Jump Tail Dependence in Lévy Copula Models.pdf

Extremes DOI 10.1007/s10687-012-0162-1 Jump tail dependence in Lévy copula models Oliver Grothe Received: 28 June 2011 / Revised: 2 May 2012 / Accepted: 25 September 2012 c Springer Science+Business Media New York 2012 Abstract This paper investigates the dependence of extreme jumps in multivariate Lévy processes. We introduce a measure called jump tail dependence , defined as the probability of observing a large jump in one component of a process given a con- current large jump in another component. We show that this measure is determined by the Lévy copula alone and that it is independent of marginal Lévy processes. We derive a consistent nonparametric estimator for jump tail dependence and establish its asymptotic distribution. Regarding the economic relevance of the measure, a simula- tion study illustrates that jump tail dependence has a substantial impact on financial portfolio distributions and optimal portfolio weights. Keywords Multivariate Lévy processes · Dependence of jumps · Nonparametric estimation · Strong consistency · High frequency financial data · Portfolios AMS 2000 Subject Classifications 60G51 · 62G32 · 91G70 1 Introduction This paper proposes a measure of the dependence of extreme changes in multivariate Lévy processes. In contrast to the existing literature, the proposed measure is constant with respect to different return frequencies and therefore allows for a clear interpre- tation. The measure is based on the theory of Lévy copulas and is independent of the marginal distributions of the processes. O. Grothe (B) Department of Economic and Social Statistics, University of Cologne, Albertus-Magnus-Platz, 50923 Köln, Germany e-mail: grothe@statistik.uni-koeln.de O. Grothe In the literature, the tendency of returns of time series to show common extreme realizations is measured b