An Empirical Study On The Correlation Between Idiosyncrasy Volatility And Expected Return

The correlation between idiosyncratic volatility and expected return is a perennial issue in financial circles.As a proxy index of idiosyncratic risk,idiosyncratic volatility is concerned has no correlation with expected return in the classic asset pricing theory,because investors could avoid it by holding diversified portfolios.However,the assumption of the classic asset pricing theory is so stringent that it is hard to satisfy it in reality.Some scholars believe that investors can't avoid idiosyncratic risk entirely,idiosyncratic risk should obtain risk premium as the systemic risk do,therefore,the correlation should be positive.Other scholars believe the correlation is negative because of short-sales constraints and heterogeneous beliefs.Although empirical studies on this problem don't come to a unified conclusion,most of the empirical research believe that the correlation is negative,this phenomenon is known as "idiosyncratic volatility puzzle".Many domestic studies show that idiosyncratic volatility puzzle also exists in China stock market,and it's wildly attributed to the strict restrictions on short selling.This paper uses the portfolios method,the two-dimensional analysis method and the Fama-Macbeth method to test whether idiosyncratic volatility puzzle exists in China stock market,we also try to find the causes of the negative correlation.In order to find out the relation between idiosyncratic volatility puzzle and margin trading,we take March 2010,when the securities margin trading was officially carried,as the time division point.We find that idiosyncratic volatility puzzle actually exists in China stock market,no matter which method is used to calculate idiosyncratic volatility.The negative correlation has weakened after the introduction of securities margin trading,but it's still significant.We attribute it to the special investor structure in China stock market.We also find that turnover rate can partly explain idiosyncratic volatility puzzle.Our unique contribution lies in the following points:first,in order to test the stability of idiosyncratic volatility puzzle,in addition to the mainstreaming idiosyncratic volatility measurement method,which is known as Fama-Macbeth method,we use bankruptcy risk Z model to measure idiosyncratic volatility.We also use various method to remove serial correlation.Second,as securities margin trading has been running for more than five years,we have enough data to test whether the idiosyncratic volatility puzzle has changed after securities margin trading was carried.one shortcoming of this article is that among all the risk factors we choose,turnover rate is the single one which could explain idiosyncratic volatility puzzle,we fail to find all causes of this phenomenon.The other shortcoming lies in that the dividion of the data,it should has been more detailed.We could use latest idiosyncratic volatility measurement model and choose more risk factors in future studies.