A Study On Risk Management Of International Portfolio

This paper combines the Realized GARCH model in the realized volatility method,the POT model in the extreme theory,and the Vine Copula model for high-dimensional correlation modeling to build a Realized GARCH-EVT-Vine Copula model.The model is using to conduct research on the risk management of international portfolio.In this paper,the Value-at-Risk(VaR)of the international asset portfolio is taken as a risk metric,and the accuracy of VaR prediction generated from Realized GARCH-EVT-Vine Copula model is tested by statistical methods,and compared with other common models.In the empirical part,this paper selected the Chinese Shanghai-Shenzhen 300 Index(HS300),the US Standard&Poor’s 500 Index(SP500),the Hong Kong Hang Seng Index(HSI),and the Japan Economic Average Index(N225)for building an equal-weighted portfolio.The time period of the sample data is from January 5,2009 to April 6,2011,with a total of 490 joint trading days.The conclusions obtained in this paper are:(1)The skew-t distribution is superior to the normal distribution and t distribution in characterizing the sharpness,fat tail and the skewness of financial returns.(2)The four stock indexes show obvious regional correlation.(3)Models that do not use the extreme value theory have significantly lower performance in VaR prediction than models using extreme value theory.(4)The Realized GARCH model outperforms the traditional GJR GARCH model in VaR forecasting.(5)The Realized GARCH-EVT-C Vine Copula model has the best performance in VaR forecasting among all models.