AI-augmented ESG investing : Measuring and predicting sustainability impact in capital markets
Mohd Shahid Alimohd.shahid@iilm.eduSchool of ManagementIILM UniversityGurugram, Haryana, 122003, IndiaView full profile → , Ritika Bajaj WaghrayDr.ritikawaghray@vjim.edu.inDepartment of FinanceVignana Jyothi Institute of ManagementHyderabad, Telangana, 500118, IndiaView full profile → , *Afzalur RahmanCorresponding authorafzalur.rahman@woxsen.edu.inSchool of BusinessWoxsen UniversityHyderabad, Telangana, 502345, IndiaView full profile → , M. A. Honest Mohideenhonest.mohideen@gmail.comDepartment of CommerceMotilal Nehru CollegeUniversity of DelhiBenito Juarez Marg, South Campus, New Delhi, 110021, IndiaView full profile → , Refia Wiquarrefia.wiquar@sharda.ac.inSharda School of Business StudiesSharda UniversityGreater Noida, Uttar Pradesh, 201310, IndiaView full profile → , Mohsina Hayatmohsina.hayat@jainuniversity.ac.inSchool of CommerceJain (Deemed to be University)Bengaluru, Karnataka, 560069, IndiaView full profile →
* Corresponding author · click or hover a name for details
- Received:
- 01 Dec 2025
- Published Online:
- 31 Oct 2026
- Article type:
- Research Article
- Language:
- EN
- Article no.:
- JSMS-1754
- Pages:
- 1151–1160
Abstract
This paper examines whether artificial intelligence (AI) improves the information that environmental, social and governance (ESG) ratings provide to investors. Sustainalytics ESG risk ratings, market indicators and natural language processing (NLP) features from business descriptions are combined for 429 S&P 500 firms (81,843 firm-months, 2010–2026). Random forest, XGBoost and LSTM models tested out of sample (January 2024 – September 2026) identify firms entering high-volatility months far more accurately than ESG ratings alone (AUC 0.82–0.84 versus 0.50; p < 0.001). However, no model predicts returns better than the ratings-only benchmark, and AI portfolio Sharpe ratios do not differ significantly from that of a best-in-class ESG portfolio (0.81 versus 0.84). AI therefore adds most value as a forward-looking risk-monitoring tool rather than as a source of excess returns.
Keywords
Subject Classifications
References
[1] A. Amel-Zadeh and G. Serafeim, “Why and how investors use ESG information: Evidence from a global survey,” Financ. Anal. J., vol. 74, no. 3, pp. 87–103 (2018).
[2] F. Berg, J. F. Kölbel, and R. Rigobon, “Aggregate confusion: The divergence of ESG ratings,” Rev. Finance, vol. 26, no. 6, pp. 1315–1344 (2022).
[3] M. A. Delmas and V. C. Burbano, “The drivers of greenwashing,” Calif. Manage. Rev., vol. 54, no. 1, pp. 64–87 (2011).
[4] K. V. Lins, H. Servaes, and A. Tamayo, “Social capital, trust, and firm performance: The value of corporate social responsibility during the financial crisis,” J. Finance, vol. 72, no. 4, pp. 1785–1824 (2017).
[5] A. Banu, A. Rahman, and M. Murari, “A study on financial performance of Nifty 50 Indian companies with strong corporate governance,” Journal of Information and Optimization Sciences, vol. 46, no. 8, pp. 2403–2412 (2025), doi: 10.47974/JIOS-2055.
[6] V. D’Amato, R. D’Ecclesia, and S. Levantesi, “ESG score prediction through random forest algorithm,” Comput. Manag. Sci., vol. 19, pp. 347–373 (2022), doi: 10.1007/s10287-021-00419-3.
[7] S. Gu, B. Kelly, and D. Xiu, “Empirical asset pricing via machine learning,” Rev. Financ. Stud., vol. 33, no. 5, pp. 2223–2273 (2020).
[8] R. G. Eccles, I. Ioannou, and G. Serafeim, “The impact of corporate sustainability on organizational processes and performance,” Manage. Sci., vol. 60, no. 11, pp. 2835–2857 (2014).
[9] P. Krueger, Z. Sautner, and L. T. Starks, “The importance of climate risks for institutional investors,” Rev. Financ. Stud., vol. 33, no. 3, pp. 1067–1111 (2020).
[10] F. Rizwan and A. M. A. Jenita, “An analysis of impact of CSR practices on investment behaviour,” J. Stat. Manag. Syst., vol. 25, no. 5, pp. 1011–1020 (2022), doi: 10.1080/09720510.2022.2083828.
[11] R. Gibson Brandon, P. Krueger, and P. S. Schmidt, “ESG rating disagreement and stock returns,” Financ. Anal. J., vol. 77, no. 4, pp. 104–127 (2021).
[12] D. M. Christensen, G. Serafeim, and A. Sikochi, “Why is corporate virtue in the eye of the beholder? The case of ESG ratings,” Account. Rev., vol. 97, no. 1, pp. 147–175 (2022).
[13] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discov. Data Mining (KDD), pp. 785–794 (2016).
[14] S. Y. In, D. Rook, and A. Monk, “Integrating alternative data (also known as ESG data) in investment decision making,” Glob. Econ. Rev., vol. 48, no. 3, pp. 237–260 (2019).
[15] P. Dugar, “S&P 500 ESG risk ratings” [Data set], Kaggle (2023), doi: 10.34740/KAGGLE/DS/3660201.
[16] Johnbrick123, “sp500-data: S&P 500 daily price history,” GitHub repository, (2026).
[17] Board of Governors of the Federal Reserve System, “3-month Treasury bill secondary market rate, discount basis (TB3MS),” FRED, Federal Reserve Bank of St. Louis, (2026).
[18] L. Breiman, “Random forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32 (2001), doi: 10.1023/A:1010933404324.
[19] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Comput., vol. 9, no. 8, pp. 1735–1780 (1997), doi: 10.1162/neco.1997.9.8.1735.
[20] F. X. Diebold and R. S. Mariano, “Comparing predictive accuracy,” J. Bus. Econ. Stat., vol. 13, no. 3, pp. 253–263 (1995), doi: 10.1080/07350015.1995.10524599.
[21] W. K. Newey and K. D. West, “A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix,” Econometrica, vol. 55, no. 3, pp. 703–708 (1987), doi: 10.2307/1913610.
[22] C. Memmel, “Performance hypothesis testing with the Sharpe ratio,” Finance Lett., vol. 1, no. 1, pp. 21–23 (2003).
