摘要
采用了新信息过程服从广义误差分布(Generalized Error D istribution)的ARMA-EGARCH-M模型对中国沪深两市的波动情况进行了分析研究,以捕捉沪深两市股指收益率分布的尖峰厚尾性、波动集群性以及非对称性,实证研究表明,相对新信息过程服从正态分布的GARCH类模型,广义误差分布的假定能够更好得捕捉中国股票市场波动特征,并在实证结果的基础上分析了造成这些波动特点的若干原因.
This article analyses the volatility of China' s stock market using ARMA -EGARCH -M model based on GED( Generalized Error Distribution) assumption to get the characteristics of leptokurtosis, fat- tail, volatility clustering and asymmetric, the empirical result shows that, compared with the GARCH models based on ND (Normal Distribution), the model based on GED assumption performs better on getting the characteristics of the vol- atility of China' s stock market, and some profound reasons for the volatility on empirical results are given.
出处
《南昌大学学报(工科版)》
CAS
2005年第4期67-70,84,共5页
Journal of Nanchang University(Engineering & Technology)
基金
国家自然科学基金资助项目(70371042)