Extreme Value Copula Estimation for Subordinated Processes

• Activity Number: 168
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 10:30 AM to 12:20 PM
• Sponsor: IMS
• Author(s): Jan Beran*+
• Companies: University of Konstanz
• Keywords: copula ; time series ; long-range dependence ; Gaussian subordination ; extreme value copula ; Pickands dependence function

Abstract

We consider multivariate time series with cross-sectional distributions characterized by a copula. The existence of multivariate stationary processes with (almost) arbitrary marginal copula distributions and short or long-range dependence and (almost) arbitrary marginals is established. Estimation of the unknown copulas is considered, with special emphasis on extreme value copulas. In the case of long memory, limit theorems are derived for subordinated processes. The asymptotic properties turn out to be very different from the case of iid or short-range dependent observations. Simulations and data examples illustrate the results.