Extreme Value Theory And Financial Market Risk Measurement:Empirical Evidence Of SSEC And S&P500

摘要：

In financial risk management, a lot of investors focus on the extreme event with lower probability and larger loss. This paper use ARMA( 1, 1)-GJR(1, 1) to model conditional loss and volatility of SSEC and S&P500, and apply extreme value theory(EVT) to model the extreme tail of standard residual series, and then estimate quantile and calculate corresponding Value-at-Risk( VaR), at last, we test the accuracy of VaR model used by Back-testing method of Kupiec(1995) and ChristoffersenC 1998). Our results show that the conditional losses series exhibits skewed and fattailed distribution; the conditional volatility exhibits significant leverage effect for SSEC and S & P500; the GPD model fits excess losses well. VaR models used in this paper measure dynamic risk in SSEC and S & P500 stock markets at 95％, 99％ levels accurately, the VaR model do not significant prefer to either market SSEC or S & P500.

关键词： financial market Value-at-Risk extreme value theory dynamic risk measuring Back-Testing method

作者: LIN Yu

作者单位: Business school. Chengdu University of Technology, Chengdu, P. R. China, 610059

会议类型: 国际会议

会议名称: 2010 International Conference on Management(2010管理国际大会)

会议地点: 上海

会议语种:英文

页码: 229-237

在线出版日期: 2010-07-24（万方平台首次上网日期，不代表论文的发表时间）