Research

Job Market Paper

Abstract: The illiquidity premium for publicly traded equities has been low in recent years. I argue that it is masked by short sale constrained stocks, which tend to be illiquid and are known to have low expected returns. Once short sale constrained stocks are removed, the illiquidity premium re-emerges. I develop a model to better understand the strong relationship between liquidity and short sale constraints. The model predicts that the adverse selection cost and inventory carrying cost of providing liquidity increases with the cost of borrowing shares to short, resulting in a deterioration in liquidity. I find empirical evidence that liquidity deteriorates for short sale constrained stocks through both of these channels.

Published Papers

with Andrei Gonçalves 

Abstract: Recent evidence indicates the value premium declined over time. In this paper, we argue this decline happened because book equity, BE, is no longer a good proxy for fundamental equity, FE, defined as the equity value originating purely from cash flows. Specifically, we estimate FE for public US firms (from 1973 to 2018) and find that the premium associated with the fundamental-to-market ratio, FE/ME, subsumes the BE/ME premium and has been stable while the cross-sectional correlation between FE/ME and BE/ME decreased over time, inducing an apparent decline in the value premium. We also show that FE/ME provides a better value premium signal than several alternative valuation ratios beyond BE/ME. 

with Joseph Engelberg, Richard Evans, Adam Reed, and Matthew Ringgenberg 

Abstract: We find that equity loan fees are the best predictor of cross-sectional returns. When compared to 102 other anomalies, the loan fee anomaly has the highest monthly long-short return (1.17%), has the highest monthly Sharpe Ratio (0.40), and unlike other anomalies, exhibits strong persistence throughout the sample. We show that 28% of the loan fee anomaly can be explained by its selective exposure to the best performing anomalies, while 72% is due to unique information possessed by short sellers. Our results show that short sellers' willingness to pay prices the cross-section of stocks and these "best ideas'' outperform other anomalies. 

Works-in-Progress

Why Does Short Interest Predict Stock Returns?

with Adam Reed and Matthew Ringgenberg 

Abstract: High short interest predicts low future stock returns even though short interest data is publicly observable. We show theoretically that the marginal cost of short selling increases rapidly as positions grow because higher short sales demand can lead to higher loan fees on pre-existing positions.  Thus, the cost of shorting one additional share exceeds the return to shorting one additional share before mispricing is eliminated. As a result, short selling predicts returns in equilibrium. Using transaction level data, we document evidence consistent with the model.  Marginal loan fees are significantly higher than observed loan fees.