Measurement of Dynamic Portfolio VaR Based on Mixed Vine Copula Model

Abstract

The measurement of portfolio VaR has been a hot issue in the field of the academic and the industry. This paper applies three kinds of Vine Copula model to describe high-dimensional dependency structure between multiple assets, introduces mixed binary copula function to improve the accuracy of tail dependence structure. We use six important stock markets as stock portfolio to test this model. The empirical results show that introducing mixed Copula function can improve the measurement reliability of Vine Copula model, and the reliability of mixed R-Vine model is highest in three kinds of mixed Vine Copula models.