ESG Investing: Too Good to Be True?
Good vs. Bad Corporates
BlackRock is aggressively launching products with high environmental, social, and governance (ESG) ratings. The firm’s CEO, Larry Fink, recently predicted that assets under management (AUM) in the ESG category will grow from the current $25 billion to $400 billion in 2028.
Investing in good corporates is intuitively appealing. Who doesn’t want to make money while doing good? But do stocks that rank high on ESG metrics actually outperform? After all, sin stocks — think gambling, alcohol, tobacco, and firearms — generate above-average returns, according to research. It would be a bit paradoxical if ESG and more-virtue-challenged factors both outperformed. For insight into these issues, we explored ESG factor performance in the US stock market.
Methodology
We explored ESG data from a US provider that aggregates ESG scores from multiple sources. Using data dating back to 2009, we divided ESG stocks into four main groups: Citizenship, Environmental, Employees, and Governance. Next, we created beta-neutral long-short portfolios composed of the top and bottom 10% of US stocks as ranked according to these four factors. We only included companies with market capitalizations in excess of $1 billion. The resulting portfolios are rebalanced monthly and include 10 basis points (bps) of cost per transaction.
ESG Factor Performance in the United States
Except for the Employees factor, all ESG categories generated positive returns from 2009 to 2018. Based on these results, being a good corporate would seem to pay off.
Source: FactorResearch
Are ESG Factors New?
But ESG factor performance shares some common trends, so the underlying portfolios likely have overlapping stocks. This is fairly logical. Companies that care about the environment, for example, may show similar regard for corporate governance and their employees.
For more insight into the drivers of ESG factor performance, we conducted a factor exposure analysis. We found that ESG factors are biased towards common equity factors. Of particular note are the negative exposures to both the Value and Size factors as well as positive exposures to the Low Volatility and Quality factors. The R-squared calculations of the regression analysis from which the factor betas are derived average 0.5, implying that the common equity factors explain the performance of the ESG factors reasonably well.
Factor Exposure Analysis (Factor Betas), 2009–2018
Source: FactorResearch
The negative exposure to the Value factor suggests that stocks with high ESG scores have higher valuations than their less-conscientious counterparts. One possible explanation for this is that cheap stocks face temporary or structural issues that make achieving high ESG scores less of a priority for the companies.
Average Price-to-Book Multiples, 2009–2018
Source: FactorResearch
Smaller companies also have fewer resources. Indeed, the average market capitalization of the top 10% of the highest-ranking ESG stocks is almost twice that of the bottom 10%.
Average Market Capitalizations (US Billions), 2009–2018
Source: FactorResearch
Stocks with high ESG scores also display strong positive exposure to the Low Volatility factor. So lower-ranked stocks may be more volatile. There doesn’t seem to be an easy explanation for this, though it may reflect biases to certain sectors of the stock market.
Average Betas, 2009–2018
Source: FactorResearch
The positive exposure to the Quality factor, which we define as a combination of debt-over-equity and return-on-equity, has a simpler explanation: Less-levered companies tend to have management with a more long-term perspective, and highly profitable firms have more resources with which to achieve high ESG scores.
What Has Been Driving ESG Factor Performance?
According to the analysis thus far, ESG factors generate positive excess returns. Good corporates pay off, in other words, which would seem to be good news for all investors.
However, this performance may be a product of exposure to common equity factors. The ESG factors exhibited negative exposure to the Value and Size factors, which both generated negative returns since 2009.
More importantly, ESG factors had large positive exposures to the Low Volatility and Quality factors. The Low Volatility factor was the best-performing factor since 2009 and accounts for much of the positive excess returns derived from ESG factors.
Selected Equity Factors in the United States
Source: FactorResearch
Changing the Perspective
The factor lens is just one way through which to view the ESG factor puzzle. What does a sector analysis of companies that rank high and low on citizenship look like?
It turns out there are large sectoral biases. This is hardly a surprise. Energy stocks tend to have low citizenship scores. After all, oil-fracking companies in the United States are more associated with earthquakes than friendly community engagement.
How does the breakdown by sectors reconcile with the factor exposure analysis?
The negative exposure to the Value factor is reflected in the significant overweight in technology stocks, which trade at higher valuations. Tech companies also tend to have less debt and high profit margins, which explains the positive exposure to the Quality factor. Finally, the positive exposure to the Low Volatility factor can be reconciled with an underweight in sectors like Energy and Materials, which are more volatile than the average stock.
Citizenship ESG Factor: Breakdown by Sectors
Source: FactorResearch
Further Thoughts
The notion that companies that care about the environment, look after their employees, and exhibit good governance outperform is likely too good to be true.
The drivers of performance since 2009 were common equity factors. While factor exposure might change over time, ESG investors currently run the risk of missing out if small and cheap stocks start outperforming low-risk and high-quality stocks.
Historically, speculating on small and cheap stocks hasn’t been a bad bet. That may be true again.
For more insights from Nicolas Rabener and the Factor Research team, sign up for their email newsletter.
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©Getty Images/Ablestock.com
Professional Learning for CFA Institute Members
CFA Institute members are empowered to self-determine and self-report professional learning (PL) credits earned, including content on Enterprising Investor. Members can record credits easily using their online PL tracker.
It’s a fair question: do ESG stocks outperform? But that misses the point for many. I will invest in ESG regardless because I want to support firms that are good corporate citizens. if the return is the same or slightly lower than the relevant index, I don’t care! that’s the point. My value system ranks “doing good” over “doing well.” Read my book, “Money, Meaning, and Mindsets” for more on this topic.
I completely agree. The question, whether ESG Factor Investing has the intended effect of improving our world at all is much more relevant. The strong forces towards increasing market efficiency would counteract, as can be seen from other factor Investing. For example, Value Investing looses its outperformance rapidly since detection. Why should the ESG factor behave differently? If not so, market forces would take care of even capital allocation to the ESG as well as to the sin factor.
“We explored ESG data from a US provider that aggregates ESG scores from multiple sources.” This methodology has a serious weakness. Unlike eg “momentum” or “value” factor. There is no agreed definitions of an “ESG” factor (see also Axioma ESG research looking at idiosyncractic vs common ESG factors).
And so, this study doesn’t answer the question posed at all.
Indeed, you can see why theoretically many so-called “ESG factors” might be idiosyncractic and stock-specific and go in to the large “residual” quant bucket.
For instance, CEOs and fundamental investors typically agree that “corporate culture” is important for “value creation” and that this may fall across a “S” or “G” dimension. But there is no well defined “culture factor”. Indeed arguably most culture is stock-specific.
For full disclosure the ESG data provider should be named and also the underlying sources, and also whether this analysis has undergone any other peer review or not.
I comment in a personal capacity.
Hi Ben, I think you’re making a lot of valid points and more research on ESG is needed. We wanted to name the data provider, but after seeing the slightly negative results on ESG, they asked us to remove their name as selling ESG data represents their business.
However, most of the results are kind of intuitive and are likely to be verified by other researchers. Companies that rank high on ESG are more probable to come from the tech than the energy sector given structural industry differences. Sectoral biases lead to factor tilts.
Taking a further step back: Companies that rank high on ESG need to fulfill the requirements of stakeholders, not just shareholders, which consumes resources where the benefit for shareholders is debatable. Investing in ESG should not be seen as a potential source of alpha from our perspective, but more a personal preference.
Sell-side professionals are trying hard to invent something new to offer common public. Pursue of brokerage fee is the reason for such kind of inventions. Good luck!
This analysis and the research behind it is deeply flawed:
– Small sample size (not even 15 years)
– No adjustment was made for risk differentials between Standard and ESG approaches.
– The fact that the ESG approach “slightly” underperforms (no delta is given) confirms, if anything, that the analysis is inconclusive.
– Author is openly biased against a stakeholder governance approach.
– Fact: More than 90% of the over 2200 academic studies that have looked at ESG factors have positive findings.
– If you want well-balanced, candid ESG analysis from an undisputed ESG expert, check out Mary Jane McQuillen, featured in the Dec-5th-2019 Enterprising Investor.
Very disappointed that the CFA Institute would publish this type of analysis on its letterhead.
Hi Peter, thanks for your perspective. A few comments:
– We agree that the lookback is relatively short, however, that is a general industry issue with ESG data. It’s not only that ESG data is often only available for a few years, but that more and different data has become available recently, which means investors need to be careful with backtesting ESG data.
– Our analysis shows positive returns from backtesting ESG data as all four ESG categories generated positive excess returns from 2009 to 2018, so it is in line with most other research that was published. Our research simply shows that the returns can be explained by other factors, which again is in line with the consensus.
– However, it is worth highlighting that a large amount of research is published by ESG data providers and asset managers selling ESG products, which is not conflict-free and needs to be reviewed carefully.
– The research note you referenced in your second comment benchmarks realized returns from ESG-focused ETFs in the US since 2005 and shows a slight underperformance. Unfortunate for investors, as usual, there is a difference between theory (backtesting) and reality.
– I’m not biased against a stakeholder approach and believe it’s great that investors are nudging companies to become better citizens. However, there seem to be very few free lunches in finance and the data doesn’t support that ESG is one.
Best regards, Nicolas
For more perspective on Nicolas Rabener’s views when it comes to ESG, see this this link:
https://www.factorresearch.com/research-why-pension-funds-millennials-should-avoid-esg
Here’s the problem with your approach:
– When the evidence suggests that ESG underperforms, even ever so slightly and even with a small sample, you imply that the research is conclusive.
– When the evidence suggests that ESG outperforms, you attribute the excess performance to “other factors”.
So your conclusion remains the same… regardless of the data.
Suspect.
And then there’s this statement:
“Unless there is absolute certainty that ESG strategies will not underperform plain beta benchmarks, they should be avoided.”
I think most unbiased researchers would have probably concluded the opposite: If ESG strategies can be reasonably expected to perform, on a risk-adjusted basis, as well as their standard benchmarks, then there is no reasonable basis for investors to avoid them.
Perhaps coincidental.
And finally:
In other exchanges we’ve had, you stated that even if the data sample you used in your Dec-2019 research (Factor Research) was too small, that your conclusions were still valid because they are supported by “common sense”, since companies that score well on ESG factors tend to take a stakeholder-centric, as opposed to a shareholder-centric approach to value maximization.
This simplistic, zero-sum reasoning makes a terrible assumption: That resources directed towards ESG improvements do not generate value to shareholders.
Yet independent research suggests exactly the opposite: That relatively small investments in ESG improvements generate disproportionately large benefits in the form of reduced (reputational, operational, legal and regulatory) risks, reduced cost of capital, and improved cash flow stability and sustainability.
There are also indications that companies that exhibit good ESG performance are also better managed and less susceptible to fraud, making them superior investments.
One would have expected at least some of these well-established arguments to appear in any reasonably balanced review of the pros and cons of ESG investing. But strangely, none of them appear in your research.
I call that a pattern.
Hi Peter,
Thanks for the additional comments.
I think for the benefit of other readers and this discussion, it’s worth clarifying that we’re discussing two different research notes.
1) ESG Investing: Too Good to be True? This research note uses ESG data from a well-known provider and shows that theoretical long-short portfolios generated positive excess returns since 2009. However, the analysis highlights that most returns can be explained by common equity factors, which is in line with other researchers’ findings. Here is a quote from the latest JP Morgan Quant conference summary: “Since ESG is aligned with Quality/Low Vol…..”.
2) Why Pension Funds & Millennials Should Avoid ESG. This research note analyses realized returns from ESG ETFs in the US and highlights that these slightly underperformed their benchmark since 2005. Given the evidence, we make the case that financially-impaired investors like public pension funds (only 70% funding in the US) and Millenials should avoid ESG as they need to maximize returns and avoid unnecessary risks.
Both two articles are about ESG investing, but one with theoretical and the other with realized returns, and have different objectives.
Regarding the data: If I recall correctly, you initially wrote that the 15 years of realized returns is somewhat limited and I simply agreed to that. In some of our other analysis we go back to 1926. Naturally this isn’t possible with ESG.
I think your other points essentially express that you believe that a stakeholder approach generates more for shareholders than a shareholder approach. I also believe that certain corporate characteristics like having a sound corporate governance are accretive to shareholders.
However, there will also be capital misallocations due to ESG and screening out entire industries, some which are highly profitable, isn’t sensible, so overall my theory is that ESG strategies will slightly underperform benchmarks. My theory is only supported by 15 years of realized returns and so it will be interesting to see if that will be validated over time. Personally, I would be delighted to be wrong.
Finally, introducing a different perspective to this discussion: Private equity firms are focused aggressively on creating shareholder value, typically at the expense of stakeholders like employees. They seem to be rather successful with this approach, at least judging by the AUM they’ve been gathering.
Best regards, Nicolas
Our disagreement is not about my beliefs, or your beliefs.
It’s about drawing irresponsible conclusions based on razor thin evidence and flimsy analysis.
When you have two equally plausible but mutually exclusive theories, the standard of evidence required to declare one of them correct is very high.
You cannot, for example, flip a coin 20 times, get 9 Heads and 11 Tails and then reasonably conclude that “Tails Never Fails”. Anyone convinced by that line of reasoning would be fundamentally misunderstanding probability theory and the law of large numbers.
In the case at bar, the theories are:
A. Positive ESG performance requires diverting resources away from shareholders, results in capital misallocation, suboptimal diversification, higher research costs, etc…
B. Positive ESG performance reduces reputational, operation, legal and regulatory risk, reduces the cost of capital, improves free cash flow stability and sustainability, signals superior management, reduces the occurrence of fraud, etc…
Since both of these theories appear to be equally plausible at face value, as a researcher you ought to have considered, before even beginning, what type of analysis is best suited to reasonably conclude one way or another. Had you done this, you would have realized that ESG factors, being qualitative in nature, are not particularly well suited for quantitative factor analysis. Error #1.
Nonetheless, having concluded that you wanted to proceed with the methodology you ultimately selected, you ought to have quickly realized that your performance data sample was much too small to render a meaningful analysis. Error #2.
But since you proceeded anyway, surely when you saw the results of your analysis, you ought to have questioned the materiality of the performance differential observed, and whether it could be reasonably attributed to chance. Error #3.
And when you decided to publish your research, you had full control over the language you used. You could have easily stated your factual findings, while humbly indicating that your research is inconclusive. Instead, you chose to use your research to create a veneer of credibility and make irresponsible investment recommendations to an entire generation investors and a troubled pension system. Error #4.
I give you full credit for being a gentleman in your debates, but your analysis does not belong in a CFA publication.
This is my last post.
Hi Peter, thanks for the additional feedback.
I’ve already highlighted the consensus from the latest JP Morgan quant conference, where other sophisticated investors also explain ESG returns with factors and derive similar conclusions.
Furthermore, the latest investor survey from Morgan Stanley show the following results when asked “What are the primary drivers for expanding the ESG/SRI efforts?”:
– Investor demand / help asset raising: 46%
– Incremental alpha: 17%
– Risk mitigation: 0%
– Making an impact: 12%
– ESG is not focus at this time: 25%
It seems that ESG is more about marketing and sales rather than alpha generation or risk reduction according to these investors, which aligns with our perspective.
Best regards, Nicolas
Best ESG papers of 2019, for those who are interested:
https://www.savvyinvestor.net/blog/awards-best-esg-white-paper-2019
Over the years, a lot of companies have adopted the process of ESG investing. The Socially Responsible Investing companies provide an insight to the investors to the value of the company. Due to ESG, environmental concern has become a predominant factor for most of the companies, because a lot of these companies refrain from contributing to the climatic changes.
Mr. Rabener,
Nearly a year has gone by since our last exchange.
ESG strategies have performed relatively well in 2020 (not that that has any particular significance in my view), such that, if you were to perform your analysis again including the updated information now available, you would probably find that ESG strategies are now slightly ahead.
This now leaves you with a difficult dilemma:
– If you continue to believe in your analytical approach, then your conclusion must change.
– If you continue to believe in your conclusion, then your analytical approach was flawed.
Of course, we could be back here next year and maybe ESG strategies are behind again. I suspect that you will carefully time the update of your analysis to secure a consistent conclusion over time.
And that has always been my point: your analysis presents conclusion-driven evidence, not evidence-driven conclusions.
Hi Peter, great to continue this dialogue. A few thoughts from our side:
– More ESG data sets have become available over the year and existing ones revised, which makes it more challenging to evaluate ESG investing given the heterogeneity. As of today, there are > 600 ESG frameworks.
– If we would have used the same data set as in our original analysis, then the long-short ESG portfolio would have outperformed YTD 2020, which is explained by the large overweight to technology stocks.
– Most research in finance points to almost zero free lunches being available to investors, except for diversification. Being skeptical of ESG representing a free lunch does not pose an issue when conducting analysis, as long as we’re open to being wrong, which we are. Unless you’re a robot, it’s probably impossible to be unbiased anyways. No dilemma here from our perspective.
– It is worth mentioning that I was on various ESG panels throughout the last year and none of the ESG-focused asset managers or investors made the case for ESG generating outperformance. It either focused on using ESG as a risk management tool, which is also debatable, or fulfilling personal preferences. Naturally just anecdotal evidence, but worth sharing.
Best regards, Nicolas
Hi Nicolas,
Here are a few counterpoints:
– Heads I Win, Tails You Lose: You say that your personal biases don’t affect your research, but look at the game you’re playing: When the evidence supports your a priori belief, you consider it conclusive, but when it doesn’t, it is because of some exogenous factor, like the outperformance of the technology sector.
– On Free Lunches: ESG investing is hardly a free lunch; in fact, analyzing extra-financial information is much more costly than analyzing financial information. You are treating it as a free lunch only because you are relying on quick and dirty metrics to identify “ESG stocks”, much like how naive investors use P/E and P/B ratios to identify “Value stocks”. Indeed, why would the market reward either one of these naive approaches? Perhaps that’s why your long-short portfolio is increasingly looking like a random walk.
– On ESG performance: As you know, I never made the claim that ESG investing outperforms traditional investing. You made the claim that ESG underperforms and therefore should be avoided, I simply critiqued the flimsiness of your analysis and challenged your conclusion.
My position has always been that ESG investment strategies can perform as well as traditional investment strategies when adjusted for cost and risk differentials, and so far, your research supports my claim better than it supports yours.
Peter
Hi Peter, thanks for the additional comments, although our discussion is becoming somewhat circular.
– It would be great to be totally unbiased when conducting analysis, but that seems impossible given our human nature. We can only try to be as open-minded, thorough, and data-driven as possible.
– Regarding the “quick and dirty” ESG metrics: our data is from a dedicated ESG provider that aggregates ESG scores from the major data providers, which essentially represents meta-scores.
– Classic equity factors like value or momentum are backed by hundreds of academic research papers, although that doesn’t mean consistent outperformance. There is much less evidence for ESG, which would make the scientifically-minded investor somewhat cautious.
– Our research simply highlights that ESG portfolios have large factor and sector bets, which in the short-term may go either way. Given the composition of the factor exposures (negative Value, Size, Momentum, positive Low Vol & Quality) and higher costs for ESG products, this would indicate underperformance in the long-term, assuming these exposures remain constant.
Best regards, Nicolas
Circular, indeed.
Good luck with your oil and tobacco stocks.
Let’s circle back in 10 years and you can show me your outperformance then.
Peter
I would think that the popularity and often mandating for funds, ETFs that invest in ESG has created a new source of demand for these securities that has contributed to the results observed. Once this demand is satiated relative to all securities, the excess returns are likely to moderate or disappear all together. Yes, a free lunch is a mirage.