Conditional Dependence Modelling with Regular Vine Copulas

Abstract

Modelling sophisticated high-dimensional dependence structures for financial assets in a portfolio framework require flexible dependence models. In this paper, a regular vine-copula based model is employed to analyze financial dependencies and co-movements of a six-dimensional portfolio of currency exchange rates starting from January 2001 to April 2018. The regular-vine copula based model employs partial correlations to construct the regular vine structure and offer superior flexibility in the selection of the distributions to model financial dependence structure. The model also captures the asymmetry between multivariate variables using bivariate copulas with flexible tail dependence. Empirical evidence suggests that co-movements in currency markets are most likely to experience a crash and boom together thus, concluding that currency markets are integrated due to the nature of the global financial systems. The C-Vine copula specification is favoured over the other