ESG in times of Crisis: Evidence from the COVID-19 Pandemic

Nowadays, there is a growing tendency towards environmental, social and governance (ESG) incorporation in investment decisions, which is reflected in firm’s market returns. The role of these factors in times of crisis is still debatable and underexplored with mixed evidence from the COVID-19 market crash. The pandemic has been an unprecedented exogenous shock, providing a scientific opportunity for in-depth research. Therefore, this paper is aimed to contribute the more accurate understanding which ESG factors are most rewarded by European investors in times of crisis on the example of the COVID-19 pandemic. The examination is based on ESG database of Refinitiv, but MSCI and S&P Global data is also explored for robustness purposes. The main dependent variable in our study is crisis-period stock return over the “collapse” period. It involves three measures of stock returns: raw, CAPM-adjusted and market-adjusted stock returns. Existing academic literature is largely devoted to US markets, while our dataset consists of a new sample of 8 European countries, the choice of which is motivated by availability of high-quality ESG metrics by Refinitiv and liquid financial markets. The final sample has 1157 European companies. The empirical analysis shows that general ESG score, environmental score and social score have been irrelevant for European investors during the pandemic-induced market crash. Nonetheless, there is statistically significant evidence that corporate governance performance has a positive and economically sizeable effect on crisis-period stock return performance. The relevance of ESG rating differs for firms with various levels of financial constraints: the positive effect of corporate governance on crisis-period returns is more pronounced for firms with low cash holdings. Exploring the role of specific ESG components, product responsibility, corporate governance management and shareholders score have been particularly rewarded by European investors. The study involves the two-stage least squares model which produces qualitatively similar conclusions, a strong instrument is found which is represented by average governance score for country-industry subset excluding the firm in focus. Our findings are robust to controlling for industry and country fixed effects, various firm characteristics, different dates, model specifications, return measurements, ESG providers and estimation methods.