Abstract

This study empirically investigates the financial impact of ESG factors and market dynamics on service sector firm excess returns using a 2012-2021 panel dataset of 14 firms. Panel regression and time series analysis, along with machine learning models, were employed. Findings indicate that traditional Fama-French factors significantly influence returns, while a higher aggregated ESG score unexpectedly correlates negatively with excess returns. Market level ESG factors show dynamic interdependencies with market returns but no Granger causality. Machine learning models demonstrate high predictive accuracy for future return direction, with financial and ESG features as key predictors. These exploratory findings are useful for understanding complex ESG-financial relationships in the service sector, guiding nuanced corporate sustainability strategies and investment decisions, and contributing to the ongoing debate on ESG materiality.

Keywords: Service sector, ESG, Financial performance, Stock returns, Fama-French factors, Panel data, Machine learning, Stakeholder theory, Asset pricing, Sustainable finance