ESG Honestly

Posted 11 June 2024

Where do ESG scores come from?

ESG metrics are used in all kinds of investment decisions. These are often presented just as single numbers, but underneath those numbers is a myriad of judgement calls. Investment managers have a responsibility to be clear about what those are.

If you make investment decisions on behalf of someone else, you should have a good understanding of what is best for them. We generally assume that people are the best judges of what is good for them – which is why investment products are often tailored to people’s risk preferences and time horizons. ESG considerations are no different, fundamentally. People choose ESG products because they want their investments to align with their values. So, ESG investment managers should understand what those values are and make sure the investments are aligned.

There is a special challenge that the ESG investment sector faces here, though. ‘Aligning with your values’ is a much more complicated task than matching up companies to risk profiles. Ideally, we would pick a set of investments that meet some preset and clearly defined ESG criteria – so investors can clearly see what they are getting into and decide for themselves whether it matches their values. But ESG criteria are often complex and multidimensional, which makes this harder to do.

What goes into a score?

Aggregate ESG scores are the clearest example of this. These are supposed to measure companies’ environmental, social and governance credentials and present them as a single number. With the growth of ESG investments, there are now hundreds of different ESG ratings systems, provided mostly by established ratings providers, specialist ESG researchers, or NGOs.

S&P Global’s ESG ratings, for example, are based on extensive company survey data, the answers to which are graded and weighted to reach individual E, S and G scores. The particular questions and weightings are tailored to individual industries on the basis of S&P’s research and what it considers to be the most materially relevant factors. The overall ESG score is a weighted average of the individual scores, where the weightings again vary by sector.

It may surprise you to learn that environmental factors – which usually dominate ESG discussions – tend to be underweighted in S&P’s scores, with an average sectoral weighting of just over 30%. The lowest ‘E’ weighting, assigned to both the Biotechnology and Professional Services sectors, is just 14%.

Whose values?

The above is just one example, but that process – individual company assessment, aggregated according to sector-specific weightings to yield a single number for cross-sector comparisons – is the norm for ESG ratings providers. Quirks in those processes can sometimes lead to strange results, like an oil company with a higher ESG score than a renewable tech company.

The most worrying part of this is that little justification is given for the particular weightings. What counts as good or bad for a “social” score, as well as how important that social score is relative to other factors, is contentious. If investment professionals have to make those calls, the least they should give is a justification for why they did, allowing non-specialists to better decide if they agree or not.

Often these justifications are either hidden in hard-to-decipher formulas and data, or simply not given. In some cases, this is not an oversight but a deliberate decision: ESG ratings and research are mostly made for investors, and research companies often don’t want to show their working out for free.

A clearer path 

The reason why these issues are controversial is that ESG ratings are often used as part of selection criteria for ESG investment products, meaning that investors have a harder time figuring out whether out whether those products align with what they care about. But of course, not all products use these metrics.

Tatton’s Ethical portfolios, for example, are primarily focused on generating returns while minimising exposure to no-go industries – where the exclusions are decided by surveying our clients. We look at how funds deal with ESG ratings only insofar as it affects their screening for those industries. Our process is liable to change if and when our clients’ preferences do.

Other approaches, like impact investing, focus on specific ESG goals that their clients care about and try to improve companies according to that vision. The point isn’t that one process is better than the other simpliciter, but that anyone selling products with potentially contentious labels like “ethical” or “ESG” needs to be transparent about what that label amounts to.

This is the exact thinking underlying new labelling and disclosure rules from the EU and UK. They are not perfect, but are least aimed at standardisation, which should make it easier for investors to see if something billed as ESG matches their own judgements. If you made a value judgement on someone else’s behalf, you should at least be able to tell them why.

Aggregated scores can mask serious underlying issues by balancing out positive and negative corporate behaviors. For instance, a company might have poor employment practices, however could still achieve a neutral overall score due to its carbon neutrality, which could be seen as a positive.

[1] Average weighting in terms of number of sectors (according to GICS sectoral classification) not by companies or market cap. Full breakdown available at https://portal.s1.spglobal.com/survey/documents/CSA_Weights.pdf