Hedge funds portfolio optimisation using a vine copula-GARCH-EVT-CVaR model

Abstract

This paper investigates the conditional value-at-risk (CVaR) hedge funds portfolio optimisation approach using a univariate GARCH type model, extreme value theory (EVT) and the vine copula to determine the optimal allocation for hedge funds portfolio. First, we apply the generalised pareto distribution (GPD) to model the tails of the innovation of each hedge funds strategy return. Second, we capture the interdependence structure between hedge funds strategies and construct vine copula-GARCH-EVT model. Then, we combine it with Monte Carlo simulation and mean-CVaR model to optimise hedge funds portfolio, in order to estimate the risk more accurately. The empirical results of five Hedge funds indexes show that the C-vine copula can better characterise the interdependence structure between the different hedge funds strategies and the performance of C-vine copula-GARCH-EVT-CVaR model is better that of multivariate copulas-GARCH-EVT-CVaR models in portfolio optimisation.