Data has shown that investment strategies that address environmental, social and governance (ESG)[1] issues constitute a third of professionally managed U.S. assets. That has led some to claim that asset prices have been driven up to the point where investors should expect poor performance going forward.
That narrative is false because it relies on what I consider to be an overly broad definition of an ESG asset.
The AUM of assets addressing ESG issues depends on how one defines that universe. A constrained definition – one which I consider reasonable – shows that those assets are much less than what is claimed on an absolute and relative basis.
I have been guilty of perpetuating that false narrative, as have been other authors whose articles have appeared in Advisor Perspectives.
A prominent source of that narrative is the US SIF Foundation’s widely followed biennial report, which stated that one third of U.S.-domiciled, professionally managed assets addressed ESG considerations as of the end of 2019, either via ESG incorporation[2] or filing or co-filing shareholder resolutions on ESG issues. Of the $51.4 trillion in total U.S. assets under professional management, the report claimed that there were $17.1 trillion using sustainable investing strategies.
If one-third of professionally managed assets were following the same strategy – be it addressing ESG issues or any other – it would likely represent what researchers call “overgrazing” – too many dollars chasing too few securities, driving prices above their intrinsic value and depressing future returns. Though the US SIF Foundation does not directly discuss performance in its biennial report, it does provide a list of studies and meta-studies demonstrating the strong performance of ESG funds in recent years.
Claims of overgrazing have been made against value investing and other quantitative strategies, but it is difficult to measure asset flows to a strategy, and most overgrazing claims remain an unproven theory.[3]
Nevertheless, asset flows matter, and it is critical to understand the limits of any strategy. Other than a global, market-capitalization-weighted portfolio, every strategy has an asset limit beyond which returns will suffer.
David Snowball, the editor of the Mutual Fund Observer, raised this issue. He is also a professor at Augustana College, where he has been teaching statistics as rhetorical tools for the last 40+ years. In an email correspondence, he called the global estimates of ESG assets, “laughably unreliable.” He wrote, “Social problems are most compelling when they have a number attached: the number of abused elders, homeless people, missing children, victims of stalking, undocumented immigrants. Numbers make problems real for us. People involved in causes know that, and they know they need to find the ‘right’ number ... and so they do.”
Let’s look at the claim that one-third of assets address ESG issues. I will show that ratio uses a numerator that is based on a broad definition of what constitutes an ESG asset and a denominator that is understated.
The size of the ESG universe
The researchers at the US SIF Foundation arrived at its $17.1 trillion estimate by employing a combination of direct surveying as well as primary and secondary research. Through this process, the report identified 530 asset owners, 384 money managers and 1,204 community investing institutions in 2020 engaged in some form of sustainable investing.
The US SIF Foundation’s findings reflect a rigorous process that incorporates an inclusive definition, clearly stated, of what constitutes ESG investing. This definition reflects the breadth and depth of ESG incorporation across the financial markets. The US SIF Foundation’s methodology is clear (available for review as chapter 5 in the US SIF Foundation’s Trends report), and its staff told me it has been consistently applied for years.
Despite that rigor and clarity, the inclusion criteria in US SIF Foundation’s definition are very broad.
The survey asked respondents to indicate which of a series of ESG criteria it used in its investment process, such as climate change/carbon, anti-corruption, board issues, tobacco, etc. There were 37 criteria across five categories: community (8), environment (7), social (7), governance (5), and product and industry (10).
If an institution indicated that it used at least one of those 37 criteria, such as screening for independent board directors, it was included in the ESG totals.
On this definitional issue, here’s what Snowball wrote:
I suspect part of the problem is the proliferation of "soft" ESG disciplines, who claim to account for ESG factors, "to the extent that those factors directly affect financial performance," which seems to be one of the ways that non-ESG strategies grab some marketing or moral edge. Similarly, I suspect that some of the negatively screened funds ("we'll buy anything except porn and fossil fuels") allows the entire remainder of the portfolio (the meat packer Tyson foods) to count toward the ESG total. That is, a billion dollar fund whose sole virtue is avoiding two industries puts all of the remainder of the portfolio into nominal ESG land.
Indeed, academic research has shown that asset managers do not always adhere to the ESG investment principles they proclaim.
The US SIF Foundation faced a computational task in that it had to adjust for any assets that were double- or triple-counted. For example, an institution (e.g., Harvard University) could use a money manager (e.g., Russell Investments) that could buy investment products (e.g., BlackRock ETFs). In those situations, it had to make the appropriate adjustments to avoid double or triple counting of assets. I have no reason to suspect there were errors in those adjustments. But I would have liked to verify the data.[4] I requested an anonymized version of the data it collected, but this request was denied, as was my request for a blank copy of the survey.
The US SIF Foundation’s research identified $11.5 trillion (69% of the $17.1 trillion) in “undisclosed investment vehicles.” That undisclosed component has become an increasingly large portion of the assets the US SIF Foundation tracks, and it grew by 53% since the last time the survey was done in 2018. Most of the “opaque” AUM came from public PRI Transparency Reports or other institutional public disclosures, which showed that money managers say they have large portions of ESG AUM but provide no additional information on $9.5 trillion of vehicles. Of that $9.5 trillion pool of assets with no investment vehicles disclosed, $6.4 trillion did not have specific ESG factors associated with them. For further details and analysis, see page 36 of the US SIF Foundation’s report section titled, “Undisclosed Investment Vehicle Assets.”
The effect of the broad definition used by the US SIF Foundation shows up in the 718 mutual funds it identified with $3.06 trillion in assets and 94 ETFs across $21 billion that address at least one ESG criterion. For a comparison, I obtained data from Morningstar Direct and found that it identified only $260 billion in ESG assets across 392 mutual funds and ETFs as of March 10, 2020. Morningstar includes only funds and ETFs that have an explicit sustainable or responsible investment mandate.
I have interviewed mutual fund managers who do not explicitly market an ESG product. I have asked them about the reported surge in ESG-related assets, and they have responded that they have always included those criteria in their process. But they chose not to use the explicit ESG label, so they are not included in the Morningstar data. But they would be included by the US SIF Foundation.
To illustrate the difference between the US SIF Foundation and Morningstar estimates, consider that there are approximately $18 trillion in mutual fund assets and $4 trillion in ETFs. Based on Morningstar data, ESG assets represent 1% of that total; based on US SIF Foundation data, ESG assets are 14% of the total.
The growth in ESG assets is undeniable, particularly over the last decade. But whether you choose to include only funds with an explicit ESG mandate (as does Morningstar) or funds that employ as few as one criterion (as does the US SIF Foundation) makes a big difference.
The US SIF Foundation used a broad and generous definition of ESG (albeit a definition that is clearly stated), which explains why its results differ so greatly from Morningstar’s. That, coupled with the fact that 69% of its assets did not have an associated investment vehicle, means that its estimate of $17.1 trillion is an extreme, upper limit of what should be considered ESG assets.
The size of the professionally managed universe
Cerulli Associates provided the estimate of $51.4 trillion in U.S.-domiciled, professionally managed assets. I spoke with Bing Waldert, its managing director of U.S. research, to understand this calculation.
Cerulli used what I call a “top-down” approach to size the investment universe. It obtained data from regulatory bodies and reporting agencies to determine the amount of assets managed in various channels, such as advisory, insurance, endowments, and pensions. It also obtained data by product type to assess the amount of assets in vehicles such as mutual funds, ETFs, hedge funds, and separately managed accounts. It then went through the same process as the US SIF Foundation to adjust for double- or triple-counting before arriving at its final total.
One difficulty with this approach is that it will miss any assets held by an institution that is not subject to the reporting requirements that Cerulli used. I do not know what those institutions might be, but I know that omission makes Cerulli’s estimate problematic.
I know this because I calculated the size of the professionally managed universe in an alternative, “bottom-up” fashion. The market capitalization of the U.S. stock market is approximately $51 trillion, and the U.S. bond market has approximately $40 trillion in assets (excluding debt held by U.S. government entities, such as the Federal Reserve and the Social Security trust). Included in those $91 trillion in assets are approximately $8.5 trillion in self-directed brokerage accounts, according to Cerulli. Cerulli also estimated that approximately 20% of those brokerage assets ($1.7 trillion) are in individual securities; the rest are in mutual funds, ETFs or other products.
Thus, there are only $1.7 trillion in assets that are not professionally managed. The remaining $89.3 trillion is professionally managed. (Since this is a bottom-up approach, there is no need to adjust for double- or triple-counting.)
My bottom-up estimate could be too large if, for example, some of the shares of the U.S. equities are not available in the public float, or some of the assets are held by non-U.S. domiciled entities (which is surely the case). But my estimate is also too conservative because it does not include non-U.S. securities managed by U.S. institutions or asset classes other than stocks and bonds.
The size of the U.S.-domiciled, professionally managed universe is much closer to $89.3 trillion than $51.4 trillion.
No overgrazing
The most conservative estimate of the relative size of the ESG universe is the $260 billion identified by Morningstar divided by the $89.3 trillion of professionally managed assets, or a market share of 0.29%. That is too conservative, because it does not include so-called “impact” funds with an ESG mandate or other investment products, like separately managed accounts, that have an explicit ESG mandate.
But even if my estimate is too conservative by a factor of 10, it still means that that ESG investors should not be concerned about overgrazing. There is plenty of fertile ground for ESG products to graze upon to find appropriate securities to purchase.
Where does this leave ESG investors? There are material differences in the way ESG investing is defined. The US SIF Foundation has chosen the broadest possible definition. But advisors must make decisions based upon their clients’ specific preferences. This leaves them with a challenging due-diligence process of understanding the rules used by various products and aligning them with their clients’ goals and aspirations.
The most important determinant of investment returns is expenses, and ESG funds can still be expensive, as is the case with any actively managed strategy. But there are a growing number of passive ESG funds available. The weighted-average expense ratio of the funds in the Morningstar universe is 63 basis points; 22% of those funds have expense ratios of 1% or higher.
Beware of the inclusion rules of the ESG funds you choose and the costs you pay, but don’t worry about overgrazing.
Robert Huebscher is the editor of Advisor Perspectives. The US SIF Foundation’s staff reviewed this article and provided valuable feedback.
[1] In this article, I use the term “ESG” to broadly include strategies with an ESG, SRI or socially responsible mandate. US SIF’s definitions can be found here.
[2] For US SIF Foundation descriptions of the various ESG incorporation strategies, see here.
[3] That has not stopped some researchers from trying.
[4] With academic studies, it is routine for researchers to provide their data to those wishing to verify their results. US SIF Foundation notifies those who fill out the survey that it will not publicly share their information as a way to encourage institutions to fill out the survey. The US SIF Foundation treats this data confidentially and does not provide it to anyone, including its member institutions.
Read more articles by Robert Huebscher