Machine learning models have been successfully employed to cross-sectionally predict stock returns using lagged stock characteristics as inputs. The analyzed paper challenges the conventional wisdom that more training data leads to superior machine learning models for stock return predictions. Instead, the research demonstrates that training market capitalization group-specific machine learning models can yield superior results for stock-level return predictions and long-short portfolios. The paper showcases the impact of model regularization and highlights the importance of careful model design choices.

Traditional asset pricing literature has yielded numerous anomaly variables for predicting stock returns, but real-world outcomes often disappoint. Many of these predictors work best in small-cap stocks, and their profitability tends to decline over time, particularly in the United States. As market efficiency improves, exploiting these anomalies becomes harder. The fusion of machine learning with finance research offers promise. Machine learning can handle extensive data, identify reliable predictors, and model complex relationships. The question is whether these promises can deliver more accurate stock return predictions…