【摘要】 从分析中国股市指数收益率的统计特征入手,以SV模型为基础,在多种分布情形下测算了沪深两市时变风险值V aR及ES。结果表明:基于GED分布的SV模型(SV-GED模型)较好地刻画了高频时间序列的尖峰肥尾性及波动集聚性与持续性等特性,并对两市指数进行较准确的预测,ES相比V aR能够较准确地估计尾部风险。更多还原

【Abstract】 The statistical characteristics of index returns ratios in the Chinese stock market are analyzed and the VaR and ES of Shanghai and Shenzhen Stock Exchange.Market based on SV model are measured under different distributions.It shows that SV model based on GED can give better estimation to the index of two market when fat-tailed densities,volatility clusting and volatility persistence are taken into account in the conditional variance.In addition,ES can give better estimation to tail risk than VaR.更多还原