Adriana Z. Robertson & Sarath Sanga1Adriana Z. Robertson is the Donald N. Pritzker Professor of Business Law at the University of Chicago Law School. Sarath Sanga is a Professor of Law at the Northwestern University Pritzker School of Law and the William Nelson Cromwell Visiting Professor of Law at Harvard Law School. We thank Jill Fisch, Kate Judge, Elizabeth Pollman, Christina Skinner, David Weisbach, and the University of Chicago Law Review Online team for valuable suggestions and discussions. This Essay benefited from comments by workshop participants at the 1st Annual Women in Law & Finance Conference. Talla Khelghati provided exceptional research assistance. All errors are our own.
* * *
What does it mean for a fund to deliver ESG results to its investors? Money has poured into ESG funds in recent years. But scholars, commentators, and regulators have all expressed concerns that ESG fund promises are too vague and ESG fund disclosures are too meager to enable investors to make informed decisions.
In this Essay, we identify a core challenge to evaluating ESG funds: the problem of using company-level ESG characteristics to construct a portfolio-level ESG score. Commentators and market participants have largely ignored this aggregation problem and instead relied on rules of thumb like “buy the ESG leaders” as ad hoc guides for ESG investment strategies and ESG fund performance evaluation. Such strategies, however, fail because they are often inconsistent with the purpose of pursuing ESG objectives in the first place.
Our main claim is that an ESG fund must establish a consistent link among the fund’s stated ESG purpose, its ESG investing strategy, and—crucially—the portfolio-level ESG metric. Otherwise, it is not possible to coherently evaluate the fund. We propose a practical approach to constructing portfolio-level ESG metrics and explain how, in light of our analysis, ESG fund managers and the SEC should act. We explain, for example, how fund managers can ensure consistency between their ESG theory and investment strategy and why standardizing ESG metrics would—contrary to popular intuition—make it more difficult to evaluate ESG funds. Our analysis also suggests that the direct impact of the SEC’s proposed anti-greenwashing and ESG fund disclosure rules may be nil, while the indirect and unintended benefits of those rules may be substantial.
Interest in ESG investing2We recognize that important nuances distinguish ESG from the kindred concepts of socially responsible investing (SRI) and corporate social responsibility (CSR). We focus on ESG because the term and the concept has largely overtaken and subsumed SRI and CSR in the academic and industry discourse. has exploded over the last two decades. The idea that firms and investors should expressly consider environmental, social, and governance factors has moved from a relatively niche community to the mainstream. A dizzying array of ESG mutual funds have sprung up to meet investor demand for ESG investment products. Even beyond specialized ESG funds, asset managers, with a few notable exceptions, have also joined the cause. This includes some of the world’s largest money managers. In his 2021 letter to CEOs, Larry Fink, CEO of BlackRock, went so far as to commit to “supporting the goal of net zero greenhouse gas emissions by 2050 or sooner.” BlackRock is hardly alone in this: Vanguard, State Street, Fidelity, and many others have expressed their commitment to various forms of ESG.
Notwithstanding this excitement, scholars and commentators have raised significant concerns about these product offerings, often revolving around so-called “greenwashing” that “misleads investors or stakeholders, inhibits corporate accountability, or crowds out other concepts and proposed solutions.” The financial press is full of reports of greenwashing and other dodgy behavior by funds purporting to be focused on ESG issues.
This, in turn, has drawn the attention of regulators. Under Chair Gary Gensler, the SEC has taken an active interest in policing ESG disclosure in its investigation, enforcement, and rulemaking capacities. While the best-known example of the latter is the proposed “climate rule,” which would mandate issuer-level disclosures related to carbon emissions, when it comes to ESG mutual funds, two other proposed rulemakings are more consequential. The first, colloquially known as the proposed “ESG Names Rule,” would regulate the names of funds marketing themselves as having ESG characteristics, while the second, colloquially known as the proposed “ESG Fund Disclosure Rule,” would require ESG funds to include certain disclosures in their regulatory filings. If adopted, these two rules would represent a significant intervention by the SEC in the ESG funds space.
While these developments represent a welcome step forward, on their own they are not enough to cut through the existing confusion. The reason for this is simple: when it comes to ESG, it’s impossible to characterize a portfolio without first articulating a theory of why you are considering ESG in the first place. We demonstrate this by way of an example in Part I, and then discuss the state of play in ESG investing in Part II.
In Part III, we offer a framework for holding ESG funds accountable to their investors. Our framework is premised on one core claim: investment advisers and funds should articulate their theory of the incremental benefits of the ESG characteristic of interest. In the language of economics, this is the marginal social value of the ESG characteristic, which can be increasing, decreasing, or constant. As we demonstrate, this theory determines how the fund or adviser should form a portfolio given the ESG characteristics of the available companies, as well as how it should report on the portfolio’s ESG characteristics. In other words, the theory is simply the advisor’s investing thesis as applied to the relevant ESG characteristic. We leverage this analysis in Part IV to provide best practices for ESG fund managers and a critique of the SEC’s proposed rules on ESG fund names and disclosures.
I. A Simple Example
Consider an investor whose ESG priority is the representation of women on corporate boards. This example is hardly hypothetical: the representation of women on corporate boards has been a major ESG focus over the last several years. It first drew mainstream attention with State Street’s “Fearless Girl” campaign in 2017, before taking off much more broadly. Along with carbon emissions, the representation of women on corporate boards is, to date, one of the clearest success stories for ESG investing.
Suppose that our investor is given the choice between two otherwise identical portfolios. The first, called “High/Low,” is a 50/50 blend of two companies, each with an eight-person board. One company has two women on its board, while the other has none. The second, called “Average,” is also a 50/50 blend of two companies. Like High/Low, each of the companies has an eight-person board. In Average, however, both companies have one female director. Because the portfolios are otherwise identical, this implies that they have the same financial risk and return characteristics.
Which portfolio does our investor like better? The answer is that it depends. Specifically, it depends on the reason she is concerned with the representation of women on corporate boards to begin with. For example, perhaps the investor believes that some representation is much better than none, so going from no women to one woman is more important than going from one to two. In other words, this investor believes that the marginal value of adding additional women is decreasing—the first is more valuable than the second, who is in turn more valuable than the third. This investor is extremely unhappy that one of the companies in the High/Low portfolio has no women on its board, and she will strongly prefer the Average portfolio.
Alternatively, the investor might think that one woman on a corporate board, while better than none, is not nearly as effective as two. This investor believes that a single woman on the board smacks of tokenism—a concern that goes away when there are two women. In other words, she thinks that the marginal value of adding additional women is increasing; it’s not that the first isn’t valuable, it’s just that the second is much more valuable than the first. This investor strongly prefers the High/Low portfolio. She is willing to tolerate the fact that one of the companies has no women in order to get the benefits of two women at the other company.
Finally, perhaps the investor just cares about total representation. This investor doesn’t care how women are allocated between the companies in her portfolio; she just cares about the total. This investor thinks that the marginal value of women is constant: she’s just as happy to see an additional woman on the board of any company, regardless of how many it already has. This investor will have no preference between the two portfolios.
Each of these investors presents a distinct and coherent view of the value of ESG. These views generate different preferences over ESG funds because they differ over the marginal social value of the ESG characteristic. Whether one of these investors is “right” in some objective sense is irrelevant. As with the rest of the securities regulatory regime, we have very little to say about what any particular investor should buy.
Instead, what matters is how a fund manager should act in light of such heterogeneous investment preferences. To adequately respond to these preferences, a fund would have to (1) commit to a theory of marginal value, (2) disclose that theory to potential investors, (3) make investments consistent with that theory, and finally (4) adopt and report a fund-level ESG metric that is consistent with that theory. Indeed, as we will argue below, a fund that merely commits to, say, “prioritizing investments in companies with female directors” actually commits to very little, if anything, and in any case provides no objective mechanism for investors to evaluate the fund’s ESG performance. This is our main claim, and we unpack it in detail in Part III. Before doing so, we first summarize the current state of ESG investing.
II. The State of Play in ESG Investing
The core problem illustrated by the example in Part I arises because the investor is choosing not among investments in individual companies but among aggregations of companies—i.e., among portfolios. To choose among portfolios, the investor must have a theory of trading off ESG characteristics between firms. For this reason, below we will draw a distinction between evaluating ESG at the issuer level (and thus holding specific companies accountable for their ESG promises) versus evaluating ESG at the fund level (and thus holding ESG funds accountable for their ESG promises).
A. Issuer-Level Challenges
Despite its popularity with investors, ESG investing faces major challenges. For our purposes, the two most important are the challenge of defining ESG at the issuer (i.e., company) level and the challenge of measuring issuer-level ESG.
There seem to be almost as many conceptions of ESG as there are people talking about it, so the first challenge in constructing issuer-level ESG metrics is defining the concept. The most basic distinction is between “value” (or “risk-return”) based ESG and “values” based ESG (also known as “collateral benefit” ESG).
A value-based ESG investor is concerned with one or more ESG characteristics because she believes that they are material to the financial terms of her investment. This can generally occur in two ways. First, the investor might believe that an ESG characteristic captures material financial risk that is not otherwise being priced by the market. Alternatively, she might believe that an ESG characteristic is a signal that a company is well (or poorly) positioned to earn above market (or below market) returns going forward. Again, this implies that the investor believes that the market has not already impounded this information into prices.
In contrast, a values-based ESG investor cares about the ESG characteristics of her portfolio independently of any financial benefits of those characteristics (either in the form of increased financial return or reduced financial risk). Her reason for doing so is therefore not based on her own private interests, but rather because she cares about some social value. For this to have any content, it follows that she must be willing to sacrifice some risk-return benefit—i.e., to sacrifice some financial return or take on some additional financial risk—for the sake of those characteristics. Naturally, this need not mean that she is willing to forego financial returns entirely, nor does it mean that she is willing to take on an unlimited amount of financial risk. It simply means that her willingness to make this sacrifice is nonzero, and perhaps even nontrivial.
Closely related to traditional values-based ESG is impact-based ESG. Whereas most traditional values-based ESG investors focus on a company’s characteristics at the time of the investment—perhaps under a theory that companies that do well on these metrics should be rewarded with a lower cost of capital—impact-based ESG investors focus on improving the ESG characteristics of their portfolio companies. Impact-based investors often invest in companies with poor ESG characteristics, on the theory that these companies have the most room for improvement. While both values- and impact-based investors often have similar underlying values, these two approaches can lead to very different investing strategies and very different approaches to stewardship.
Even once the definitional issue is resolved, we must still contend with a challenging measurement problem. There are hundreds of ESG ratings providers. These include some well-known names in the financial data space, including S&P, MSCI, ISS, Refinitiv, and Morningstar. On top of these, there are a plethora of other ESG ratings providers, each with their own methodology (both conceptually and practically).
Ratings from different providers are famously inconsistent. One prominent Article studied several leading ESG ratings and demonstrated that the correlations between ESG ratings across providers were 54% on average. Some have interpreted the proliferation of different ratings as evidence that there is no “there there.” Others have articulated concerns about rating shopping—where issuers or asset managers might cherry pick an ESG rating that provides their desired result.
Impact-based ESG investing also raises its own measurement challenges. Rather than requiring a measure of the relevant ESG characteristic(s) at the time of the investment, impact-based ESG investing requires an evaluation of the change in the relevant characteristic. Moreover, to be true to the concept of impact-based ESG investing, we would really want a measure of the change in the relevant ESG characteristic that was caused by the investment. This measurement challenge adds an additional layer of complexity to evaluating impact-based claims.
B. Portfolio-Level Challenges
Even once we solve these issuer-level challenges, we still have to grapple with the question of how to aggregate from the issuer level to the portfolio level. As the example in Part I makes clear, moving from the issuer level to the portfolio level adds a layer of complexity that has so far been overlooked.
The example in Part I was designed to obviate the traditional concerns about ESG metrics. This is a setting in which measuring the ESG characteristics is easy. Unlike, say, quantifying how inclusive a company is, or how ethical its supply chain is, all we have to do is count the number of women on the corporate board. This is information that can easily be obtained from the company’s website, or from its regular SEC filings.
It is also a setting where only one ESG characteristic—women on corporate boards—is relevant. We therefore avoid the challenge of combining different ESG metrics and thus trading off, say, women on boards, labor practices, and carbon emissions. As a result, the example cleanly illustrates why the problem of aggregating ESG characteristics to the portfolio level is distinct from the problem of issuer-level ESG metrics. Even when the latter is solved, the former remains a problem.
III. The Missing Link: Marginal Social Value
Without a clear connection between its underlying theory of ESG, its ESG investing strategy, and the ESG metrics it uses, our investors in Part I have no way to know what a “women in leadership” fund is actually doing, let alone whether what the fund is doing is consistent with the investor’s preferences.
So, what is a values-based ESG fund manager to do? In our view, ESG fund managers must articulate a theory of the marginal social value of the underlying ESG characteristic and adopt an investment strategy that is consistent with that theory. At the end of the year, it should report a portfolio-level ESG metric that is also consistent with that theory to the investors so that they can evaluate the fund’s performance at achieving its goal. This claim is both analytical and normative: doing otherwise leads to incoherent strategies that do not achieve investors’ ends. While it seems obvious, to our knowledge, this is a point that has been largely overlooked in the debate thus far.
A. Decreasing, Constant, and Increasing Marginal Value
The phrase “marginal value” is used to describe the incremental value that a person places on a given thing. When marginal value is decreasing, it means that the person values incremental amounts of the thing in question (a product, attribute, or whatever else) more when she has less of it to begin with. For example, a very busy person might value an extra day off more than a person who already has a lot of free time. This doesn’t mean that the latter person doesn’t value an additional day off; it just means that the former person values it more.
One example of an ESG characteristic that is likely to have declining marginal value is labor practices. For example, the benefit from moving from a reliance on forced labor in a corporate supply chain to the use of sweatshops is almost certainly larger than the benefit from moving from sweatshops to “fair trade” levels of labor practices. To be absolutely clear, this is in no way an argument in favor of sweatshops, nor does it represent a statement of any kind about the value of fair-trade practices. Rather, it is a relative statement about the abhorrence of forced labor.
Another example is the first investor in Part I: an investor who believes that going from zero women on a corporate board to one woman is more valuable than going from one woman to two is an investor who believes that the marginal value of adding women is decreasing. Other things equal, if she were given the choice between adding a woman to the board of a company that has no women and adding a woman to the board of a company that already has one, she would prefer the former.
Marginal value can also be constant, meaning that a person values all incremental amounts of that thing the same way. In the ESG context, the most natural example of this is carbon emissions. Other things equal, an investor who cares about her portfolio’s carbon footprint is probably focused on the total carbon emissions in her portfolio. Each ton of carbon emitted has roughly the same impact on the environment, regardless of the company emitting it.3One exception to this is so-called “hotspots.” See generally David Weisbach, Regulatory Trading, 90 U. Chi. L. Rev. (forthcoming 2023). Other things equal, then, an investor would be equally happy about a reduction in emissions in any part of her portfolio. The portfolio’s ESG measure—at least as it pertains to carbon emissions—is then simply the weighted average of the emissions of the constituent issuers.
The third investor in Part I, who believed that the value of women on corporate boards derives primarily from increasing the number of women at the board level, could be another example of an investor with a theory of constant marginal value. Such an investor might argue that total representation of women is the only thing that matters and that it is not worth sacrificing representation for the sake of a more uniform distribution of women across boards.
Finally, marginal value can also be increasing, meaning that increases are more valuable the more one already has. The second investor in Part I, who believes that a critical mass of women is needed for the full benefits of female representation to manifest, is an example of this. While such an investor would prefer a board with one woman to a board with none, she also thinks that going from one woman to two is more valuable than going from zero to one. Accordingly, such an investor would prefer to “concentrate” the women onto the boards of a relatively small number of firms rather than spreading them out over as many firms as possible.
In addition to being decreasing, constant or increasing, marginal value can also change. For example, one might think that going from one woman on an eight-person board to two women is more valuable than going from zero to one (our increasing marginal value example) without believing that going from six women to seven is more valuable than going from zero to one. There is no contradiction here. It’s simply the case that, according to this particular investor, the marginal value is increasing at low levels of female representation, but then becomes decreasing at very high levels of representation.
B. Solving the Aggregation Problem for a Values-Based Investor
What should a values-based ESG investor do with this analysis? To see how this works in a concrete setting, consider the example from Part I. Since we are interested in a values-based ESG investor, our investor is willing to sacrifice some risk-return benefits in order to gain ESG characteristics. Given that there is a trade-off between her portfolio’s financial performance and the ESG characteristic (here, the representation of women on corporate boards), how should our investor make the trade-off between risk-return benefits versus ESG benefits?
The answer is that it depends upon her theory of the marginal social value of women on boards. Suppose she believes the marginal value is declining. This investor would prefer that all firms have at least one woman to a small number of firms with many women. This investor should therefore focus on the worst performing firms in her portfolio (i.e., firms with the fewest women) because dropping such firms can potentially produce the biggest “ESG bang for her buck.” In deciding which of the worst performers to drop, our investor would of course want to consider the foregone risk-return benefits. In practice, then, constructing this optimal ESG portfolio may entail a complicated trade-off between ESG and risk-return benefits.
However, one simplified and reasonable way of executing this strategy is to adopt a “minimum screening” test—that is, to simply drop the worst ESG performers. This is a common approach to ESG investing. For example, our investor might make sure that all the firms in her portfolio have at least one woman on the board, or, if that’s already satisfied, she might move the minimum up to two women. While she would—other things being equal—like to raise that minimum requirement as high as possible, other things aren’t equal. Rather, every ESG constraint that she adds to her portfolio necessarily entails a deviation from her hypothetically “optimal” non-ESG portfolio (i.e., from the portfolio that only maximizes her risk-return benefits). Because she is a values-based ESG investor, she is prepared to tolerate some deviation from that optimal non-ESG portfolio. But naturally, she will want to do so in the most efficient way possible; that is, to deviate in a way that makes the greatest impact on her ESG goal. And given her theory of declining marginal social value, one way for her to get the biggest ESG bang for her buck is by dropping the worst ESG performers. For such an investor, a strategy of ESG screens therefore makes sense.
By contrast, one strategy this investor should probably not employ is a “buy leaders” based strategy—that is, a strategy of investing in the best ESG performers. From the investor’s perspective, the difference between an ESG “leader” and a company that is just below the leader is quite small in terms of social value. This is because, by assumption, the marginal social value, from this investor’s perspective, is declining. Put another way, even small deviations from the ESG leaders could potentially produce substantial risk-return benefits. By solely focusing on leaders, our investor inefficiently over-constrains her portfolio and gets very little ESG bang for her buck.
To be sure, it is possible that a “buy leaders” strategy is optimal for an investor who believes in declining marginal value of ESG. It could be that, although the marginal ESG benefits of moving from a portfolio that “excludes the worst” to one that “includes only the best” may be relatively small, our investor may so value those small ESG benefits as to be willing to sacrifice enormous risk-return benefits. But this would be a very particular investor with extreme pro-ESG preferences. Such investors no doubt exist. But the point here is that the “buy leaders” strategy, while common in the real world, is consistent only with a very particular kind of investor preference.
To give another example, consider an investor who believes that marginal social value is constant. This investor would only care about the average level of the ESG characteristic in her portfolio. As a result, it’s also unlikely that she would prefer to “buy leaders” because she gets the ESG bang for her buck at all ESG levels. The fact that a company is a leader is not particularly important to her. This is not to say that she will necessarily dislike the “buy leaders” portfolio; but whether it is her optimal portfolio will depend upon the trade-off between the average ESG characteristics of the portfolio and the risk-return cost of achieving those characteristics.
The only time it does make sense for a values-based ESG investor to use a “buy leaders” approach—if only as a simplified rule of thumb—is if she believes that the marginal social value is increasing. This is because, all else equal, the marginal ESG gains of including only ESG leaders may be much larger than the gains from merely excluding the worst ESG performers.
The normative punchline of this Section is that a values-based ESG investor should adopt an aggregation method that reflects her theory of marginal social value. Otherwise, the investor will forego more risk-return benefits than necessary to achieve her ESG goals. Or, alternatively, she may inadvertently choose a portfolio that fails to achieve her ESG goals because it’s inconsistent with her theory of marginal social value. This risk is compounded by the fact that seemingly intuitive strategies like “buy leaders” can turn out to be fundamentally inconsistent with most theories of values-based ESG investing. Only an investor who believes the marginal value is increasing should, at least in the first instance, be interested in such an approach. Of course, security selection is often delegated to a portfolio manager, who acts on behalf of end investors. In such cases, it is the portfolio manager who is inefficiently (if not incoherently) trading off risk-return and ESG benefits. Alternatively, to the extent that the portfolio manager’s words don’t match her actions, investors might not be getting what they thought they were getting.
C. Solving the Aggregation Problem for an Impact-Based Investor
What about an impact-based ESG investor? While impact investing is often viewed as sitting uneasily with traditional ESG metrics, our analytical approach can easily accommodate it.
Because an impact-based ESG investor is concerned with the change in ESG, it’s natural for her to focus on the change in the ESG characteristic of the companies in her portfolio over some period (say, a year, or, for more long-term evaluations, five years). As before, she should evaluate those changes—and aggregate them across her portfolio—depending on their marginal social value.
It is straightforward to see how this would play out in practice. Take an impact-based ESG investor who is focused on labor practices. Assuming that the marginal social value of improving labor practices is decreasing, an investor would prefer a fund that succeeded in improving the conditions at the worst performing firm in her portfolio over a fund that had raised the standards at the best performing firm. In fact, she might still prefer the former fund even if the increase was much more modest than the increase achieved by the latter fund.
Moreover, if all the constituents in the portfolio begin with equally grim ESG characteristics, our investor would prefer that the portfolio manager focus on achieving modest improvements at all the constituent companies (or as many as possible) rather than attempting to achieve a very large improvement at just one (or a small number of them). This again follows from the premise that the investor believes that the marginal social value of improvements is decreasing. Naturally, the opposite would be true if she believed that the marginal social value of improvements was increasing. If she thought it was constant, then she would be indifferent as to the extent to which the improvements were concentrated or spread out across firms.
As with our discussion in Part III.B, this is a normative claim. Our claim is not that investors—or portfolio managers—in fact aggregate ESG characteristics in this way, but rather that they should do so. Failing to aggregate in this way leaves investor welfare on the table. Moreover, to the extent that a portfolio manager’s statements aren’t consistent with the aggregation method implied by her portfolio selection decisions, investors are unlikely to be getting the portfolio they expected.
D. Solving the Aggregation Problem for a Value-Based Investor
Not all investors who care about ESG characteristics are prepared to sacrifice risk-return benefits to achieve ESG aims. In some cases, the investor’s thesis is that a particular ESG characteristic (or set of characteristics) captures material financial risk not otherwise reflected in market prices. For example, large institutional investors like BlackRock and Vanguard have argued that, to the extent that they consider ESG characteristics in their investing decisions, they do so only to maximize shareholder value.
At first blush, risk-based ESG investing sits uneasily with our approach to ESG aggregation. The risk-based thesis is fundamentally at odds with a values-based investor’s willingness to make risk-return sacrifices in furtherance of some ESG goal. Nevertheless, what remains true for both approaches is that the investor is focused on the ESG characteristic with some goal in mind.
An investor whose theory of some ESG characteristics is risk-based should be focused on the relationship between the ESG characteristics and financial risk. Rather than marginal social value, then, her focus should be on marginal reduction in risk. If the investor’s theory of this ESG characteristic is that it represents a risk factor not captured by traditional measures of financial risk, her focus should be on how it reduces risk. Take, for example, the labor practices example from Part III.A. An investor might not be interested in labor practices per se, but she might be concerned that a company with particularly bad labor practices in its supply chain might find itself in trouble. Perhaps it risks unfavorable publicity, which could in turn lead to boycotts, fines, or other formal sanctions.
Whatever her theory of the underlying risk, the investor should have a sense of whether the reduction in risk that comes from improving the ESG characteristic is increasing, constant, or decreasing. In the labor practices example, the reduction in risk is almost certainly decreasing: while customers may be upset to learn about sweatshops, they will almost certainly be much more outraged to learn about forced labor. The same is likely to hold for other sanctions.
Our analytical approach to ESG aggregation does not address the standard critiques of this approach to investing—namely that it assumes that this risk is not already impounded into market prices. This is deliberate. Our analytical starting point is that the investor has solved the issuer-level measurement problem and knows how to appropriately value the issuer-level ESG characteristic. Given that, our analysis leads to a normative claim: risk-based ESG investors and portfolio managers should aggregate their ESG characteristics of interest based on their theory of its marginal effect on risk.
Similarly, an investor whose theory of ESG is that the characteristic will position a firm for improved financial performance should focus on that characteristic’s marginal impact on returns. If the ESG characteristic is an indication that the company is well positioned for high financial returns, we need a theory of the value proposition—what is the marginal impact on financial performance?
Here again, our analytical approach to ESG aggregation does not speak to the merits of this approach to ESG investing. We agree that the fact that this approach presumes that the market is not correctly pricing expected risk adjusted returns sits uneasily with standard concepts of market efficiency. Even to the extent that markets are not always perfectly efficient, one might wonder whether an ESG characteristic—especially one that receives a tremendous amount of public attention—is a good candidate for undervaluation. At the same time, the entire industry of active portfolio management rests upon the premise that an investor (or portfolio manager) can identify an undervalued investment opportunity. The fact that not every investor is fully invested in broad-based index funds further indicates that a substantial portion of the market—including many sophisticated and wealthy investors—believe that these mispricings exist. Our goal in this Section is not to wade into this debate. Rather, it is to provide guidance to an investor, or, more likely, a fund manager, who believes that such opportunities exist.
IV. Lessons for Practice and Policy
What should asset managers and the SEC do with this analysis? We certainly don’t think that forcing asset managers to use terms like “marginal social value” or “aggregation method” in fund prospectuses is a good idea. Instead, asset managers should (1) articulate their theory of why the ESG characteristic of interest matters, (2) explain whether its value is increasing, decreasing, or constant, and (3) adopt an investment strategy and a portfolio-level ESG metric that is consistent with that theory. These steps are, in our view, all necessary to enable investors to hold ESG funds accountable. Section A offers an example of how an asset manager might do this in practice.
As for the SEC, it should make rules that either encourage or require ESG funds to take the steps outlined above. Some of the SEC’s proposed ESG rules are useful, if incomplete, advances in this direction. Others, particularly the core features of the ESG Names Rule (which include minimum investment and “anti-greenwashing” provisions), may accomplish little or nothing. Sections B and C explain why.
A. Best Practices in ESG Fund Disclosure: An Example
How should an ESG fund implement our analysis? Returning to the simple example in Part I, here are three examples of positions that a fund might articulate. Option 1 is consistent with a theory of declining marginal value, Option 2 is consistent with a theory of constant marginal value, and Option 3 is consistent with a theory of increasing marginal value. Each option gives an example of the (stylized) language that a fund might use to articulate a coherent theory of ESG and an investment strategy that is consistent with that theory.
Option 1: “We support gender equality because we believe that women need to have a seat at the boardroom table. We believe that the value of increasing the number of female board members is largest for companies that have very few to begin with. Accordingly, we exclude companies that don’t have at least one female director from consideration, and the gender equity score of our portfolio is defined as the percentage of female directors at the 10 worst performing companies in our portfolio.”
Option 2: “Our goal is to maximize the number of women in the boardroom, so our focus is on the total number of female directors at the companies in our portfolio. Accordingly, the gender equity score of our portfolio is the percentage of women directors at our portfolio companies, where each company’s contribution to the total is proportionate to its weight in the portfolio.”
Option 3: “We believe that it is vitally important to increase the number of women at the boardroom level. Accordingly, the gender equity score of our portfolio is the percentage of our portfolio invested in companies where at least 40% of directors are women.”
All three of these descriptions satisfy the three criteria outlined: each outlines a theory of why female boardroom representation matters and (at least implicitly) outlines whether the value of an incremental woman is increasing, decreasing, or constant. Each statement also articulates a portfolio-level ESG metric that—and this part is crucial—is consistent with that theory. The three metrics are very different from one another, and except in unique circumstances, all three would rate the same portfolio differently. Naturally, these descriptions would look different for an ESG fund focused on a different characteristic, or even a bundle of ESG characteristics. Nevertheless, we hope that these examples serve as a useful starting point.
One might object that allowing—let alone encouraging—funds to adopt different portfolio-level ESG metrics will hinder investors’ ability to comparison shop. We disagree. In our view, it would be worse—indeed, incoherent—for two funds with different approaches to ESG to use the same ESG metric. Better to eschew standardization and let the metrics multiply—provided, however, that each fund uses a metric that is consistent with its ESG theory and investment strategy.
B. The SEC’s Proposed Rules
Early in the summer of 2022, the SEC became much more active in policing funds purporting to consider ESG characteristics. In addition to investigations and enforcement under the existing regulatory regime, the SEC released two major proposed rulemakings on May 25, 2022: the ESG Names Rule and the ESG Fund Disclosure Rule. If adopted, these rules would significantly expand the SEC’s mandatory disclosure regime on ESG funds. Here, we briefly describe each.
The ESG Names Rule would amend the existing Names Rule (Rule 35(d)-1). In its current form, the Names Rule defines the meaning of a “materially deceptive and misleading name” for the purposes of the Investment Company Act (which prohibits funds from adopting such names). Under the Rule, if a fund’s name suggests a certain type of investment, then the fund must dedicate at least 80% of its assets to that type of investment. Funds also must disclose the criteria used to select the investments that contribute to the 80%. Thus, a “Japan Large Cap Equities” fund would have to invest at least 80% of its assets in equities issued by large companies based in Japan. It would also have to disclose, for example, the fund’s threshold for “large cap.”
1. The proposed ESG Names Rule.
The proposed ESG Names Rule would change the current Names Rule in two ways. First, it would expand the 80% rule to fund names that suggest a focus on “investments that have, or whose issuers have, particular characteristics” and specifically to names “indicating that the fund’s investment decisions incorporate one or more ESG factors.” For funds subject to the 80% rule, related changes to fund registration forms would further require ESG funds to (1) define ESG-related terms used in the fund’s name, (2) identify which investments contribute to the 80% basket, and (3) explain the criteria used to select investments that contribute to the 80% basket. These three requirements effectively obligate the fund to justify its ESG theory and investment decisions. We will refer to these three requirements collectively as the “justification requirements” of the ESG Names Rules.
The second way the ESG Names Rule would change the existing Names Rule is by adding an “anti-greenwashing” provision. This provision would prohibit a fund from adopting an ESG-related name if “ESG factors are generally no more significant than other factors in the investment selection process, such that ESG factors may not be determinative in deciding to include or exclude any particular investment in the portfolio.” Such funds are sometimes called “integration” funds because their investment strategy integrates ESG and non-ESG factors into investment decisions. In the absence of this rule, an integration fund could “greenwash” its investments by merely announcing an intention to consider ESG factors. Under the proposed ESG Rule, a fund name may include ESG terms only if investment decisions are primarily determined by ESG factors.
2. The proposed ESG Fund Disclosure Rule.
The proposed ESG Fund Disclosure Rule would apply a tiered disclosure regime. Funds marketing themselves as ESG focused—a definition that includes all funds with words like “ESG” or “sustainable” in their names—would be required to explain how they consider ESG in their investment decision-making, as well as their engagement practices. Additional disclosure requirements would apply for impact funds. By contrast, the required disclosures for ESG “integration” funds (funds that consider one or more ESG factors but for which those factors may not be determinative in deciding whether or not to include or exclude a particular investment) would be somewhat more limited. For example, an integration fund need only offer a “brief narrative of how it incorporates factors.”
Under the ESG Fund Disclosure Rule, an ESG focused fund would be required to include, in its summary prospectus, an ESG strategy overview table. The first row of this table would consist of “a concise description in a few sentences of the factor or factors that are the focus of the fund’s strategy.” By way of example, the SEC suggests that “a fund might disclose that it focuses on environmental factors, and in particular, on greenhouse gas emissions.” The next part of the table is a “check the box” style section containing six common approaches to ESG, where the fund would be required to indicate which of the strategies it employs. It would then be required to briefly “summarize how it incorporates ESG factors into its process for evaluating, selecting, or excluding investments,” while providing a more detailed description later in the prospectus. The “check the box” strategies are: (1) Tracks an index; (2) Applies an inclusionary screen; (3) Applies an exclusionary screen; (4) Seeks to achieve a specific impact; (5) Proxy voting; (6) Engagement with issuers; and (7) Other. Interestingly, strategies (2) and (3) carry echoes of the discussion in Part II and could each be consistent with a specific theory of the marginal social value of the ESG characteristic—increasing marginal social value (assuming the inclusionary screen approximates a “buy winners” strategy) and decreasing marginal social value.
In explaining the rulemaking, the SEC observed that funds may disagree over how to promote a given ESG goal. According to the SEC, “[e]ven when investors focus on the same ESG issue, such as climate change or labor practices, there are debates about how to address such issues, resulting in different, and sometimes opposing, assessments of whether a particular investment meets the investors’ goals in furthering that issue.” Accordingly, the SEC believes that
requiring funds and advisers to disclose with specificity their ESG investing approach can help investors and clients understand the investing approach the fund or adviser uses. . . . The proposed rules would . . . better inform investors by providing them with decision-useful information to compare, for example, two funds that both refer to their strategy as “sustainable” but employ different approaches and areas of focus to implement their sustainable strategy.
We applaud the SEC for recognizing and validating these differences of opinion. In the next Section, however, we evaluate the proposed rules and conclude that there are several crucial shortcomings.
C. Evaluating the Proposed Rules
While they represent an improvement over the status quo, with one important exception, the proposed rules would not require a fund to commit to any particular ESG metric to evaluate the fund’s ESG performance. Nor would they require a fund to articulate a theory of marginal social value, let alone explain why, given that theory, the fund uses its particular portfolio-level ESG metric (if indeed it uses one at all). As we have argued above, without a clear link between ESG theory, ESG investment strategy, and portfolio-level ESG metric, it is not possible to evaluate a fund’s ESG performance.
Notwithstanding this, it is possible that the proposed changes will push the ESG mutual fund industry toward our suggested best practices. Indeed, the principal strength of the proposed rules is that they do require an ESG fund to be more detailed and careful in explaining its ESG investment strategy. For this reason, we are optimistic about the rules’ potential to help the ESG fund industry establish a tighter link between ESG theory, ESG investment strategy, and ESG metrics. Here we discuss how the proposed rules may do this and where they may fall short.
1. The proposed ESG Names Rule.
The principal mechanism by which the ESG Names Rule may push the industry toward best practices would not be through the 80% rule (which the SEC highlights as the main innovation of the Names Rule). The 80% rule is a reasonable requirement for non-ESG funds. This is because investors typically invest in, say, “Japan Equities Large Cap” to gain exposure to a certain kind of asset class (namely, equities issued by large Japanese firms). By contrast, values- and impact-based ESG investors do not merely want financial exposure to firms with certain characteristics. They want to promote the existence and success of such firms and to encourage other firms to adopt those characteristics.
The more useful component of the proposed ESG Names Rule is the “justification requirements.” Recall that these require funds with ESG names to (1) define the ESG terms, (2) identify which investments contribute to the 80% basket, and (3) explain how the fund selects investments that contribute to the 80% basket. Items (1) and (3) require a fund to articulate some theory of ESG investing, if not necessarily the coherent kind of theory that we offered in the analysis above. These requirements thus establish a framework through which a fund could achieve “best practice” without actually mandating the best practice itself.
A final potential benefit of the proposed ESG Names Rule is that, by compelling funds to define ESG-related terms, it could indirectly help standardize ESG terminology. We have argued above that the proliferation in ESG metrics is not in itself concerning and on the contrary may be required to properly evaluate ESG funds. The proliferation and inconsistent usage of ESG terms, on the other hand, is a major impediment to coherent ESG fund evaluation. Thus, funds should not be required to use the same ESG metrics, but they should be encouraged to use a given ESG term in the same way.
The proposed Rule does not list or otherwise explicitly describe the kinds of terms that would suggest an ESG focus—and thus trigger the Names Rule. Instead, the Rule implicitly defines ESG terms by requiring that any terms within a fund’s name be used consistent with their “plain English meaning or established industry use” (emphasis added). As a perhaps unintended consequence of the Rule, “industry use” could be established through compliance with the Rule itself. This is because the ESG justification requirements effectively compel a dialogue—through disclosure requirements—among funds and investors over the meaning of ESG terms and the investment criteria they suggest. This may be an unintended but beneficial consequence of the justification requirements, even if the requirements themselves do not directly promote best practices.
2. The proposed ESG Fund Disclosure Rule.
Because the disclosure regime contemplated by the ESG Fund Disclosure Rule varies by the type of ESG fund, we consider each of the three types in turn.
Given the weak, if not illusory, nature of the “promise” that integration funds make (namely, the promise to “consider” ESG factors in investment decisions), the SEC’s decision to impose only weak disclosure requirements on these funds sounds reasonable—at least at first glance. The SEC offers by way of example a green integration fund that considers emissions only when making an investment in a high-emitting company. Such a fund, it says, could comply with the disclosure requirement by disclosing that it considers emissions only for high-emitting companies and then defining “high-emitting.” While this example might appear to give investors a sense of the manager’s evaluation process, it may ultimately leave investors no better informed about the manager’s theory of ESG or how well the fund is doing at furthering that theory. After all, it tells investors nothing about what the manager does with this information.
The SEC should have gone further with integration funds. It should have required integration funds to provide some kind of metric that summarizes the extent to which the fund’s investments reflect the fund’s stated ESG goals. For example, a “no high-emitting” integration fund could simply report the percent of the fund’s assets that are invested in high-emitting companies. Without an objective measure of the extent to which the fund “considers” a given ESG characteristic, an integration fund’s ESG promise to consider that ESG characteristic is essentially worthless.
There is more merit to the slightly stronger requirements for ESG focused funds. For example, the “check the box” section of the ESG strategy overview—which, as discussed in the prior Section, implicitly reflects a theory of marginal social value—is a step in the right direction. The further requirement that such a fund explicitly describe its investment strategy by providing a detailed explanation of the methodology it uses to evaluate investments could, we hope, become a basis for selecting and disclosing the portfolio-level metrics we have argued are necessary.
The specific requirements for green funds come closest to mandating our suggested best practices. In addition to the general requirements for ESG focused funds, green funds must specifically report “weighted average carbon intensity” (WACI) of the portfolio. As the SEC explains, “WACI is the fund’s exposure to carbon-intensive companies, expressed in tons of CO2e per million dollars of the portfolio company’s total revenue.” All else being equal, changes in portfolio-level WACI do not depend on whether the changes in emissions were concentrated in one company or spread across companies within the portfolio. Thus, the proposed Rule implicitly mandates an ESG score based on a constant marginal social value.
While a theory of constant marginal value is reasonable (at least to a first approximation), the idea that the SEC should mandate this particular theory of marginal social value runs contrary to the regulatory regime’s normal deference to investor preferences. Most concerningly, the Rule effectively requires green funds that do not adopt a theory of constant marginal value to report a portfolio-level ESG metric that is inconsistent with the fund’s articulated goal and investment strategy.
To be sure, the WACI requirement may represent a reasonable majoritarian mandatory rule—presuming that most investors’ emissions preferences reflect a theory of constant marginal value. The requirement may also serve as a common standard to compare all green funds, as well as to evaluate the green fund industry as a whole. If this is the rationale, the SEC should require funds to report WACI in addition to requiring funds to select and report the ESG metric that is consistent with the fund’s theory and investment strategy.
For our purposes, the key additional requirement for impact funds is particularly promising. In addition to the general requirements for ESG focused funds, impact funds must “discuss the fund’s progress on achieving its impact in both qualitative and quantitative terms” (emphasis added). This effectively compels the fund to offer a portfolio-level ESG metric and to explain why that metric is appropriate. This, we hope, could end up effectuating a key component of our suggested best practices. The SEC should extend this requirement to all ESG focused funds.
So far, the ESG investing discourse has overlooked a crucial analytical point: it is impossible to aggregate ESG characteristics to a portfolio level without first establishing a theory of why that ESG characteristic is important. As we show with a simple example, investors with different underlying theories aggregate the same characteristic in very different ways, even when all the usual ESG-related challenges are absent. Our best practice guidelines provide a workable solution to this problem. ESG funds would be well served to take our suggestions on board, lest they risk confusing—or even misleading—their investors.
* * *
Professor Adriana Z. Robertson is the Donald N. Pritzker Professor of Business Law at the University of Chicago Law School. Professor Sarath Sanga is a Professor of Law at the Northwestern University Pritzker School of Law and the William Nelson Cromwell Visiting Professor of Law at Harvard Law School.
We thank Jill Fisch, Kate Judge, Elizabeth Pollman, Christina Skinner, David Weisbach, and the University of Chicago Law Review Online team for valuable suggestions and discussions. This Essay benefited from comments by workshop participants at the 1st Annual Women in Law & Finance Conference. Talla Khelghati provided exceptional research assistance. All errors are our own.