Subsampled factor models for asset pricing: The rise of Vasa

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 R-2's and their distribution. While the global R-2 reflects the average forecasting accuracy, we find that high variability in stock-specific R-2'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.