choose a level u somewhat in the right tail of the distribution
We then use Gnedenkos 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

EVT
Generalized Pareto Distribution
This has two parameters x (the shape parameter) and b (the scale
parameter)
The cumulative distribution is

x
1 1
b

1 / x

EVT
Estimating the two parameters x (the shape parameter) and b (the
scale parameter) using Maximum Log Likelihood Estimate
The observations, xi, are sorted in descending order. Suppose that
there are nu observations greater than u
We choose x and b to maximize
1 / x 1
1

x( v i u )

ln
1

b
b
i 1

nu

EVT & Power Law

Our estimator for the cumulative

probabilit y that the

variable v is greater th an x is
nu
x u
1 x

n
b

Setting u

1 / x

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

Prob( v x ) Kx -
where
nu
K
n

x
b

1 / x

Extreme value theory th erefore explains

holds so widely

why the power law

VAR and Expected Shortfall using EVT

The estimate of VaR when the confidence level
is q is obtained by solving
nu
q 1
n

VaR u

1 x
b

1 / x

It is
x

b n

VaR u
(1 q ) 1
x nu