Estimation of multivariate generalized gamma convolutions through Laguerre expansions.

The generalized gamma convolutions class of distributions appeared in Thorin’s work while looking for the infinite divisibility log-Normal and Pareto distributions. Although these have been extensively studied univariate case, multivariate case dependence structures that can arise from it received little interest literature. Furthermore, only one projection procedure was recently constructed, no estimation procedures are available. By expanding densities into a tensorized Laguerre basis, we bridge gap provide performant both cases. We some insights about performance procedures, convergent series density convolutions, which is shown to be more stable than Moschopoulos’s Mathai’s series. furthermore discuss examples.