Earlier this week, an article in The Wall Street Journal compared ESG scores of companies like Exxon and Tesla from multiple ratings agencies, including MSCI and Sustainalytics, in order to highlight the lack of standardization in evaluating companies using current environmental, social, and governance (ESG) metrics.
So what was the conclusion when these ratings were examined from the ESG data industry, which provides environmental, social, and governance data that can’t be found on a balance sheet?
Exxon was rated more highly overall by two of the three biggest ESG ratings firms.
The article concluded that “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.”
It’s easy to understand why media outlets and many traditional investors think ESG investing is missing the mark.
But the fact is, for all of the careful research and correct conclusions, this particular article is assessing an outdated version of ESG – state of the art in years past, but far from the real value that today’s ESG and intangible factor investors have discovered.
While the article made several strong and interesting points, here is what the author got right and wrong from our perspective.
Not quite right
Comparing Exxon to Tesla is comparing apples and oranges. Professional investors wouldn’t compare the EBITDA of a startup carmaker to that of an integrated oil exploration multinational. Neither would a sophisticated investor consider ESG factors without comparing a company to its peers in its own industry.
Luckily, ESG factors that are financially material for each industry have been carefully constructed by the Sustainability Accounting Standards Board, or SASB. Established in 2011, SASB is an independent, non-profit, private-sector standards setting organization whose board includes a former chair of the SEC, former FASB Chairman, and at one time, Michael Bloomberg. Its standards are a public good, much like FASB’s standards are.
Truvalue Labs provides financially material ESG data tailored to SASB’s standards for each industry, and that allows true apples-to-apples comparisons like this analysis of Tesla and GM.
Right on the money
What did the Journal get right? ESG ratings firms can’t agree on many companies’ scores.
Why not? At its core, it’s hard for human experts of any discipline to agree on decision-making methods, and even when methods are shared, opinions vary.
That problem, called inter-rater reliability, is detailed in this post on Thomson Reuters’ Lipper Alpha blog. And it can be seen in the Journal’s excellent charts in the article, which show how little raters agree about Tesla.
There are three main dimensions of this: how definitional categories are constructed, how these are measured and how weightings among the categories are calculated. Add to this there is a strong tendency among traditional raters to focus on problems/scandals—negative behavior; as well as a significant time lag between ratings and events.
The problem of human raters is a serious one for investment professionals, and solving that problem is the main mission of Truvalue Labs, which brings the advanced computer science of machine learning to bear on mining big data for both negative and positive company events and analyzing them with machine learning, which means that they are analyzed consistently, transparently and objectively.
Truvalue Labs is proud to partner with SASB, State Street, and other serious investment players in this effort to build a stable, research-backed framework for ESG and intangible investing.
Beyond an objective dataset based on SASB materiality, Truvalue Labs also takes a unique approach in that continuously monitoring the information flow in real-time allows for multiple updates daily, moving at the speed of markets, unlike ESG ratings firms that lag events by a month, quarter or year.
Additionally, results from a ten-year backtest of the Truvalue Labs data shows alpha can be found from ESG signals, if they’re the right signals. Dr. Stephen Malinak, Chief Data & Analytics Officer at Truvalue Labs, found upside performance using ESG data.
Investors need and increasingly want timely, transparent data, based on an open, consistent standard that is focused on materiality. Truvalue Labs covers those bases in ways that old school ESG raters and rankers do not.
Featured post originally appeared on the Tabb Forum (subscription required) as “Building a Sustainable Framework for ESG Investing.”