Exploring the Effects of ESG Scores and Carbon Emissions on Abnormal Stock Returns: A Two-Step Approach with Random Forest and Panel Regressions

As sustainable finance becomes more prevalent and climate change concerns more urgent, the role of non-financial information has never been more important. However, identifying the influential indicators remains a formidable challenge. Our research explores the impact of ESG scores and carbon emissions on abnormal stock returns, utilizing a dataset that includes granular Refinitiv ESG scores and CO2 emission data from 4,513 global publicly listed companies across two decades (2002–2022). We employ a two-step approach that combines machine learning and fixed effects panel regressions. The results show that all ESG scores analyzed in our study, except the Controversies Score, exhibit non-positive impact on abnormal returns. A one-point increase in the main ESG Score is associated with a 0.2% decrease in abnormal returns. Furthermore, all carbon emissions metrics have a non-positive impact on abnormal returns, however, the effect varies considerably depending on the metric used. A 1% increase in annual CO2 emissions correspond to a 0.027% decrease in abnormal returns. Our results suggest that superior carbon performance might be a more effective factor than higher ESG scores in achieving higher abnormal returns. The implications of our study extend beyond the financial markets, highlighting the importance of carbon emissions within the corporate sustainability initiatives and environmental policy.