Predicting Financial Success Through ESG Metrics: Leveraging Machine Learning and Explainable AI

This study examined the relationship between Environmental, Social and Governance (ESG) performance and financial outcomes via unified combination of machine learning and deep learning frameworks. A global and publicly available dataset of firms was analyzed by combining ESG indicators with financial metrics such as revenue, profit margin and market capitalization. Machine learning models such as XGBoost, LightGBM and TabNet, a state of the art are applied for both regression and classification analysis throughout the dataset. The results have shown strong predictive performance with classification accuracy reaching 0.88 and AUC values above 0.90 while the regression models has achieve R2value above than 0.90 and RMSE as low as 0.34. Explainable AI methods such as SHAP and LIME have identified the market capitalization, water usage and carbon emissions as the most influential key drivers in the domain while governance remaining the leading ESG factor across most settings, while cross domain evaluation have further confirmed robustness with domain level accuracies of 0.69 and 0.78. Scenario simulations also indicated that targeted ESG improvements such as a 10% increase in governance scores produced measurable although modest gains in profitability. These findings were then incorporated into a prototype decision support system which allowed managers to test ESG strategies in an interactive way. Overall the framework delivers robust interpretable and practical insights that align sustainability practices with financial performance.