We estimate the stock’s likelihood of extreme returns by measuring the extent to which the stock’s trades are correlated with market-wide and industry-wide trades during normal times, referred to as herding. We find that stocks whose trades herd most with aggregate-level trades experience most negative (positive) returns during market crashes (booms). While herding generates extreme returns in both sides, investors appear to demand compensation for the possibility of extreme low returns. This is the case even when we control for standard asset pricing variables and other tail risk proxies.

We argue that anomalies may experience prolonged decay after discovery and propose a Bayesian framework to study how that impacts portfolio decisions. Using the January effect and short-term index autocorrelations as examples of disappearing anomalies, we find that prolonged decay is empirically important, particularly for small-cap anomalies. Papers that document new anomalies without accounting for such decay may actually underestimate the original strength of the anomaly and imply an overstated level of the anomaly out of sample. We show that allowing for potential decay in the context of portfolio choice leads to out-of-sample outperformance relative to other approaches.

Recent years have seen increased demand from institutional investors for passive replication products that track the performance of hedge fund strategies using liquid investable assets such as futures contracts. In practice, linear replication methods suffer from poor tracking performance and high turnover. We propose a model combination approach to index replication that pools information from a diverse set of pre-specified factor models. Compared with existing methods, the pooled clone strategies yield consistently lower tracking errors, generate less severe portfolio drawdowns, and require substantially smaller trading volume. The pooled hedge fund clones also provide economic benefits in a portfolio allocation context.

We find that higher stock lending fees predict significantly lower future returns after controlling for shorting demand for US stocks during the period 2007–10. These results suggest that active institutional investors on the supply side play an important role in the return predictability of fees and they not only respond to demand but also price in additional information around earnings news announcements. Overall, we find evidence that stock lenders are informed and, together with short sellers, contribute to the price discovery process.