ReSGA: A Large Tail Risk Model for Learning Value-at-Risk and Expected Shortfall
2026-06-03 • Machine Learning
Machine Learning
AI summaryⓘ
The authors developed a new model called ReSGA to better predict financial risks, specifically Value-at-Risk and Expected Shortfall, by using a large number of parameters to capture complex patterns in asset data. They tested this model on nearly a century of US stock data and found it performed better than many other traditional and machine learning methods. Their results also showed that the model's accuracy mainly comes from the complexity of the data rather than just the model itself. Additionally, the model is interpretable and works well across different markets. This makes it useful for managing financial risks more effectively.
Value-at-Risk (VaR)Expected Shortfall (ES)autoencodercross-sectional dependencetemporal dynamicsfinancial risk modelingbacktestingmachine learningtransfer learningfeature importance
Authors
Yichi Zhang, Ke Zhu, Zhoufan Zhu
Abstract
Learning Value-at-Risk (VaR) and Expected Shortfall (ES) is important for managing financial risks effectively. Existing approaches with limited parameters are vulnerable to model misspecification in the era of big data. To address this limitation, we propose a large tail risk model, the retrieval-enhanced self-grouping autoencoder (ReSGA), which is designed with millions of parameters to exploit the rich cross-sectional dependence and long-term temporal dynamics of assets using their characteristics. Applied to monthly US equity returns from 1926 to 2023 with 153 firm characteristics, ReSGA outperforms twelve econometric and machine learning competitors in terms of out-of-sample loss and statistical backtesting. In addition, its forecast advantages can translate into significant economic gains from long-short decile portfolios that are constructed by a new size-enhanced left-side momentum strategy. To clarify the role of complexity, we further conduct a systematic scaling analysis and demonstrate that improvements in joint VaR-ES forecasting are primarily driven by data complexity rather than model complexity. Finally, our analyses of group-importance and transfer-learning exhibit the interpretability and cross-market generalizability of ReSGA.