Note on new prospects on vines

Pierre-André Maugis (1) , Dominique Guegan (1, 2)

1 CES - Centre d'économie de la Sorbonne
2 PSE - Paris School of Economics

In this paper, we present a new methodology based on vine copulas to estimate multivariate distributions in high dimensions, taking advantage of the diversity of vine copulas. Considering the huge number of vine copulas in dimension n, we introduce an efficient selection algorithm to build and select vine copulas with respect to any test T. Our methodology offers a great flexibility to practitioners to compute VaR associated to a portfolio in high dimension.

Mots clés: Vines; multivariate copulas; model selection

Domaines: Economies et finances; Gestion et management; Méthodes et statistiques; Probabilités [math.PR]; Statistiques [math.ST]; Théorie [stat.TH]

Pierre-André Maugis, Dominique Guegan. Note on new prospects on vines. Insurance Markets and Companies : Analyses and Actuarial Computations, 2010, 1 (1), pp.15-22. ⟨halshs-00471362⟩