Essays in empirical and behavioral asset pricing

Young, Trevor (2021) Essays in empirical and behavioral asset pricing. Doctoral thesis, University of London: London Business School.

Abstract

This thesis presents three essays. The first chapter examines the information content of aggregate mutual fund alpha. Recent evidence shows that mutual funds collectively buy overvalued stocks. I hypothesize that aggregate alpha arises in part from such stocks becoming even more overvalued, and thus is a measure of market overvaluation. A one-standard-deviation increase in aggregate alpha corresponds to a 0.83 percentage point decrease in the following month's excess market return, and aggregate alpha yields a monthly out-of-sample R2 of 3.64%. Moreover, higher aggregate alpha predicts higher anomaly returns and lower aggregate earnings surprises. The market-return predictability stems from funds with high investment in overvalued stocks, and from small-cap or growth investment styles. The evidence is difficult to reconcile with rational explanations based on benchmarking, flow, or catering. The second chapter, joint with Justin Birru, tests a straightforward time-series prediction of sentiment; sentiment should exhibit its strongest effects on asset prices at times when valuations are most subjective. Consistent with this, we show that a one-standard-deviation increase in aggregate uncertainty amplifies the predictive ability of sentiment for market returns by two to four times relative to when uncertainty is at its mean. For the cross-section of returns, the predictive ability of sentiment for test assets expected to be most sensitive to sentiment and for anomalies is substantially larger in times of higher uncertainty. The final chapter, joint with Justin Birru and Hannes Mohrschladt, provides a unifying explanation for a number of existing risk puzzles. Systematic mispricing primarily affects speculative stocks and tends to take the form of overpricing, predicting lower average returns. Because speculative stocks are typically deemed risky by rational models, failing to control for exposure to systematic mispricing can bias tests of risk-return tradeoffs. Controlling for the effects of systematic mispricing, we recover robust positive risk-return relations for a large number of cross-sectional risk proxies, including many low-risk, distress, and illiquidity anomalies.