Navigating in the ESG score jungle- A cross-sectional approach to determine the ESG risk factor

Abstract

This thesis examines the relationship between ESG scores and yearly excess return between 2010 and 2020 on the S&P 500 Index. With a solid theoretical background regarding investor preferences, we ask whether investors accept lower returns for holding greener assets. Our method is a cross-sectional approach, using pooled time-series regressions and Fama-MacBeth regressions, where we seek to determine the ESG risk score factor. We find significant evidence that ESG scores have a negative relationship with yearly excess return in all our regressions when controlling for other return predictors and the Sin Stock anomaly. This relationship holds for the overall ESG score and the separate ESG pillar scores, Environmental, Social, and Governance. Our results prove to be consistent with previous research regarding ESG-motivated investors. We found inconsistent results with previous research regarding the Governance pillar score, arguing that the Governance pillar score may not be an appropriate proxy. Our results remain consistent while conducting further robustness tests with clustering on the sector level.