Study Of The Investor Sentiment Based On PLS And Market Anomalies

Market anomalies are empirical results that are inconsistent with the Efficient Market Hypothesis theory.Their existence indicates the inefficiency of market and the deficiency of the current asset pricing models(Schwert,2003).The classic financial theory insists that stock price reflects the fundamental values,and the prices are always moving within the reasonable range.Actually,there are plenty of empirical results supporting that stock prices are not always moving within the reasonable range.Some indicators or events are significantly correlated with the stock future returns.On account of the limits of the classic financial theory' explanatory power,many scholars are combining psychology,sociology and experimental economics etc.to explore the market anomalies from the aspect of behavioral finance.Investor sentiment is one of the core concept in behavioral finance.The theory of investor sentiment argues that many investors make the same correlated mistakes.Lots of foreign literature supports the fact that investor sentiment is a systematic irrational factor that will make the asset price derive from its fundamental value.It is not so early for domestic scholars to explore A-share market anomalies from the aspect of investor sentiment theory.Many scholars prefer the mainstream measurement,Principal Component Analysis,to measure the investor sentiment in A-share market.The accuracy of investor sentiment measurement determines its predictability of stock returns.Huang et al.(2015)has provided abundant empirical evidence to prove the fact that Partial Least Square can eliminate a common noise component in sentiment proxies,which overcomes the weakness of PCA.Therefore,this paper is going to use PLS measure the investor sentiment,and conduct the study as followed:first,use PLS to construct investor sentiment indicator SPLS and compare it with the sentiment indicator SCICSI constructed by PCA method to verify the superiority of SPLS;second,build up long leg and short leg strategies according to 10 anomalies indicators and verify their significance and independence in A-share market;third,study the relationship of the improved sentiment indicator SPLS and the future excess stock returns,as well as the benchmark-adjusted returns of the long leg and short leg portfolios.The research meaning of this paper is that it explores several A-share market anomalies at the same time from the angle of investor sentiment based on the improvement of sentiment indicator accuracy,and finds out the following empirical results:(1)investor sentiment can explain 9 A-share significantly existing market anomalies;(2)investor sentiment has significantly positive correlation with the stock future returns,even under the control of MKT,SMB and HML factors;(3)as the sentiment goes high,the overvalued stocks are mispriced even harder,while the undervalued ones start to recover to the fundamental values.