Investing Classics: ESG & Sustainable InvestingWhere Values Meet Prices

ESG's root problem is category confusion. E (carbon emissions), S (labor, supply chains) and G (governance, shareholder rights) frequently contradict each other — a well-governed tobacco company scores high on G and low on E/S; how do you sum that into one number? Worse, the rating agencies disagree. MIT Sloan's study Aggregate Confusion found that the pairwise correlation across six major ESG raters averaged just about 0.54, versus 0.99 for credit ratings. The same company can be an honors student to one rater and a laggard to another. When the underlying signal is that inconsistent, what you buy is mostly the noise of a rating methodology, not real sustainability.

In May 2022, S&P removed Tesla from its ESG index while ExxonMobil sat in that index's top ten; Musk publicly called "ESG a scam." The point isn't who's right — it's what a scoring system that ejects the EV leader and keeps the oil major is actually measuring. The answer: S&P's ESG score weights "how a company discloses and manages ESG risk" far more than "the business's net impact on the planet." Two completely different things, hidden under one name.

ESG isn't worthless — the G (governance) dimension has solid empirical support: bad governance really does foreshadow long-term risk. The problem is lending G's credibility to the fuzzy E and S.

• In the ESG score I use, are E/S/G viewed separately, or blended into one total?
• Does it measure "net impact on the world," or "the company's risk-management process"? The two get conflated constantly.
• Switch rating agencies — does this company's rank flip completely?
• Do I use ESG because it truly improves risk-return, or just to "look responsible"?

The economics of greenwashing are simple: a "green" label attracts money and commands a higher fee, while the cost of verifying the claim falls on someone else — information asymmetry creates room for marketing arbitrage. The regulatory response is to turn "said" into "must do": the EU's SFDR requires funds to classify and disclose by degree of sustainability; the U.S. SEC enforces directly against false statements. But regulation breeds a new game of its own: once "green" becomes an arbitrageable label, the rules defining the label become the new battleground.

DWS, Deutsche Bank's asset-management arm, once claimed ESG was "in its DNA." In 2021 its former head of sustainability blew the whistle that this was not the case; in May 2022 German police raided its Frankfurt headquarters and the CEO soon resigned; in September 2023 the SEC fined DWS $19 million over ESG-related misstatements. A larger-scale signal: after SFDR took effect, from late 2022 into early 2023 large numbers of funds originally labeled top-tier (Article 9) were reclassified down to Article 8, covering well over a hundred billion dollars in assets — the business hadn't changed; the bar for "green" had finally been enforced.

Not every label change is greenwashing — sometimes it's the healthy result of tighter rules and clearer definitions. Real greenwashing is about the gap between "said" and "done."

• Is this fund's "sustainability" in the holdings, or only in the name and cover?
• If you stripped the ESG label from its top ten holdings, could you tell any difference?
• The premium fee it charges — does it buy verifiable substance, or an adjective?
• When regulators tighten definitions, does it raise its standards, or quietly reword?

You must separate three things that are constantly conflated: (1) negative screening (don't buy tobacco, weapons) — this changes your portfolio, not the world; (2) ESG integration — treating ESG as a risk factor in valuation, still aimed at returns; (3) impact investing — using primary-market capital (venture, project finance, private credit) to do things that would not happen without your money. The first two have almost no additionality in secondary markets: the stock you sell is instantly bought by someone else, and the company's cost of capital doesn't budge. Real additionality lives where capital is scarce — early-stage clean tech, developing-world infrastructure, transition projects nobody else will lend to.

Academia has a sober quantification of this. Berk and van Binsbergen's study The Impact of Impact Investing (2021) found that broad divestment has a negligible effect on the cost of capital of the divested firms — because there is always an indifferent buyer to step in at a slightly higher expected return. Their conclusion: to actually change corporate behavior, engagement (staying invested and pressing) is far more effective than divestment. In other words, selling the stock feels good, but holding it and voting against management — or steering new capital toward what is genuinely starved of it — is what might create additionality.

Additionality is extremely hard to measure and the easiest thing to overstate — nearly every impact fund tells a moving impact story.

• If I didn't invest this money, would the good thing still happen? (If yes, additionality ≈ 0.)
• Is this primary market (new capital into a real entity), or secondary (shares changing hands)?
• Is the impact metric an auditable real-world output, or a narrative the fund tells about itself?
• Am I pursuing real change, or buying the emotional comfort of "I took part"?

The logic is clean: if a group of investors refuses on moral grounds to buy "sin stocks" (tobacco, gambling, fossil fuels), those stocks' prices are pushed down and their future expected returns therefore rise — and that premium is earned by the investors who don't care. That is the source of the "sin premium." So values investing inherently contains a trade-off: you exchange (potential) excess return for sleep that matches your convictions. This is not a moral failure — it's a choice that must be honestly priced. What is truly dangerous is the free-lunch narrative that "responsible also earns more" — it dodges the trade-off, and the moment sin stocks outperform, its adherents waver.

In The Future for Investors (2005), Siegel found that among the original 1957 constituents of the S&P 500, the long-run return champion was Philip Morris (tobacco, now Altria), returning roughly 19.75% annualized from 1957 to 2003, far ahead of the index — driven partly by chronically low valuations compounded with reinvested high dividends. The other side: Norway's sovereign fund (GPFG) maintains a long-standing ethics-based exclusion list (tobacco, certain weapons, severe polluters) and openly acknowledges these exclusions have a measurable, usually negative but acceptable effect on returns — it chooses to bear that cost transparently rather than pretend it doesn't exist.

Values investing doesn't have to underperform — diversifying and avoiding genuine long-term risks (e.g. stranded assets) can even help. The key is not to fool yourself.

• Am I excluding an asset class out of conviction, or out of a prediction that "it will underperform"? (The latter is often wrong.)
• Would I hold the exclusion even if that asset class beat me for three years straight?
• Is the (expected) price I pay for my values small enough to bear over the long run?
• Am I buying the "responsible and more profitable" free-lunch narrative? If so, beware.