Il giorno martedì 13 maggio alle ore 15.00 presso la aula seminari del Dipartimento di Statistica e Metodi Quantitativi della Università di Milano Bicocca, al IV piano dell'edificio U7, il dott. Ruodu Wang della Università di Waterloo (Canada) terrà un seminario su:
Admissible distributions of risk aggregation
Abstract
Modeling inter-dependence between individual risks often faces statistical as well as probabilistic challenges, in which dependence uncertainty naturally arises. In this talk, we consider problems of risk aggregation with dependence uncertainty. The core question of interest is to characterize admissible distributions of risk aggregation, where marginal distributions of individual risks are assumed known but the joint dependence structure is unknown. In the first part of the talk, we will summarize finite dimensional results based on the notion of joint mixability, and discuss their relevance to optimization problems in risk management. In the second part of the talk, we will study asymptotic problems of admissible risk aggregation with identical margins. The results are concise and, in the meanwhile, surprising. We hope to deliver the following message: with the marginal distribution known and dependence structure unknown, we know essentially nothing about the asymptotic shape of the sum of random variables.
Tutti gli interessati sono invitati a partecipare.
Fabio Bellini