ESG Fragility and Downside Risk: An Interpretable Machine Learning Framework for Predicting Corporate Vulnerability During Market Turbulence
Keywords:
ESG fragility, downside risk, interpretable machine learning, corporate vulnerability, market turbulence, SHAP, stress index, financial governance, algorithmic transparencyAbstract
Environmental, social, and governance (ESG) ratings have become central to modern investment decision-making, yet the fragility of ESG allocations during periods of acute market turbulence remains poorly understood. This paper develops an interpretable machine learning framework designed to predict corporate vulnerability by integrating ESG signals with downside risk metrics. The framework leverages gradient-boosted trees and attention-based neural architectures to capture nonlinear interactions between ESG subcomponents, macroeconomic stress conditions, and firm-level financial indicators. A key contribution is the incorporation of a PCA-APT stress index that distills systemic risk factors into a tractable vulnerability score. To address the opacity of black-box models, the framework employs SHAP-based explanations and counterfactual analysis, enabling stakeholders to trace the drivers of vulnerability predictions. Empirical validation on a large panel of U.S. listed firms from 2010 to 2024 demonstrates that the model outperforms traditional logistic regression and random forest baselines in identifying firms that experience severe drawdowns during volatility clusters. The study also reveals that governance factors contribute disproportionately to downside risk during crises, while environmental and social dimensions exhibit time-varying relevance. Additionally, the framework highlights the dangers of leakage in benchmark design, advocating for economically credible evaluation protocols. The paper concludes by discussing policy implications for ESG rating agencies, regulatory stress-testing infrastructure, and the governance of algorithmic risk assessment in financial systems. The proposed approach offers a transparent, deployable tool for portfolio risk management and systemic surveillance, with broader implications for the robustness of socio-technical financial infrastructures.
References
1. Berg, F., Koelbel, J. F., & Rigobon, R. (2022). Aggregate confusion: The divergence of ESG ratings. Review of Finance, 26(6), 1315–1344.
2. Liu, T. (2026). A Comparative Study of Transformer-Based and Classical Models for Financial Time-Series Forecasting. Journal of Risk and Financial Management, 19(3), 203.
3. Ang, A., Chen, J., & Xing, Y. (2006). Downside risk. Review of Financial Studies, 19(4), 1191–1239.
4. Taleb, N. N. (2007). The Black Swan: The impact of the highly improbable. Random House.
5. Christensen, H. B., Serafeim, G., & Sikochi, A. (2022). Why is corporate virtue in the eye of the beholder? The case of ESG ratings. The Accounting Review, 97(1), 147–175.
6. Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30, 4765–4774.
7. Xue, P., & Ye, Y. (2026). Attention-enhanced reinforcement learning for dynamic portfolio optimization. Intelligent Systems with Applications, 200622.
8. Liu, T. (2026). Leakage-Safe Benchmark Design for Market-Stress Early Warning: An Economically Credible Evaluation.
9. Liu, T. (2026). PCA-APT Stress Index for Market Drawdowns.
10. Hu, L., & Shen, Y. (2026). A predictive analytics approach for forecasting global stock index returns using deep learning techniques. Decision Analytics Journal, 100685.
11. Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52(1), 57–82.
12. Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56.
13. Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd ed.). Springer.
14. Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). “Why should I trust you?” Explaining the predictions of any classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 1135–1144.
15. Khandani, A. E., Kim, A. J., & Lo, A. W. (2010). Consumer credit-risk models via machine-learning algorithms. Journal of Banking & Finance, 34(11), 2767–2787.
16. Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–794.
17. Giglio, S., Kelly, B., & Pruitt, S. (2016). Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), 457–471.
18. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30, 5998–6008.
19. Girardi, G., & Ergün, A. T. (2013). Systemic risk measurement: Multivariate GARCH estimation of CoVaR. Journal of Banking & Finance, 37(8), 3169–3180.
20. Edmans, A. (2011). Does the stock market fully value intangibles? Employee satisfaction and equity prices. Journal of Financial Economics, 101(3), 621–640.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Global Financial Analytics Research Review
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
