This repository contains empirical results in paper to estimate a general non-linear asset pricing model with a deep neural network applied to all U.S. equity data combined with all relevant macroeconomic and firm-specific information.

• Empirical Results
• Related Resources
• Author

We compare our GAN model, with a simple forecasting feedforward network model labeled as FFN, the linear special case of GAN labeled as LS and a regularized linear model labeled as EN.

• Sharpe Ratio

• Explained Variation

• Cross-Sectional R2

• Deep Learning in Asset Pricing (Chen, Pelger and Zhu 2019)
• Pre-trained Models
• Data
• GitHub Repository

• Luyang Chen