Learning Behavior Of Investors Who Have Heterogeneous Beliefs And Stock Market Volatility

The stock market is unpredictable.The Shanghai Composite Index breached 2000-points in 2008,and the eternal spectacle happened in A-share market in the first half of 2015,which is thousands of shares fell or rose by their daily limits and thousands of shares were suspended repeatedly.These are sufficient to show that the wild swings of Chinese stock market.As the core of the stock market,investors are bounded rationality limited by physiology,knowledge and ability,so they can only make a satisfactory decision,but not the optimal solution.Furthermore,Investors learn imperceptibly to form the beliefs for the future.They have heterogeneous beliefs,which lead to different expectations and different investment behaviors for different individuals.Then they can affect the stock price.In order to explore the relationship between learning behavior of investors and stock market volatility in detail,firstly,by studying the Granger causality between the number of Shanghai A-share new accounts and Shanghai A-share Index,we tested the effect of changes in investors’ scale on the stock index.Then we constructed the asset pricing model with the change of investors’ scale.Secondly,according to experience to divide the investor groups,investors in the same group have homogeneous beliefs,and investors in different groups have heterogeneous beliefs.Then we established the microscopic mechanism of investors’ learning behavior.The investors have heterogeneous beliefs.They learn to form the beliefs,which including the growth rate of dividends and of the stock price.They form their beliefs about the growth rate of the stock price through learning from three aspects,which including learning from personal experience,social learning and policy learning.Social learning is that investors guess or reference to other investors’ expectations and policy learning is that investors judge the pros and cons of the economic policy on the stock market.Thirdly,we also explored the feedback loop that investors learn about the stock equilibrium price which is the result of their collective actions given their beliefs through learning.Lastly,we used the dividend series of Shanghai A-share for the period March 2008 to December 2015 as the simulation data of dividends for numerical simulation.The correlation coefficient between the analog values and the actual values of the price-dividend ratio is more than 90%,and the fluctuation trend and range of the two are basically the same.This thesis pointed out that the learning behavior of investors who have heterogeneous beliefs is an important reason for the stock market volatility.At different times,different weights are given to different learning behaviors.During the boom and bust,the vast majority of investors preferred social learning and policy learning.While in the relatively stable period,the weight of the vast majority of investors learn from their personal experience has increased.The increase of inexperienced investors will exacerbate the volatility of the stock market.While investors summarize and reflect on some deviations at the end of the period,which can effectively reduce the volatility of the stock market.Therefore we recommend that investors should keep learning to accumulate experience.New entrants shouldn’t be fanatical or panic blindly.Investors need to communicate with each other for reduce errors of someone guess what others expect,which is due to lack of information.Moreover,in order to ensure the healthy and stable development of the stock market,regulators need to establish a real-time monitoring platform of investor confidence index,and they should consider the impact of economic policies on investor expectations.