Exchangeability of copulas

Abstract

When studying the dependence structure of multivariate random vectors, the copula is crucial. A copula is called exchangeable, if it is invariant under any permutation of its arguments. In this thesis, a limit for the absolute distance between a copula and a version of itself with permuted arguments is derived. Furthermore, nonparametric test procedures for testing the hypothesis of exchangeability of a given sample of multivariate data are discussed as well as their asymptotic behavior.