On background driving distribution functions (BDDF) for some selfdecomposable variables

Abstract

Many classical variables (statistics) are selfdecomposable. They admit the random integral representations via Lévy processes. In this note are given formulas for their background driving distribution functions (BDDF). This may be used for a simulation of those variables. Among the examples discussed are: gamma variables, hyperbolic characteristic functions, Student t-distributions, stochastic area under planar Brownian motions, inverse Gaussian variable, logistic distributions, non-central chi-square, Bessel densities and Fisher z-distributions. Found representations might be of use in statistical applications.