基于稳定分布条件的GARCH模型研究

Study of GARCH Model with Stable Distribution

本文应用GARCH模型对1995～2008年的沪深A股指数收益率序列进行分析,对正态分布、学生t分布、广义误差分布以及稳定分布下的GARCH模型进行对比研究,发现基于极大似然准则和AIC信息准则下,新信息服从稳定分布的GARCH模型优于其他模型。

This paper use GARCH model to fit shanghai and shenzhen stock market index returns（ 1995 - 2008）, compares the results of GARCH model with normal distribution ,student-t distribution, GED distribution and stable paretian distribution. We find that a GARCH model with stable paretian innovation fits returns clearly better the other models based on the MLE and AIC criteria.