Why an ESG rating is like drowning

Averages can be useful metrics if you want to understand the changes over time in a certain population or data set, like housing prices, consumer confidence, or the indebtedness of corporates. The above story, however, shows – in a rather harsh way – that an average isn’t always the best metric to go by. The increasing use of ESG ratings² seems to suggest that this lesson is not sufficiently understood.

In recent years, sustainability has taken center stage in the financial sector. Triggered by (among others) the European Commission’s Green Deal, to fight the threats from climate change and environmental degradation, in November 2020, the European Central Bank published clear expectations on the way banks should manage climate-related and environmental risks. Another example is the introduction of the Sustainable Finance Disclosure Regulation (SFDR) that requires banks and asset managers to disclose information on how they integrate sustainability risks and potential adverse sustainability impacts in their investment process.

To put these new expectations and regulations into practice, extensive use is being made of ESG ratings. These aim to measure the performance of a company on Environmental, Social, and Governance (ESG) aspects; a bit like how credit ratings measure a company’s Probability of Default (PD). Stemming from the breadth of ESG topics, the number of indicators underlying an ESG rating typically dwarfs the number of indicators used to determine credit ratings. More than 100 indicators is not exceptional. These can range from environmental indicators like the company’s level of greenhouse gas emissions, water usage, and pollution, to social and governance indicators like the number of accidents in the workplace, the use of child labor, and the presence of anti-corruption policies. Not surprisingly, academic literature shows that ESG ratings differ considerably between rating providers.³

In practice, an ESG rating often is the (weighted) average of all individual indicators. This may not give the best indication of the ‘depth of the river’: Tesla Inc. may score rather well on certain environmental aspects, being the frontrunner in electric vehicles. From a social and governance perspective, however, it may not be considered best-in-class. Consequently, averaging the scores over the E, S, and G pillars does not lead to a proper understanding of the sustainability risks involved in this company. Or think of it this way: is child labor (negative score) less of a problem if a company’s employees enjoy freedom of association (positive score)?⁴

The issue also surfaces within the three ESG pillars as a wide range of indicators will be considered to determine a company’s performance for any of the three pillars. As an example, a Kuwait-based oil drilling company will likely score not so well on greenhouse gas emissions, but may obtain a good score on deforestation (given the lack of trees in the desert to begin with). Again, blindly averaging all environmental indicators will not lead to a very useful metric for the environmental performance of the company, nor will it help understanding to what climate risks a company is exposed to.

To truly understand a company’s sustainability risk profile, it is therefore important to assess all (material) indicators individually. Only in this way it is possible to properly address the multifaceted nature of ESG. Those who use only the aggregated ESG rating risk drowning.

Footnotes:

1 “Behar proverbs” – Classified and arranged by John Christian – 1891 – page 120 

2 In this column, an ESG rating is defined as a single (aggregated) measure for the performance of a company with respect to environmental, social, and governance factors. 

3 “Aggregate confusion: the divergence of ESG ratings” – Berg et. al – 15 April 2022 

4 Freedom of association grants employees the right to join a trade union.