Could the use of ESG controversy ratings increase risk for investors?
Ralf Barkemeyer, Christophe Revelli and Anatole Douaud explore the limitations of ESG controversies assessments in their article « Selection bias in ESG controversies as a risk for sustainable investors ».
Their conclusions are echoing other results set forth by the emerging body of literature on ESG assessments:
- The selection of the media sources used by rating agencies to assess ESG controversies is significantly biased.
- Companies headquartered in English-speaking countries are 5 times more likely than the average to be covered as part of a controversy. This probability is 3 times higher for French-speaking ones.
- The authors stress the selection bias studied in the paper is the result of media sources concentration, leading ESG controversy data to lack reliability except to some extent for large France-, UK- or US-based companies.
- They observe that, as with other kind of ESG scores, controversy rating methodologies are black boxes showing shortcomings and high margins of errors when used by finance practitioners.
- In spite of the local and context-specific nature of ESG controversies, media concentration incentivises large companies to adopt a centralised ESG approach around their headquarters' country.
- Further research could add to these results by investigating the factors that lead companies to be covered in the media more than others, and if this kind of selection bias can be quantified for other kind of ESG assessments than controversy ratings.
The distortion of the ESG perception of companies based in countries with poor media coverage resulting from the selection bias described in this article can represent an undetected risk for investors relying on ESG controversy ratings.
The emergence of increasingly efficient technologies to automatically scrape, translate and interpret text data may however have significant consequences for the results put forth in this article in the coming years.