Deep Learning for Conditional Asset Pricing Models

2025/09/05 by Hongyi Liu, Liu, Hongyi
Decision Sciences · Economics, Econometrics and Finance · #Computational Finance (q-fin.CP) #FOS: Economics and business #Financial Markets and Investment Strategies #Stochastic processes and financial applications #Stock Market Forecasting Methods

paper · pdf · doi:10.48550/arxiv.2509.04812

openalex publication_date 2025/09/05 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We propose a new pseudo-Siamese Network for Asset Pricing (SNAP) model, based on deep learning approaches, for conditional asset pricing. Our model allows for the deep alpha, deep beta and deep factor risk premia conditional on high dimensional observable information of financial characteristics and macroeconomic states, while storing the long-term dependency of the informative features through long short-term memory network. We apply this method to monthly U.S. stock returns from 1970-2019 and find that our pseudo-SNAP model outperforms the benchmark approaches in terms of out-of-sample prediction and out-of-sample Sharpe ratio. In addition, we also apply our method to calculate deep mispricing errors which we use to construct an arbitrage portfolio K-Means clustering. We find that the arbitrage portfolio has significant alphas.

Citations

• Long Short-Term Memory
• Learning long-term dependencies with gradient descent is difficult
• Multilayer feedforward networks are universal approximators

Related

• Asset Pricing and Deep Learning
• Deep Learning in Asset Pricing
• Deep Learning for Predicting Asset Returns
• Deep Partial Least Squares for Empirical Asset Pricing
• From Factor Models to Deep Learning: Machine Learning in Reshaping Empirical Asset Pricing