ABSTRACT Farmer, Schmidt, and Timmermann (FST) document time‐variation in market return predictability, identifying “pockets” of significant predictability through kernel regressions. However, our analysis reveals a critical discrepancy between the method outlined by FST and the code actually implemented. Instead of using a one‐sided kernel, which guarantees out‐of‐sample forecasts, they perform in‐sample estimation with a two‐sided kernel. As a result, future information leaks into the forecasting model, undermining its reliability. Rectifying this error qualitatively alters the findings, invalidating most conclusions of the FST study. Thus, attempts to exploit such “pockets”—should they exist—offer little help in forecasting market returns.

Abstract We propose CTREND, a new trend factor for cryptocurrency returns, which aggregates price and volume information across different time horizons. Using data on more than 3,000 coins, we employ machine learning methods to exploit information from various technical indicators. The resulting signal reliably predicts cryptocurrency returns. The effect cannot be subsumed by known factors and remains robust across different subperiods, market states, and alternative research designs. Moreover, it survives the impact of transaction costs and persists in big and liquid coins. Finally, an asset pricing model that incorporates CTREND outperforms competing factor models, providing a superior explanation of cryptocurrency returns.