Garch-M模型的沪市险值实证研究

Empirical Analysis of VaR in Shanghai Stock Market Based on Garch-M Model

由于我国股票市场每日报酬时间序列的非正态性和厚尾特性,且呈现波动集群性,基于正态假设的静态模型存在很大的缺陷,而Arch类模型具有良好的处理厚尾能力,能较好地描述股价等金融变量的波动特征,因此,将Garch-M模型引入VaR方法中,通过实证分析,结果表明,Garch-M模型能显著提高预测的准确性。

For the non-normal characteristic and heavy-tailed distribution in China＇s stock markets, there are defects in using the model based on normal assumptions. Arch model can deal with heavy-tailed distribution and describe the fluctuations of stock price well. This paper carries out an empirical analysis with the value-at-risk （VaR） method based on Garch-M model, and the results show that the Garch-M model improves the prediction precision.