We propose a new method, variable subsample aggregation (VASA), for equity return prediction using a large‐dimensional set of factors. To demonstrate the effectiveness, robustness, and dimension reduction power of VASA, we perform a comparative analysis between state‐of‐the‐art machine learning algorithms. As a performance measure, we explore not only the global predictive but also the stock‐specific
R2's and their distribution. While the global
R2 reflects the average forecasting accuracy, we find that high variability in stock‐specific
R2's can be detrimental for the portfolio performance. Since VASA shows minimal variability, portfolios formed on this method outperform the portfolios based on random forests and neural nets.