Purpose-The purpose of this paper is to analyze the trading behaviors of retail investors and investigate their impacts on stock rets.Design/methodology/approach-As retail investors are considered as the main noise traders in the capital market,using the trading records of Chinese retail investors from 2005 to 2009,the authors study their trading preferences and the correlation of their trades.Then,they use a multifactor model to test whether the co-movement of stock rets could be explained by individual sentiment.Findings-The authors’ results show that the small-cap stocks are obviously preferred by retail investors.Meanwhile,the net stock demands of retail investors are systematically correlated,even when the effect of market risk is excluded.In the perspective of the net stock demands,the authors use BSI to measure the individual sentiment,finding that individual sentiment plays an important role in the formation of the cross-section of stock rets.However,the authors’ results imply that BSI is a reverse indicator to predict the future rets,which implies that the trading behaviors of retail investors are irrational.Originality/value-Consistent with behavioral theory,the authors’ findings support the viewpoint that stock rets could be affected by the systematic correlated trading of retail investors.To some extent,their findings highlight the need to know more details of individual investors’ trading behaviors through which the fluctuations of asset prices can be better understood.