| Description |
This dissertation is composed of 3 chapters on the topics of investor sentiment and institutional trading momentum. In the first chapter, I investigate whether the returns to cross-sectional anomalies reported in the finance literature are due to investor sentiment. I present evidence of a weak relation between cross-sectional anomalies and investor sentiment. Using a larger collection of cross-sectional anomalies, I find that only a small subsample of these anomalies exhibits a relation with investor sentiment. This result does not appear to be due to certain anomalies being more sensitive to changes in macroeconomic conditions. Further I show that the predictive power of sentiment diminishes significantly after controlling for the Fama and French factors. These results suggest that the returns to active trading strategies are generally not due to sentiment-driven mispricing. In the second chapter, I investigate whether the relation between investor sentiment and cross-sectional anomalies is due to short sale constraints. I find that the average security in these strategies is not hard-to-short. Furthermore, the short leg does not appear to be harder to short or more overvalued than the long leg. However, I find that these strategies are more illiquid and have higher institutional ownership following low sentiment. These results imply that the relation between investor sentiment and profitable trading strategies could be due to illiquidity and institutional trading, rather than short sale constraints. Finally, in the third chapter I investigate whether the collective trades of financial institutions create mispricing in the stock market. Previous studies have generally found a positive relation between institutional demand and short-term returns, consistent with the interpretation that institutional trading pushes prices towards fundamental values. However, these studies do not control for the general and firm-specific trends in institutional ownership. After removing the trend in institutional ownership using the Hodrick-Prescott filter, I find strong evidence that financial institutions create substantial mispricing in the market. There is a large reversal in both returns and ownership following periods when ownership is abnormally high or low. |
