The Impact Of Analyst Rating Change On Herd Behavior In Chinese Funds Market

In recent years,securities investment fund industry is in rapid development stage,especially after the formal promulgation of The Law of the People7 s Republic of China on Securities Investment Fund on June 1,2004.The number of funds increased from 134 in 2004 to 5087 in 2018,the fund scale increased from 407.83 billion yuan to 8,932.91 billion yuan.The fund pools scattered funds and hands them over to professional institutions for portfolio investment,which increases profits,diversifies the investment risk,and activates the financial market.The fund has become an indispensable part of China's capital market.China's financial market is not completely effective,and one of the important manifestations of market inefficiency is herding behavior.At present,some scholars have studied the herd behavior of the domestic fund market(see[4],[7],[8],[10]for details).Based on this,this paper further analyzes the impact of analyst rating changes on herd behavior in Chinese funds market.With the continuous development and expansion of China's securities analyst team,research on the herd behavior of fund market and the impact of analyst rating changes on herd behavior have far-reaching significance for the sustainable development of the entire market and the correct investment decisions of fund managers.Firstly,this paper gives a basic overview of herd behavior,introduces the man-ifestation of herd behavior in China's securities markett,classifies herd behavior from different perspectives,and analyzes the production mechanism of herd be-havior from the perspective of information,reputation and compensation.Then we introduce several well-known empirical models for measuring herd behavior:LSV model and PCM model based on institutional investor trading behavior,CK model and CCK model based on the dispersion of equity returns,and indicate the basic form,principle and defects of these models.Then,this paper uses the LSV model to measure the overall level of herd behav-ior of fund managers when investing in stocks,and divides herding measure(HM)into buy-herding measure(BHM)and sell-herding measure(SHM)according to Wermers' method.After that,this paper introduces six control variables and four dummy variables that affect herd behavior,and establishes a regression model of herd behavior and analyst rating change(RATECHG).Fama-MacBeth regres-sion results not only confirms the significant impact of some control variables on herd behavior,but also indicates that the impact of analyst rating changes on herd behavior is significant,and the impact of analyst rating downgrade on sell-herding is particularly significant.In order to further analyze the internal mechanism of the relationship between analyst rating changes and herd behavior,this paper revises the above regression model:adds the interaction term of the analyst rating change and stock infor-mation environment(RATECHG×INFO)in the original regression model.The results shows that the impact of analyst rating changes on herd behavior is still significant.The public information of listed companies reduces the impact of analyst rating changes.That is to say,fund managers refers to stock disclosure information which analysts' rating results bases on,rather than just focus on analysts' rating results when making investment decisions.At the end of the paper,it shows that since the fund management company only announces the top 10 stock positions of the investment portfolio at the end of each quarter,the empirical study based on these stocks may underestimate herd behavior,especially sell-herding in financial market,therefore,the conclusions drawn in this paper may require further revision and improvement.