MAX Effect Research Based On Microstructure

MAX effect is one of anomalies in financial markets,which is contrary to the efficient market theory.It is discovered by Bali et al.(2011),refers to the phenomenon that securities in the stock market with maximum daily returns in the previous period,the expected returns are lower.Although some discussions were focused on the vision,the effect of market microstructure is not found.For one thing,according to the market microstructure theory,the organizational structure of the stock market,the trading mechanism and information disclosure system could influence the formation of the stock price.In addition,they also affect the returns of securities.For another,micro structure factors,such as bid-ask spread,can lead to market friction.Except for real value,microstructure noise is also contained within the stock price,but causes asset pricing biases at the same time.Based on the above reasons,this paper tries to empirically test the MAX effect from two aspects,the market differences and the market micro structure noise.Empirical data includes common stocks in NYSE,AMEX and NASDAQ for the period from January 1927 to December 2014.Then the paper builds MAX index based on the maximum daily returns in last month.Based on short selling stocks with low MAX and buying long stocks with high MAX,the paper constructs portfolios.First of all,with time series analysis and cross-section regression,the article tests MAX effect in three stock markets respectively.The results show that there are significant differences between the high and low MAX portfolios in the three stock markets.After controlling the systemic risk by commonly used asset pricing factor models,the abnormal returns between the portfolios are still significant.The above results imply that the MAX effect is not caused by the market differences.Last but not least,the paper constructs two stock return series,which are calculated with mid-bid-ask and bid respectively in order to eliminate the microstructure noises in returns.Then,the paper uses portfolio analysis and unilateral T test for returns with and without noise.Although the differences between the long and short portfolio returns are still significant,spreads decreased significantly.The results mean that the market microstructure noise is not main cause of the MAX effect,but has a certain decreasing effect.Based on the microstructure of stock market,the paper,which applies the market microstructure theory to the financial empirical study,empirically studies the causes of MAX effect.The paper provides empirical support for the market microstructure theory.At the same time,it offers a new avenue for the research of MAX effect,and enriches the existing research literature of MAX effect.Finally,it also has certain reference significance to the financial investment and theory.