【摘要】 采用混频数据抽样模型(MIDAS)研究了混频投资者情绪对中国股市收益率及其波动的影响.通过构建日度、周度及月度这三种不同频率的投资者情绪,实证结果发现,混频情绪对当期收益率及其波动都存在显著的正向影响,并且与传统回归模型相比,MIDAS模型具有更强的解释能力.本文进一步使用GARCH-MIDAS模型研究了混频情绪对收益率波动长期成分的影响,发现混频情绪能够显著影响收益的长期波动.更多还原

【Abstract】 This paper studies investor sentiment with three different frequencies: daily,weekly and monthly,and applies MIDAS model to study the effects of mixed-frequency investor sentiment on stock market return and volatility. It is found that the mixed-frequency sentiment has a significant positive impact on both return and volatility and that the MIDAS model outperforms the traditional regression model. The GARCH-MIDAS model is used to study the effect of mixed-frequency sentiment on the stock volatility,and it is found that the mixed-frequency sentiment significantly affects the long-term volatility.更多还原