论文部分内容阅读
本文将Realized GARCH模型推广至基于周历日的时变参数情形以刻画杠杆和溢出效应的周内特征并避免传统GARCH类模型在拟合长记忆性与周内效应时两者相互干扰问题.将新模型应用于上海股票市场2001至2013数据的分析发现:我国股市波动率存在时变的杠杆效应和溢出效应.实证结果表明:新模型无论在样本外的预测能力还是在样本内的拟合度上都明显优于现有模型.
In this paper, we generalize the Realized GARCH model to the case of time-varying parameters based on the days of the week to describe the characteristics of the week of leverage and spillover and avoid the mutual interference between the traditional GARCH model and the long-term memory effect. The analysis of the data from 2001 to 2013 in Shanghai stock market shows that the volatility of China’s stock market has time-varying leverage effect and spillover effect.The empirical results show that the new model’s predictive ability outside the sample or within the sample On both are significantly better than the existing model.