基于极值理论的沪深股市VaR和CVaR分析

将VaR和CVaR结合起来能全面描述金融时间序列与尾部相关的风险。考虑沪深股指收益序列胖尾特性,极值理论方法能够对沪深股市VaR和CVaR进行较好估计,运用基于Boot strap和极大似然估计方法解决极值理论数据不足的缺陷,从而给出对VaR和CVaR的点估计和区间估计。

Analyses of VaR and CVaR on Shanghai and Shenzhen Stock Markets by the Use of Extreme Value Theory

Uniting VaR and CVaR can fully describe tailrelated risk. Since Shanghai and Shenzhen A Stock Market return distribution exhibit fat tails, the writers of this paper make use of extreme value theory to analyze VaR and CvaR on the two markets. Bootstrap method is used to reproduce subsamples to offset insufficient data. Bootstrap and likelihoodbased methods are also used to give point estimation and construct confidence intervals for the VaR and CVaR.