ABSTRACT
Much work has been done by researchers and practitioners to explore non-normal properties of multiple asset returns. Among these properties, tail dependence measures the dependence structure of random variables in an extreme environment. With a volatile financial environment, tail dependence becomes more important for asset allocation and risk management. For instance, to increase risk-adjusted returns, risk managers typically pursue diversification strategies by exploiting positive and negative correlations of various asset returns. However, this diversification strategy will become less effective or even fail in the extreme environment and result in significant portfolio losses. Hence, one task faced by researchers and practitioners is to precisely quantify tail dependence of considered assets and find a proper model that captures different tail dependence patterns well in a multivariate setting.
