Deep Fundamental Factor Models
This improves financial factor modeling for investors by providing better performance and interpretability, though it is an incremental extension of existing methods.
The authors developed deep fundamental factor models to automatically capture non-linearity and interaction effects in financial factor modeling, achieving information ratios approximately 1.5 times greater than OLS factor models on a dataset of 3290 assets.
Deep fundamental factor models are developed to automatically capture non-linearity and interaction effects in factor modeling. Uncertainty quantification provides interpretability with interval estimation, ranking of factor importances and estimation of interaction effects. With no hidden layers we recover a linear factor model and for one or more hidden layers, uncertainty bands for the sensitivity to each input naturally arise from the network weights. Using 3290 assets in the Russell 1000 index over a period of December 1989 to January 2018, we assess a 49 factor model and generate information ratios that are approximately 1.5x greater than the OLS factor model. Furthermore, we compare our deep fundamental factor model with a quadratic LASSO model and demonstrate the superior performance and robustness to outliers. The Python source code and the data used for this study are provided.