/

Enterprise Risk Management Formula Book

Appendix A.2: Probability Distributions: Continuous (univariate) distributions (c) generalised Pareto, lognormal, Student’s t

[this page | pdf]

Distribution name

Generalised Pareto distribution (GPD)

Common notation

Parameters

 = shape parameter

 = location parameter

 = scale parameter ()

Domain

Probability density function

where

Cumulative distribution function

Mean

Variance

Skewness

(Excess) kurtosis

Other comments

GPD is used in the peaks over thresholds variant of extreme value theory

 

 

Distribution name

Lognormal distribution

Common notation

Parameters

 = scale parameter ()

 = location parameter

Domain

Probability density function

Cumulative distribution function

Mean

Variance

Skewness

(Excess) kurtosis

Characteristic function

No simple expression that is not divergent

Other comments

The median of a lognormal distribution is  and its mode is .

 

The truncated moments of   are:

 

 

Distribution name

(Standard) Student’s t distribution

Common notation

Parameters

 = degrees of freedom (, usually  is an integer although in some situations a non-integral  can arise)

Domain

Probability density function

Cumulative distribution function

where

Mean

Variance

Skewness

(Excess) kurtosis

Characteristic function

where  is a Bessel function

Other comments

The Student’s t distribution (more simply the t distribution) arises when estimating the mean of a normally distributed population when sample sizes are small and the population standard deviation is unknown.

 

It is a special case of the generalised hyperbolic distribution.

 

Its non-central moments if  is even and  are:

 

If  is even and  then , if  is odd and  then  and if  is odd and   then  is undefined.

 


NAVIGATION LINKS
Contents | Prev | Next


Desktop view | Switch to Mobile