In a satisfying way, it turns out that there are only so many shapes that probability densities can assume as they head off toward infinity. Extreme value theory makes this notion precise, and gives us some tools to work with them.

See also densities and intensities, survival analysis.

🏗

## Generalized Pareto Distribution

Best intro from Hosking and Wallis (1987):

The generalized Pareto distribution is the distribution of a random variable \(X\) defined by \(X=\) \(\alpha\left(1-e^{-k Y}\right) / k,\) where \(Y\) is a random variable with the standard exponential distribution. The generalized Pareto distribution has distribution function

\[ \begin{aligned} F(x) &=1-(1-k x / \alpha)^{1 / k}, & & k \neq 0 \\ &=1-\exp (-x / \alpha), & & k=0 \end{aligned} \] and density function

\[ \begin{aligned} f(x) &=\alpha^{-1}(1-k x / \alpha)^{1 / k-1}, & & k \neq 0 \\ &=\alpha^{-1} \exp (-x / \alpha), & & k=0 \end{aligned} \] the range of \(x\) is \(0 \leq x<\infty\) for \(k \leq 0\) and \(0 \leq x \leq \alpha / k\) for \(k>0 .\) The parameters of the distribution are \(\alpha,\) the scale parameter, and \(k,\) the shape parameter. The special cases \(k=0\) and \(k=1\) yield, respectively, the exponential distribution with mean \(\alpha\) and the uniform distribution on \([0, \alpha] ;\) Pareto distributions are obtained when \(k<0 .\)

and

- The failure rate \(r(x)=f(x) /\{1-F(x)\}\) is given by \(r(x)=1 /(\alpha-k x)\) and is monotonic in \(x,\) decreasing if \(k<0,\) constant if \(k=0,\) and increasing if \(k>0\)
- If the random variable \(X\) has a generalized Pareto distribution, then the conditional distribution of \(X-t\) given \(X \geq t\) is also generalized Pareto, with the same value of \(k\)
- Let \(Z=\max \left(0, X_{1}, \ldots, X_{N}\right),\) where the \(X_{i}\) are independent and identically distributed as (1) and \(N\) has a Poisson distribution. Then \(Z\) has, essentially, a generalized extreme value (GEV) distribution as defined by Jenkinson (1955); that is, there exist quantities \(\beta, \gamma,\) and \(\delta,\) independent of \(z,\) such that
\[ \begin{aligned} F_{Z}(z) &=\operatorname{Pr}(Z \leq z) \\ &=\exp \left[-\{1-\delta(z-\gamma) / \beta\}^{1 / \delta}\right], \quad z \geq 0 \end{aligned} \] furthermore, \(\delta=k ;\) that is, the shape parameters of the GEV and the GPD are equal.

## Generalized Extreme Value distributions

🏗️

## Burr distribution

🏗️

## References

*The Annals of Probability*2 (5, 5): 792–804. https://doi.org/10.1214/aop/1176996548.

*PLOS ONE*13 (5, 5): e0196456. https://doi.org/10.1371/journal.pone.0196456.

*Journal of the American Statistical Association*92 (440, 440): 1609–20. https://doi.org/10.1080/01621459.1997.10473683.

*Mathematical Geology*21 (8, 8): 829–42. https://doi.org/10.1007/BF00894450.

*Statistical Extremes and Applications*, edited by J. Tiago de Oliveira, 461–82. NATO ASI Series. Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-94-017-3069-3_34.

*Extremal Events in Finance and Insurance*. Springer Berlin Heidelberg.

*Modelling Extremal Events*, 21–57. Applications of Mathematics 33. Springer Berlin Heidelberg. http://link.springer.com/chapter/10.1007/978-3-642-33483-2_2.

*Mathematical Proceedings of the Cambridge Philosophical Society*24 (2, 2): 180–90. https://doi.org/10.1017/S0305004100015681.

*Journal of Risk and Financial Management*11 (1, 1): 10. https://doi.org/10.3390/jrfm11010010.

*Journal of the Royal Statistical Society: Series B (Methodological)*38 (3, 3): 257–64. https://doi.org/10.1111/j.2517-6161.1976.tb01591.x.

*Technometrics*35 (2, 2): 185–91. https://doi.org/10.1080/00401706.1993.10485040.

*Technometrics*29 (3, 3): 339–49. https://doi.org/10.1080/00401706.1987.10488243.

*Communications in Statistics - Theory and Methods*40 (14, 14): 2500–2510. https://doi.org/10.1080/03610920903324874.

*Communications in Statistics - Theory and Methods*48 (8): 2014–38. https://doi.org/10.1080/03610926.2018.1441418.

*The Journal of Operational Risk*1 (2, 2): 51–57. https://doi.org/10.21314/JOP.2006.009.

*Computer Networks*40 (3, 3): 459–74. https://doi.org/10.1016/S1389-1286(02)00306-7.

*ASTIN Bulletin: The Journal of the IAA*27 (1, 1): 117–37. https://doi.org/10.2143/AST.27.1.563210.

*Quantitative Risk Management : Concepts, Techniques and Tools*. Princeton: Princeton Univ. Press.

*The Annals of Statistics*3 (1, 1): 119–31. https://doi.org/10.1214/aos/1176343003.

*Biometrika*72 (1, 1): 67–90. https://doi.org/10.2307/2336336.

*Scandinavian Actuarial Journal*1951 (1-2, 1-2): 158–75. https://doi.org/10.1080/03461238.1951.10432137.

*Institute of Mathematical Statistics Lecture Notes - Monograph Series*, 272–83. Beachwood, Ohio, USA: Institute of Mathematical Statistics. https://doi.org/10.1214/074921706000001102.

*Mathematics*7 (5, 5): 406. https://doi.org/10.3390/math7050406.