高频视角下考虑收益非对称性结构的VaR和ES风险测度

Rsik Management of Chinese Stock Market with Asymmetric Structure of Return in the Perspective of High-frequency Data

考虑中国股市指数收益率分布和波动的非对称性结构,采用偏t分布拟合收益率的有偏分布形态,利用RS-捕捉波动率的杠杆效应,并构建ARFIMA-GARCH和SKST-RS-模型分别预测RS-和刻画收益率波动的动态结构,进而改进VaR和ES并测度卖空限制市场的下侧风险。通过Kupiec LR和动态分位数检验,实证分析了ES和VaR的风险管理效果。结果表明:基于日内高频收益的SKST-RS-模型的VaR预测能力强于SKST-RV模型和基于日间收益率的GARCH类模型;在VaR估计市场极端风险失效时,ES能够有效地对尾部极端风险进行管理。

Considering the asymmetric structure of index return^s distribution and volatility in Chinese stock market, we use skew-t distribution to fit index return, apply RS- to capture leverage effect of volatility, construct ARFIMA-GARCH model and SKST-RS- model to forecast RS-, to depict the dynamics of volatility respectively, then calculate VaR and ES to measure down-side risk in the market with the feature of short-sale constraint. Furthermore, we analyze the effect of VaR and ES by Kupiec LR test and Dynamic quantile regression. The results show that SKST-RS- model based on high- frequency data performs batter in forecasting VaR than SKST-RV and GARCH type models based on low-frequency data. When VaR fails to estimate extreme risk in market, ES could manage extreme risk of tail efficiently.