Why Sustainable Investing May Be Unprovable

One of the most persistent and polarizing questions in capital markets today is whether sustainable investing (including ESG) improves risk-adjusted returns. After more than a decade of data, backtests, and academic studies, our assessment remains unsatisfying: we can’t prove it.

Not only can’t we prove it (or disprove it), but the more uncomfortable possibility is that it may not be possible to know.

Not because the data is insufficient, or because the methodologies might be flawed—though both are true—but because of how markets actually function.

Price Discovery is from Trading, Not Investing

Markets are largely dominated by long-term investors—pensions, endowments, insurance companies, and increasingly passive index funds. These are institutions with multi-decade liabilities and investment horizons. Public markets are often described as mechanisms for efficiently incorporating information into prices. But that description overlooks a structural reality. 

Public markets are priced at the margin by much shorter-term actors—hedge funds, quant strategies, market makers, and increasingly retail flows. These participants dominate trading volume and respond primarily to price signals, positioning, liquidity, and near-term catalysts.

“Asking whether sustainable investing works by looking at short-term returns is a bit like trying to understand climate trends by watching the weather day to day.”

A Structural Mismatch — Not a Data Problem

Sustainable investing is, by construction, a long-duration proposition. Whether the issue is climate risk, governance quality, or social license to operate, the economic consequences tend to unfold over years or even decades.

But price formation—especially the data used to evaluate investment performance—is heavily influenced by participants operating on time horizons of hours, days, or weeks.

The result is a market that may be structurally incapable of producing a clear signal on whether sustainable investing “works.”

Why the Evidence Will Always Be Mixed

This helps explain why the evidence is so persistently mixed. Some studies show ESG outperformance. Others show underperformance. Many show no statistically significant effect at all. Results vary dramatically depending on time period, factor exposures, and methodology. This is often treated as an empirical puzzle—one that can be solved with better data or more refined models.

But what if the price signal itself is unstable because the system generating it is dominated by actors who are not optimizing for the same time horizon as the risks sustainable investing is trying to capture? In that case, asking whether sustainable investing improves risk-adjusted returns is a bit like trying to understand climate trends by watching the weather day to day. The measurement framework and the underlying time horizon are misaligned.

Past Performance should not be used to justify action

For years, much of the sustainable investing industry has leaned on the argument that sustainable investing or ESG “outperforms.” That claim has helped align sustainability with fiduciary duty and broaden adoption.

But if markets cannot reliably reflect long-term risks in observed returns, then performance may be the wrong lens entirely to justify doing or not doing ESG. That does not mean sustainable investing lacks economic value. It may, in fact, create meaningful long-term value—by identifying risks and opportunities that are real, but not consistently or efficiently priced. We happen to think that positioning portfolios for climate change (both adaptation and mitigation) will create such value.

The challenge is that a market dominated by short-term price setters may never provide definitive proof. Both proponents and critics of sustainable investing should be prepared for that possibility. The debate over whether ESG “works” may never be settled through performance data alone—not because the answer is trivial, but because the system itself is not designed to deliver one.

Which raises a more important question:

If markets cannot be relied upon to validate long-term risks and opportunities in a timely or consistent way, what happens when those risks begin to manifest in the real economy—unevenly, and across regions?

Part II of this series will attempt to address this question.