An Explainable Modeling Method for Analyzing ESG–Finance Interactions in Strategic Decision-Making

Environmental, Social, and Governance (ESG) is a well-established criterion for corporate social performance, but its financial materiality remains elusive. This study aims to unpack the nuanced and often contentious relationship between ESG and corporate financial performance, focusing not just on if ESG matters but more critically, on how, when, and for whom it becomes a significant value driver. Using a Random Forest model and SHapley Additive exPlanations (SHAP), we analyze a large panel dataset of publicly listed firms across Asia, the United States, and Europe from 2017 to 2024. The results show that firms with high ESG ratings exhibited greater resilience during the COVID-19 crisis period, confirming ESG's role as a risk mitigation factor. Moreover, our findings refine stakeholder theory by showing that satisfying stakeholder interests is not uniformly rewarded; rather, its financial benefits are conditional on both industry context and existing financial strength. These findings provide a data-driven rationale for strategic ESG investments.