B0504
Title: Financial risk measures in complex networks: The effect of asymptotic independence
Authors: Vicky Fasen-Hartmann - Karlsruhe Institute of Technology (Germany) [presenting]
Bikramjit Das - Singapore University of Technology and Design (Singapore)
Abstract: Risk measures are investigated for a financial network of agents with portfolios of heavy-tailed risk objects. Financial returns are usually empirically observed to be heavy-tailed, and it is well-known that tail risks between two such objects are often asymptotically tail-independent, i.e., extreme values are less likely to occur simultaneously. Surprisingly, asymptotic tail independence in dimensions larger than two has received little attention in the literature; the notion of mutual asymptotic tail independence is first established for general d-dimensions and is compared with the notion of pairwise asymptotic independence commonly used to access bivariate tail dependence. The focus is on two particular dependence structures ideally suited for modelling risk in any general dimension: the well-known Gaussian copula, once popular in financial modelling, and the Marshall-Olkin copula, which is widely used for modelling systemic risk in large systems. Using bipartite graphs to capture risks in a financial network and multivariate regular variation, the effect of asymptotic tail independence (both mutual and pairwise) is studied on the asymptotic properties of popular systemic tail risk measures.
