Does the time horizon of the return predictive effect of investor sentiment vary with stock characteristics? A Granger causality analysis in the frequency domain

Abstract

Behavioral theories posit that investor sentiment exhibits predictive power for stock returns, whereas there is little study have investigated the relationship between the time horizon of the predictive effect of investor sentiment and the firm characteristics. To this end, by using a Granger causality analysis in the frequency domain proposed by Lemmens et al. (2008), this paper examine whether the time horizon of the predictive effect of investor sentiment on the U.S. returns of stocks vary with different firm characteristics (e.g., firm size (Size), book-to-market equity (B/M) rate, operating profitability (OP) and investment (Inv)). The empirical results indicate that investor sentiment has a long-term (more than 12 months) or short-term (less than 12 months) predictive effect on stock returns with different firm characteristics. Specifically, the investor sentiment has strong predictability in the stock returns for smaller Size stocks, lower B/M stocks and lower OP stocks, both in the short term and long term, but only has a short-term predictability for higher quantile ones. The investor sentiment merely has predictability for the returns of smaller Inv stocks in the short term, but has a strong short-term and long-term predictability for larger Inv stocks. These results have important implications for the investors for the planning of the short and the long run stock investment strategy.