Hedge Funds: Strategy, Risk and Performance Evaluation

Abstract

The study presents an application of multivariate regime switching copula models, in order to model the joint distributions of selected hedge fund classes/strategies and traditional capital markets. Our empirical interest focuses on testing for the presence of any asymmetric dependence structures between these sample hedge fund classes and benchmark capital market returns. A number of different specifications of canonical vine copulas are incorporated into a Markov switching two-regime framework to capture hedge fund reactions in bear and bull market phases. We apply this empirical framework on core and complementary hedge fund strategies as well as on traditional asset classes and produce substantial evidence of asymmetric dependence structures between hedge fund strategies and conventional capital markets. Furthermore, our empirical results reveal a high dependence regime with high estimated correlations and a low dependence regime with lower estimated correlations.