The Research Of Herding Behavior In Stock Market Based On Quantile Regression Model

Policy change, the financial crisis, institutional change event shocks often have huge impact on the financial markets; the stock price has been frequently intervened by the changing environment in financial markets. As the main inducing factors of the financial crisis, the emergence of the herd behavior poses challenges to the traditional theory of financial crisis. Herding behavior refers to the irrational behavior of investors which can not explained by traditional investment decision theory, When investors to make a major investment decision, They often ignore their own information, and to emulate the behavior of other investors, Finally the share price deviate from the rational price described in asset pricing theory, Thus inevitably caused the stock market investment risk, Eventually led to the investment market environment worsening day by day. But most scholars detect characteristic herding in the stock market through the establishment of traditional regression analysis model.The regression coefficients estimated by traditional regression analysis is unbiased and valid when the model is a random disturbance term with zero mean and variance of a normal distribution. However, in practical economic phenomena, the economic data showed heteroscedasticity, autoregression and other problems often cause these estimates are no longer optimal and stable. Therefore, this paper introduce the quantile regression analysis when discuss herd behavior characteristics of emerging markets,More concerned about the herd behavior characteristic differences in different stock market conditions.Under the background of large T data, the smooth data process could not guarantee variable relationship didn't exist structural changes, Therefore, In view of the fat-tailed features of financial data distribution and the individual behavior degeneration of cross-section data. By introducing the quantile regression method to construct the herding behavior quantile regression model of stock market and to describing the qualitative variation, capture the commonness and characteristics of herd behavior in the rise and fall financial and economic system environment. At the same time, take the emerging stock markets such as China, India,Brazil as research objects, comprise the herd behavior of those market, Further more, Depict the characteristics of herd behavior in different stock market.The results indicate that the model parameter of 1?and 2?is positive in the six emerging market. It means the cross-sectional absolute deviation is increased as the market price. At the other hand, the model parameter of 3?and 4?is negative, It means no matter in rising or falling market, the linear relationship of the cross-sectional absolute deviation with the market price can not be found. From the analysis result, we can find that the local characteristics exist in the herd behavior research. In this paper, we can conclude that the herd behavior is more likely exist in the low market(the market price is negative), In order to evade the loss happened in the low market and the high risk in the stock market, the investors in those emerging market are more inclined to imitate the other investors, and make the same investment decisions, this consistent decision will trigger the herd behaviour.