An Empirical Study Of Investor Sentiment And The Market Returns’ Asymmetric Volatility

Behavioral finance theory holds that individual investors are irrational in investment decisions because of the influence of cognitive bias and emotional influence. This paper focuses on the research of investor sentiment and t he volatility of market returns. The research status of the related problems at domestic and abroad is reviewed systematically and the basic theory of investor sentiment and market return volatility is carefully combed. Monthly market data from January 2003 to December 2014 are selected by means of principal component analysis method and partial least square method to construct the composite sentiment index and through the HP filter to separate the emotional sequence of irrational emotional components. We then put forward the research hypothesis and in-depth analyze the impact and impact mechanism.of investor sentiment factors on market returns asymmetry volatility.Following overall research ideas of from simple to complex and from linear to nonlinear analysis, we gradually relax the assumptions of the model and in-depth study the effects of investor sentiment on return volatility focusing the angle of sentiment threshold effect and non-linear effect.Through the EGARCH model to verify the market return volatility exist asymmetric effect and through the threshold regression model to analyze the relationship difference of investor sentiment on the return volatility under different sentiment thresholds. Using STR model as an empirical research tool and from the perspective of nonlinear perspective, the impact mechanism of investor sentiment factors on market return volatility is deeply discussed, The empirical result shows that the positive asymmetric effect exist s in Chinese stock market volatility. The impact effect of good news(positive impact) on the return volatility is significantly stronger than that of bad news(negative impacts to) and irrational sentiment has a significant amplification effect on the asymmetry of market return volatility. On this basis, from the perspective of the average correlation and the average variance,we explore the impact conduction path.of investor sentiment on the market return volatility. We established a "Overall effect-Decomposition effect" linear model based on variables of investor sentiment and market return volatility, average correlation and average variance and conducted a multiple mediating effect test which taking the average variance and the average correlation as the mediating variable based on bootstrap sampling to analyze the specific conduction mechanism of the asymmetric volatility adjustment effect in different sentiment conditions in detail. The empirical results show that the average variance has a significant impact on the market return volatility under the range of rising sentiment and the average correlation has a stronger explanatory power to return volatility during depressed sentiment interval. Irrational sentiment component makes the asymmetric effect of the average correlation and the average variance on market return volatility to be significantly enhanced.