Causal Policy Analysis in Financial Time Series Using Deep Learning X-Learner
Abstract
This study employs the X-Learner algorithm with deep learning models (MLP, LSTM, GRU, CNN) to causally estimate performance differences between cyclical (T=1) and defensive (T=0) asset classes in financial time-series data. The results show that the MLP-based policy achieved a Sharpe ratio of 0.440, while the LSTM-based policy improved to 0.990, outperforming T1 (0.880) and T0 (0.493). These findings indicate that incorporating temporal dependence through LSTM enhances causal policy learning. By reframing asset allocation as a causal decision-making problem rather than a predictive task, this study provides a foundation for interpretable and adaptive investment strategies in financial markets.
References
- Y. He, et al. "X-learner: Learning cross sources and tasks for universal visual representation." European Conference on Computer Vision. Cham: Springer Nature Switzerland, 2022.
- Smith, Bevan I., Charles Chimedza, and Jacoba H. Bührmann. "Treatment effect performance of the X-Learner in the presence of confounding and non-linearity." Mathematical and Computational Applications 28.2 (2023): 32.
- Kim, Seungmo, and Taeyeon Oh. "Employing the X-Learner Algorithm to Evaluate the Intervention Effects of Physical Activity on Determinants of Elderly Mental Health." Healthcare. Vol. 13. No. 11. MDPI, 2025.
- Künzel, Sören R., et al. "Metalearners for estimating heterogeneous treatment effects using machine learning." Proceedings of the national academy of sciences 116.10 (2019): 4156-4165.
- He, Yinan, et al. "X-Learner: Learning Cross Sources and Tasks for Universal Visual Representation (Supplementary Material)." Parameters 23: 11-761.
- Lee, Seung‐Jun, et al. "Predicting Individual Treatment Effects to Determine Duration of Dual Antiplatelet Therapy After Stent Implantation." Journal of the American Heart Association 13.19 (2024): e034862.
- Upadhyaya, P., et al. "A Retrospective Causal Inference-based Study Using Machine Learning for Identifying Treatment Effects of Various Therapies in Sepsis-induced Acute Respiratory Failure Phenotypes." C22. ARTIFICIAL INTELLIGENCE IN THE ICU: THE MACHINE WILL SEE YOU NOW. American Thoracic Society, 2024. A5071-A5071.
- N.H. Jung, T. Oh and K. H .Kim, “A Study on DRL-based Efficient Asset Allocation Model for Economic Cycle-based Portfolio Optimization”, Journal of Korean Society for Quality Management, Vol 51, No. 4, pp.573-588, 2023.
Keywords
Details
| Section | Articles |
| Issue | Vol. 2 No. 1 (2025): Volume 2 Issue 1 (November 2025) |
| Published | 2025-11-28 |
| Pages | 41-49 |
- aSSIST University / Ph. D. Candidate
- Seoul National University, Ph. D. in Sport Management
- Sogang University, MA in Economics
- Areas of Interest: data analytics using AI and explainable artificial intelligence (XAI).
- After earning a Ph.D. in Engineering from the University of Texas at Austin, he worked for various companies and is currently serving as an Assistant Professor and Program Director of the Ph.D. program in AI Convergence Engineering at the Seoul School of Integrated Sciences & Technologies (aSSIST University). His primary research areas include artificial intelligence and retail innovation.
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.