Modelling losses and locating the tail with the Pareto Positive Stable distribution

This paper focuses on modelling the severity distribution. We directly model the small, moderate and large losses with the Pareto Positive Stable (PPS) distribution and thus it is not necessary to fix a threshold for the tail behaviour. Estimation with the method of moments is straightforward. Properties, graphical tests and expressions for value-at risk and tail value-at-risk are presented. Furthermore, we show that the PPS distribution can be used to construct a statistical test for the Pareto distribution and to determine the threshold for the Pareto shape if required. An application to loss data is presented. We conclude that the PPS distribution can perform better than commonly used distributions when modelling a single loss distribution for moderate and large losses. This approach avoids the pitfalls of cut-off selection and it is very simple to implement for quantitative risk analysis.

[1]  Sidney I. Resnick,et al.  How to make a Hill Plot , 2000 .

[2]  William Feller,et al.  An Introduction to Probability Theory and Its Applications , 1967 .

[3]  B. M. Hill,et al.  A Simple General Approach to Inference About the Tail of a Distribution , 1975 .

[4]  J. Nielsen,et al.  Kernel density estimation for heavy-tailed distributions using the champernowne transformation , 2005 .

[5]  X. Gabaix.  Zipf's Law for Cities: An Explanation , 1999 .

[6]  Steven Vanduffel,et al.  Bounds and Approximations for Sums of Dependent Log-Elliptical Random Variables , 2008 .

[7]  J. Sarabia,et al.  A two-tail version of the PPS distribution with application to current account balance data , 2012 .

[8]  L. F. Velandia,et al.  Medidas de riesgo, características y técnicas de medición: una aplicación del VaR y el ES a la tasa interbancaria de Colombia , 2005 .

[9]  Raluca Vernic,et al.  Skewed bivariate models and nonparametric estimation for the CTE risk measure , 2008 .

[10]  D. Karlis,et al.  Modelling losses using an exponential-inverse Gaussian distribution , 2004 .

[11]  PAUL EMBRECHTS,et al.  Modelling of extremal events in insurance and finance , 1994, Math. Methods Oper. Res..

[12]  José María Sarabia,et al.  The Pareto-positive stable distribution: A new descriptive model for city size data , 2009 .

[13]  On the distribution of the (un)bounded sum of random variables , 2011 .

[14]  Z. Landsman,et al.  Multivariate Tweedie Distributions and Some Related Capital-at-Risk Analysis , 2010 .

[15]  H. Bauke.  Parameter estimation for power-law distributions by maximum likelihood methods , 2007, 0704.1867.

[16]  Mark E. J. Newman,et al.  Power-Law Distributions in Empirical Data , 2007, SIAM Rev..

[17]  H. Akaike.  A new look at the statistical model identification , 1974 .

[18]  N. L. Johnson,et al.  Continuous Univariate Distributions. , 1995 .

[19]  P. V. (Sundar) Balakrishnan,et al.  Power Law and Evolutionary Trends in Stock Markets , 2006 .

[20]  X. Gabaix.  Zipf's Law and the Growth of Cities , 1999 .

[21]  N. Balakrishnan,et al.  A Primer on Statistical Distributions , 2003 .

[22]  A. Azzalini.  The Skew‐normal Distribution and Related Multivariate Families * , 2005 .

[23]  E. Castillo.  Extreme value and related models with applications in engineering and science , 2005 .

[24]  Wu-Chyuan Gau,et al.  Estimation of loss reserves with lognormal development factors , 2008 .

[25]  Montserrat Guillén,et al.  Nonparametric Estimation of Value-at-Risk , 2012 .

[26]  Stuart A. Klugman,et al.  Loss Models: From Data to Decisions , 1998 .

[27]  J. McDonald,et al.  Applications of the GB2 family of distributions in modeling insurance loss processes , 1988 .

[28]  Jens Perch Nielsen,et al.  Kernel density estimation of actuarial loss functions , 2003 .

[29]  Michael Thomas,et al.  Statistical Analysis of Extreme Values , 2008 .