Double materiality and data gaps go a long way to explaining the range of results across providers, says Diane Menville, Head of ESG, Scope Group.
The lack of convergence in corporate sustainability scores from different providers continues to make investors uneasy. They want reliable, comparable assessments like credit ratings, but the divergence is deeper rooted than just the lack of good data.
Divergent ESG scores from leading providers reflect a fundamental difference between financial and non-financial ratings.
A credit rating essentially answers one question: what is the probability of default of a debtor?
An ESG score is different. There is no single comparable question. Indeed, investors and issuers today are asking a range of different questions as sustainable investing balloons into a US$40 trillion sector.
The answer for investors is to adjust their expectations and accept that double, if not multiple, ESG scores may be necessary to capture fully the sustainability of their investments
Some corporate ESG scores assess the ESG-related risks companies face such as natural disasters, climate change, regulatory penalties or the danger of stranded assets like oil fields and coal mines. Others measure the impacts of the company itself on the environment and society.
No single score can capture so-called double materiality: the balance of risks and impacts. Take the hypothetical example of a solar-panel manufacturer, paying local staff well but located on low-lying coastline vulnerable to rising sea levels. The company has a good ESG impact score, but the business is danger, hence a poor ESG risk score.
The result is that providers of ESG assessments rely on multiple different criteria, inevitably leading to different results.
The complexity of measuring sustainability, in contrast with creditworthiness, comes into sharper relief when it becomes clear that the ESG risks and impacts of a company extend far beyond its own activities to those of its suppliers and customers: the supply chain.
The so-called upstream and downstream supply-chain factors or, in the terminology of the Greenhouse Gas Protocol, ‘Scope 3’ factors, are crucial for fully assessing ESG risks and impacts. Take other hypothetical examples: either that of an apparel company which has high ESG standards at its own sites but which relies on suppliers with sub-standard practices, or an impeccably run bank which finances environmentally harmful projects. The scores for each would depend heavily on whether the ESG assessment incorporated the company’s supply-chain data or, in the case of a bank, balance-sheet data.
Relative or absolute scores?
If that were not enough, questions regarding ESG have even more nuances. A credit rating is an absolute measure. However, investors seeking to balance risk, reward and sustainability might want a score which shows how well a company rates compared with others in its sector – ‘best in class’ – rather than in absolute terms. ESG risks and impacts vary so hugely between industries – think of coal mining and software development – that absolute scores are not always helpful.
With so many different ESG questions to answer and providers’ use of different methodological approaches, it would be a surprise if results did converge.
There is also the problem of the data. For now, there is no ESG disclosure equivalent to the audited financial statements of companies, based on internationally agreed accounting standards, on which credit rating agencies rely for assessing the likelihood of a borrower’s default on its debt.
Inconsistent corporate self-disclosure is big challenge
Most providers of ESG scores depend on companies’ own ESG measures and disclosure which is patchy at best. One alternative is to use macro-economic approximations of corporate ESG impacts which at least provide broad cross-country, cross-sector indications and a comparable way to measure impacts across companies.
Even when disclosure on the data becomes standardised, other problems will have to be addressed before scores converge. One problem is aggregating E, S and G factors which are measured in such different ways: think of GHG emissions, water pollution, accident rates, and gender inequality. Solutions to this vary from attaching different weights to ESG components – which risks being arbitrary – or calculating their cost so they can be more easily compared.
Standardised ESG reporting is on its way
Standardisation is on its way, notably in Europe. The European Commission’s new Corporate Sustainability Reporting Directive (CSRD), a replacement for the Non-Financial Reporting Directive (NFRD), strengthens rules to ensure that Europe’s biggest companies report reliable and comparable sustainability information. ESG scores using the same methodologies – whether focused on risk or impacts, absolute or relative performance, or other criteria – and standardised data should start to converge, repeating the experience of credit ratings which have converged only over time.
However, it is early days. We would not be surprised to see more institutional investors use standardised data to create their own bespoke ESG scores to meet the different demands and sensitivities of their clients. Divergence might yet be a sign of competitive advantage.
 MIT published a study in August 2019 – “Aggregate Confusion: The Divergence of ESG Ratings” – by Florian Berg, Julian F. Koelbel, Roberto Rigobon which compared the normalised scores of five data providers: Asset4 (now part of Refinitiv), MSCI ESG , RobecoSAM, Morningstar’s Sustainalytics and Moody’s Vigeo Eiris.