Some have argued that environmental, social, and governance (ESG) ratings are generally more reliable than credit ratings due to how the raters are paid for their work. In new research, Suhas Sridharan and coauthors find that significant conflicts of interest can arise for ESG raters when they also provide ESG index funds to investors.
Over the past decade, investment strategies based on environmental, social, and governance (ESG) scores have evolved from a niche scheme into a core pillar of asset management. As of 2020, ESG assets under management (AUM) exceeded $35 trillion and are projected to make up a quarter of all professionally managed assets globally by 2030.
Yet as ESG investment vehicles proliferate, so do concerns about the quality and credibility of the ESG ratings that drive these investments. While they are instrumental in shaping fund composition, little is known about how ESG ratings are produced. ESG ratings claim to provide objective summaries of firms’ ESG performance. Such objectivity should lead to relatively similar ratings across different raters for the same firms. However, the fact is that there are persistent inconsistencies across raters.
Noting this, regulators like the U.S. Securities and Exchange Commission (SEC) and the European Securities and Markets Authority (ESMA) have raised doubts about the reliability of ESG ratings. To help inform potential future regulation, our paper seeks to understand the forces that influence the production of ESG information.
A natural point of comparison for understanding ESG ratings is credit ratings. Credit rating agencies face long-standing scrutiny for issuer-pay conflicts, where the firm receiving the rating pays for the rating. Substantial evidence shows that this model leads credit ratings to favor the firm being rated. In contrast, ESG ratings are typically sold under a “user-pay” model, where investors, rather than the firms, pay for the data (ratings). This structure is typically seen as less prone to conflicts of interest, because whereas firms under the issuer-pay system may not return to a rater if the rater rates them poorly (however accurately), an investor just wants the most accurate rating available.
However, that does not mean ESG raters are free from conflicts of interest. Our paper makes the important observation that many ESG rating agencies also operate as index providers, selling benchmarks that fund managers use to track ESG investments. The licensing revenues these index fund providers earn from investors is typically tied to the fund’s AUM—which, in turn, is highly sensitive to past performance. As a result, an index that has high ESG ratings and strong financial performance becomes especially attractive to potential investors. This dual business model—rating companies while profiting from indexes based on those ratings—creates potential conflicts of interest.
This leads us to the central question of our paper: do ESG raters inflate ESG scores for firms with better stock returns when they also license ESG indices? In other words, when an ESG rater stands to profit from constructing an index that both appears sustainable and performs well financially, does it have incentives to issue higher ESG ratings to more profitable firms?
To investigate these potential conflicts, we compare ESG ratings from two raters: a HighIndex rater (MSCI) that both provides ESG ratings and also engages in substantial index-licensing business, and a LowIndex rater (Refinitiv) that focuses primarily on selling ratings data, without much index business. We study over 7,500 observations for 1,449 U.S. firms between 2012 and 2019. Importantly, we compare ratings assigned by both raters to the same firms, thereby holding constant each firm’s ESG fundamentals.
Our study shows thatfirms with better stock returns receive significantly higher ESG ratings from HighIndex raters relative to LowIndex raters. This effect persists even after accounting for differences in rating methodologies and characteristics of the firms being rated. Crucially, this return-related inflation is concentrated among firms eligible for inclusion in ESG indices.
To more convincingly link this pattern with index-licensing activity, we also investigate how stock returns and ESG ratings influence whether a firm is added to (or dropped from) ESG indices. Consistent with our intuition, we find that firms with better stock returns and higher ESG ratings are more likely to be included, and to be given greater weight, in ESG indices. By showing that ESG ratings directly influence index construction, we reinforce our hypothesis that higher ratings for high-return firms boost index performance.
That ESG index fund providers give higher ratings to firms with higher stock returns than those raters that do not provide ESG index funds does not in itself suggest HighIndex raters are manipulating ratings to increase their AUM. Alternatively, we explore the idea that index licensing could give HighIndex raters better information to construct ratings. In particular, the process of researching firms for index inclusion could yield insights that inform ratings that LowIndex raters would not uncover. If this is the case, we would expect HighIndex ratings to provide earlier or more accurate information about changes in a company’s ESG activity.
To test whether HighIndex changes to their ESG ratings reflect better forward-looking ESG insights, we examine how well they anticipate future ESG outcomes. We consider four observable ESG outcomes: regulatory penalties, board gender diversity, board racial diversity, and climate exposure. Our results indicate that HighIndex ratings changes are not predictive of these future outcomes, suggesting the rating changes are not driven by superior ESG intelligence. Moreover, LowIndex ratings do not “catch up” to HighIndex ratings, which we would expect if they could potentially glean valuable insights from private information embedded in HighIndex ratings. Collectively, these findings challenge the argument that index licensing gives raters better information to construct ratings.
Overall, our study offers a powerful caution: even ESG rating agencies using a “user-pay” model are not immune to conflicts of interest. If an agency also profits from index construction, it may face subtle but powerful pressures to favor firms that boost index returns.
For investors, the takeaway is clear: ESG ratings may not be as independent as they appear. Funds tracking ESG indices may be tilted toward better-performing stocks—not necessarily better ESG performers.
Our evidence raises a critical question for regulators and stakeholders: Should the same firms that rate ESG performance be allowed to profit from indices built on those ratings? Should ESG raters be banned from also providing ESG fund indexes or, alternatively, should ESG raters be allowed to provide index funds but be required to build those index funds using the ratings from another rater? This is not just a technical question of methodology. It’s a question of market integrity and investor protection. As ESG investing becomes more central to global capital markets and policy debates—such as the EU’s Sustainable Finance Disclosure Regulation (SFDR) or the U.S. SEC’s proposed ESG fund labeling rules—our paper offers timely evidence of incentive distortions in ESG data.
In sum, our paper underscores the importance of scrutinizing not just what ESG ratings say, but who says it, and why.
Author Disclosure: The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
