A1140
Title: Construction, estimation and application of diffusion processes for extreme values
Authors: Consuelo Nava - University of Aosta Valley (Italy) [presenting]
Ramses Mena - Universidad Nacional Autonoma De Mexico (Mexico)
Abstract: A simple yet powerful method is proposed to construct strictly stationary Markovian models with given but arbitrary invariant distributions. The idea is based on a Poisson transform modulating the dependence structure in the model. An appealing feature of the approach is the ability to fully control the underlying transition probabilities and, therefore, incorporate them within standard estimation methods. Given the proposed representation of the transition density, a Gibbs sampler algorithm based on the slice method is proposed and implemented. The construction results from a Bayesian perspective. In the discrete-time case, special attention is placed on the class of generalized inverse Gaussian (GIG) distributions. The GIG class is very flexible and allows one to obtain various explicit results: it is an interesting choice, especially for econometric or financial applications involving extreme values.
