Asymmetric Effect Of Investor Sentiment On Stock Market Returns In China

Based on the historical data of China's A-share market from 2004 to 2016,we do empirical study on the asymmetric effect of investor sentiment on the stock market in different market conditions.According to China's stock market characteristics,China's stock market can be divided into three stages: rapid rise stage,wandering stage,rapid decline stage.In the different stages of the stock market,investor's sentiment and trading behavior show different characteristics.In this paper,a proxy variable sample pool is established,and the proxy variables are filtered by the correlation between the proxy variables and the stock index and the correlation between the proxy variables.Then,the principal component analysis is used to reduce the dimension and extract information of the selected proxy variables.By using MSIH(3)-VAR(2)model,we study the fluctuation response of investor sentimental fluctuation to CSI 300 index return in three regimes.We use Eviews to carry on the preliminary data processing and determine the auto regressive lag order.The fitting of the main model and the Impulse response graph are realized by OxMetrics and GiveWin.In this research,we innovatively use the MS(3)-VAR model which can avoid the subjectivity in dividing the regimes to study the influence of investor sentiment on the stock market.Compare to the two-regime model,our model is more in line with the characteristics of China's stock market.The empirical study results show that the impact of investor sentiment on the stock return in different regimes has asymmetric effect.In the rapid rise stage,wandering stage and rapid decline stage,the impact of sentimental fluctuation to stock return increases successively.And in the rapid rise stage and wandering stage,the medium-term reversal effect is small,while in the rapid rise stage,the reverse correction effect is very strong.