This thesis examines whether differences in retail investor attention of different stocks measured by number of Google searches affect returns of the short-term return reversal phenomenon. I use monthly and weekly Russell 3000 stock data and Google Trends data between 2004 and 2018 to study the returns of portfolios that have stocks assigned to them based on their past returns and past abnormal search volume index (ASVI) values. I investigate whether the differences in attention affect the subsequent returns of the different groups in dissimilar ways. I also examine the drivers and effects of ASVI in a linear regression environment to study the factor further.
My findings show that ASVI has a significant effect on the past return portfolio retuns. High abnormal retail investor attention seems to create price pressure that weakens the negative reversal of past winners and boosts the positive reversal of past losers. Thus, my results indicate that in addition to liquidity shocks, investor attention is another driver of loser return reversal.
I document that a trading strategy that buys stocks with high abnormal attention and sells stocks with stable attention generates modest but consistent returns, and a short-term reversal strategy that applies this kind of ASVI screening provides significantly stronger returns than pure reversal strategies.
I find that in linear regressions, ASVI has a consistently positive but insignificant effect on following returns. I also document evidence that this effect of attention is stronger for smaller companies, although not always significantly.
ASVI as a factor is correlated with numerous other variables such as company size and past returns but the returns of ASVI trading strategies cannot be fully explained by other factors.
