Investor Sentiment And The MAX Effect

Bali et al.(2011)uncover the MAX effect which is a new anomaly in the stock market.One of the main reasons for this anomaly is that stocks with high maximum daily returns(high MAX)over the past month perform poorly relative to stocks with low maximum daily returns(low MAX)over the past month.Another reason is investor sentiment,as speculative investors tend to prefer stocks with higher returns,but these high MAX stocks do not perform well in the following month,resulting in the MAX effect.Investor sentiment can well reflect the psychology of investors in the stock market,it will have some impact on the investors' decision-making,and it is also one of the hot issues in behavioral finance research.Therefore,based on the introduction of the three types of investor sentiment indicators in the stock market,this paper selects the closed-end fund discount,the number and the first-day returns of IPOs,Shanghai and Shenzhen A-share market turnover and consumer confidence index,six emotional indicators,based on Baker and Wurgler(2004)and Yi Zhigao and Mao Ning(2009),and the use of principal component analysis to construct the comprehensive index of investor sentiment.Thus,we can investigate the effects of investor sentiment on the MAX effect.First of all,this paper uses the daily and monthly data of all A-shares in Chinese stock market to sort stocks based on their maximum daily returns(MAX)over the past month to form decile portfolios.Decile 1(D1)is the portfolio with the lowest maximum daily returns in the past month,and D10 is the portfolio with the highest maximum daily returns in the past month.On this basis,the three-factor model of Fama and French(1992,1993)is established,which proves that the MAX effect exists in the Chinese stock market.Moreover,we show that a MAX strategy that longs a value-weighted portfolio of high MAX stocks and shorts a value-weighted portfolio of low MAX stocks produces a negative alpha based on the three-factor model of Fama and French(1992,1993).More clearly,one of the reasons for the MAX effect is that stocks with higher returns in the month do not perform as well in the next month as stocks with lower returns.Secondly,this paper will use the principal component analysis to build the investor sentiment index,applied to empirical research.Following Baker and Wurgler(2007)and Stambaugh et al.(2012),we define a high(low)sentiment month as one in which the ISI is above(below)the sample median value,and denote these sentiment states by 1 and 0 respectively.At this point,all months in the sample period are divided into high investor sentiment months and low investor sentiment months.After the establishment of three-factor model of Fama and French(1992,1993),we found the MAX effect is strongly driven by the high-sentiment state.The empirical results show that the MAX effect exists in the stock market during the period of high investor sentiment,while the investor sentiment makes the MAX effect disappear.This indicates that the MAX effect only exists in the period of high investor sentiment state.The above research on the MAX effect is only based on the portfolio level.The only benefit of that is non-parametric,because we do not need to establish any functional relationship to estimate the coefficient.However,if only the portfolio level of research,it will weaken the data contained in some of the unique information.So,finally,we have a Fama-Macbeth cross-section regression on the return of the stock to prove that there is a MAX effect in the stock market,and investor sentiment will expand the MAX effect.This article takes a fresh look at investor sentiment and the MAX effect and gives investors a better investment proposition.Our findings imply that investors should consider reducing their investments in high MAX stocks when sentiment is high since it is precisely during such periods that such stocks are overpriced.