An Empirical Analysis Of The Influence Of Herding Behavior On Stock Volatility

In the process of pricing,trading and risk control of financial derivatives,volatility is often the first priority to be considered.Through the study of stock price fluctuation,we can better grasp the operation law of the stock market and better understand the fluctuation law of the stock price.In addition,there are many possible reasons for stock price volatility,such as economic factors,political factors,company's own factors,investor factors.This paper will mainly study an irrational behavior of investors,namely herding behavior,and study its impact on stock volatility.So will there be herding behavior in stocks? If there is herding behavior,will it have a certain impact on stock volatility? Is it worsening stock volatility or reducing stock volatility? And is there a difference in the impact of herd behavior on stocks in different sectors and on different periods of stocks?This paper firstly studies the existence of the herding behavior in the GEM stocks and the Shanghai Stock Exchange through the CCK model,and classifies the two stock markets according to the situation of the bull market.According to the research results,the herding behavior in the SSE 180 Index is more significant than the herding behavior in the GEM 300 index;and in both of them,there is a herding behavior when stocks are in a bear market,but when the stock market is in a bull market,there is no herding behavior in the stock data.Secondly,the daily logarithmic rate of return data of the GEM 300 Index and the SSE 180 Index are respectively modeled by GARCH(1,1)and the volatility sequence is extracted.The stock data with the herding behavior is selected,and the Granger causality test is carried out between the cross-sectional absolute deviation degree(CSAD)sequence and the stock volatility sequence.According to the test results,the herding behavior is the Granger cause of stock volatility,so the GARCH-Jump model was established to further study the impact of herd behavior on stock volatility.The jump phenomena in the yield series is usually represented by a Poisson distribution,and for the jump intensity part of the GARCH-Jump model,the fixed constant is gradually extended to the dynamic model from the beginning.In this paper,the herding behavior measurement index(CSAD)is introduced on the basis of the GARCH-Jump model,so that the relationship between the herding behavior and stock volatility is further research.In this paper,GARCH-Jump models with different jump intensity models are established,and the logarithmic likelihood and parameter estimates of the models are calculated by maximum likelihood estimation.According to the likelihood ratio test results of different models,the GARCH-Jump model with the herding behavior index has the best fitting effect,indicating that the herding behavior does have an impact on stock volatility;according to the estimation results of the model parameters,there is a significant negative correlation between the CSAD and the jump intensity,that is to say,the occurrence of the herding behavior will enhance the jump intensity,resulting in the stock volatility becoming larger and stronger.In addition,we also find that the herding behavior on stock volatility in the GEM 300 Index is more significant than that in the SSE 180 Index.