By James Mackintosh
Is Elon Musk's electric-car maker Tesla the best, the worst or a merely middling performer on environmental issues? Is Warren Buffett's Berkshire Hathaway one of the worst-governed big U.S. companies? Is General Motors one of the most socially aware businesses, or one of the least?
Investors picking a scoring system for ESG -- environmental, social and governance -- issues facing companies can have any of those outcomes.
The differences are easy enough to understand if you dig into the details. Yet, they show just how difficult it is to take a simple approach to ESG investing, a style that is becoming ever-more popular. In addition to the billions of dollars of exchange-traded funds based on ESG indexes, increasingly fund managers are being pushed to produce portfolios with better ESG ratings, encouraged by public mutual-fund ESG scores.
The problem here isn't the ESG ratings, but that they are used as though they were some sort of objective truth. In reality they are no more than a series of judgments by the scoring companies about what matters -- and investors who blindly follow their scores are buying into those opinions, mostly without even knowing what they are.
To illustrate these differences, we can dig into the scores given to five big companies by FTSE Russell, MSCI and Sustainalytics, all used for ESG indexes and by institutional investors. The companies are Tesla, Berkshire, oil major Exxon Mobil, Google-owner Alphabet and carmaker General Motors.
Perhaps the biggest surprise is Tesla, ranked by MSCI at the top of the industry, and by FTSE as the worst carmaker globally on ESG issues. Sustainalytics puts it in the middle.
The explanation comes down to what is measured, and how the measurement is affected by disclosure.
MSCI gives Tesla a near-perfect score for environment, because it has selected two themes as the most important for the car industry: the carbon produced by its products, and the opportunities the company has in clean technology.
FTSE gives Tesla a "zero" on environment, because its scores ignore emissions from its cars, rating only emissions from its factories (to confuse things further, FTSE's separate "Green revenue" score gives Tesla 100%).
Tesla highlights another major difference in scoring: what to do when a company doesn't disclose. FTSE says it has to assume the worst if no information is provided about issues that matter to a company -- and that giving it the worst score encourages more disclosure.
Tesla, which discloses little about its operations compared with other automakers, suffers from FTSE's approach, particularly on social issues (where all three graders anyway give it low scores on how it treats workers). MSCI is more generous, assuming that if there's no disclosure the company operates in line with regional and industry norms. Sustainalytics declined to explain its methodology, but it gives points for disclosure of policies -- again, Tesla suffers -- as well as low scores for issues where there is too little disclosure to calculate, such as Tesla's renewable-energy use.
Berkshire suffers on disclosure too, again being zeroed by FTSE on both environment and social scores.
Another problem is how to put the separate environmental, social and governance scores together. Should a highly polluting company be able to offset that by having great governance and treating workers well? Sustainalytics ranks Exxon top of the five companies overall, because it puts a 40% weight on social issues, where Exxon does well thanks to strong policies for its workers, supply chain and local communities. MSCI ranks Exxon fourth of the five in part because it puts a 51% weight on environment and only 17% on social issues.
The three create their scores differently, too. MSCI selects -- using rules -- a small number of factors that matter to each company, making each important to the overall score. FTSE includes a broader set of factors, but is still rules-based. Sustainalytics also has a broad set of factors, and uses analyst judgments for some of its assessments. In all three cases the design of the rules and scoring system makes a big difference to the outcome.
And sometimes the assessments simply differ. MSCI puts Alphabet in the bottom quartile of its industry for the subcategory of corporate governance thanks to controlling shareholders and related-party transactions, although its overall governance score is lifted by a strong score on "corruption and instability" issues. FTSE takes the opposite approach, putting Alphabet in the top half of its peer group for governance, and says it is held back in part by a weak score for anticorruption assessments and training, as well as tax disclosure.x`
Investors should not treat ESG scores as settled facts to be used on their own, but as potentially worthwhile analysis that needs to be understood before being acted on. The thick ESG reports behind the scores offer useful detail about the policies and controversies around each business. But just as with financial accounts, investing without understanding is unlikely to deliver what you want.
Write to James Mackintosh at James.Mackintosh@wsj.com