ESG and Financial Performance in the Eurozone: A Cross-Industry Analysis

This paper focuses on analysing the relationship between Environmental, Social and Governance (ESG) ratings and financial performance in the Eurozone across different industries. While some theories show that companies should naturally do Corporate Social Responsibility (CSR), for others, CSR should be avoided as much as possible. The empirical literature does not confirm any of the CSR theories as there is high variety of results in the study of ESG and financial performance, showing that there is no clear answer regarding the link that they can take. This paper focuses on the 2010 - 2017 period to avoid the impact of the financial crisis on the data and uses 256 companies classified in different industries through the International Classification Benchmark. Among all ESG rating providers, Thomson Reuters’ is the one chosen for ESG ratings. Regarding financial performance, Tobin’s Q and ROA are chosen for market performance and accounting performance respectively. The main model of this study regresses ESG to financial performance and is completed by a second model which regresses individually the ESG pillars: Environmental, Social and Governance. To analyse this relationship, a Fixed Effects regression with Driscoll and Kraay standard errors is used as this method takes into account all the different characteristic of the data seen through various tests; necessity to use FE, heteroskedasticity and cross-sectional dependence. This paper shows that there is a highly significant positive relationship in the Eurozone between ESG and market performance, and more specifically a highly positive relationship between Social, Governance ratings and market performance. Regarding accounting performance, this study shows that there is no relationship in the Eurozone, but positive and negative relationships do exist when looking at specific industries. The industry analysis enables to show two important characteristics in the Eurozone. First, even if some industries are related, they can still have statistically significant ESG ratings of different mathematical signs, thus showing the interest of looking at them individually. Second, ESG pillars of a same industry can have different mathematical signs of relationship with financial performance. Finally, the sensitivity test is done with different ESG ratings, showing that the model is fairly robust