Research On Herd Behavior In The Securities Market

This paper focuses on further studies and innovations in both theoretical and empirical research on herd behavior. Firstly, it provides an overview of the recent theoretical and empirical research on herd behavior in security markets. It looks at what precisely is meant by herding, the causes of herd behavior, the existing studies in identifying the phenomenon, and the influences that herding has on security markets.Secondly, based on the traditional BHW model, this paper proposes a new theoretical research model by taking the information cost and the risk aversion of investors into consideration. Application of Bayes’rule and the inductive method shows that only the first risk-averse investor would pay for the information with the optimal cost and the information cascades starts with the second risk-averse investor.Thirdly, an ARCH testing model which is derived from the basic idea of GCAPM is proposed to detect herd behavior on the Shanghai Stock Market. Empirical test finds that a great number of portfolios exist herd behavior on the Shanghai Stock Market, especially when the market is in fluctuation. Furthermore, clustering analysis is added in the model which helps both extract characteristics from data and figure out when herding happens.As a result, some suggestions based on the conclusions are proposed to the investors and the relevant government departments.