Aâ–³Covar Measurement Method Based On Time-Varying Parametric Copula

Since19th century, the world has happened many financial crises. These crises showed an obvious feature:a country’s financial market slump resulted in the neighboring countries and even the global financial market turmoil. Due to the huge damage of sub-prime crisis, extreme risk spillover between financial institutions or markets has attracted great research interests. Because△CoVaR reflects one financial random variable s VaR conditional on the situation when another random variable incurs a VaR loss, it is widely accepted as measure of extreme risk spillover. This paper proposes a△CoVaR measurement method based on time-varying parametric Copula, which applies time-varying copula to describe the dependence between financial variables and uses the AR-GARCH(1,1)-t model to describe the marginal distribution. In this paper, we select the index monthly returns of China mainland, US and HK stock markets from2000to2011June as the original data, Empirical results on China mainland, US and HK stock markets clearly show that the method can sensitively and accurately estimate extreme risk spillover in financial crisis period, due to that it can consider time-varying volatility and time-vary ing dependence at the same time. The stock market of US has a stronger risk spillover than the stock market of China mainland. Therefore, we should improve the resistance of our financial institutions, the internationalization of financial environment, and strengthen the supervision and management of financial risks.