Environmental, Social and Governance Practices and Corporate Financial Performance

Abstract

Environmental, Social and Governance (ESG) practices have become increasingly relevant to corporate valuation, risk management, investment decisions, and long-term financial sustainability. However, conventional empirical studies frequently treat ESG as a single aggregated score, despite substantial variation in the financial relevance of environmental, social, and governance dimensions. This study proposes an explainable machine-learning framework for examining how individual ESG pillars and firm-level sustainability indicators influence corporate financial performance.
The research addresses ESG-score heterogeneity by combining structured financial variables with standardized environmental, social, and governance indicators instead of relying exclusively on a provider-generated aggregate rating. Multiple indicators of corporate financial performance, including profitability, market valuation, capital efficiency, and downside financial risk, are considered to capture heterogeneous economic effects. The proposed framework further investigates nonlinear relationships, interaction effects, and sector-dependent thresholds that may remain undetected in conventional linear ESG–performance models.