Do ESG Funds Outperform?

What the research says about ESG fund performance

The largest review of the evidence finds parity

Atz, Van Holt, Liu and Bruno reviewed over 1,100 primary peer-reviewed papers and 27 meta-reviews for the Journal of Sustainable Finance & Investment in 2022. Their conclusion is risk-return parity: at portfolio level, ESG investing performs indistinguishably from conventional investing.

One study in three finds outperformance

Roughly a third of the individual studies in that review showed superior financial performance. This is the detail most summaries drop, and it explains why both sides of the argument can cite real research. A third of the literature supports outperformance. The aggregate does not.

The largest fund dataset shows a small lead

Morgan Stanley's Institute for Sustainable Investing tracks the largest available dataset on this question. $100 invested in a sustainable fund in December 2018 was worth $162 by the end of 2025, against $152 in a traditional fund.

Geography explains much of that lead

Morgan Stanley attributes much of the difference to region. Roughly 70% of sustainable funds invest globally or in Europe, against 40% of traditional funds. When those regions lead, sustainable funds lead. That is a regional tilt, not an ESG effect.

Method matters more than the label

The 2022 review finds that ESG integration — using material ESG data alongside financial metrics — consistently outperforms negative screening and exclusionary divestment. ESGU, the largest ESG ETF in the US, weights holdings by ESG score against industry peers. ESGV, its closest competitor by size, removes fossil fuels, tobacco, weapons, gambling and nuclear outright. Both are sold as ESG funds. The research says one of those approaches works better than the other.

Why the evidence points both ways

Three reasons ESG funds might outperform

Downside resilience. ESG strategies show asymmetric benefits during shocks. High-rated firms tend to hold stronger balance sheets, lower capital costs and better stakeholder management, which buffers drawdowns.

Operational quality. A 2024 meta-analysis by Bai and Kim finds environmental and social measures correlate positively with return on equity and return on assets. Resource efficiency and labor standards act as proxies for management quality.

Cost of capital. Highly rated companies borrow and raise equity more cheaply, which compounds over long horizons.

Four reasons they might not

The diversification penalty. Exclusions restrict the investable universe. Under modern portfolio theory a restricted universe is a less efficient one, and the cost shows up whenever excluded sectors rally.

The premium effect. Capital flowing into highly rated assets bids their prices up, which lowers future expected returns. Shunned assets may offer higher expected returns as compensation for holding them. This is the mirror image of the cost-of-capital argument: the same repricing that delivers returns during the transition removes them afterward.

Fees. Actively managed ESG funds generally carry higher expense ratios than passive index equivalents, and the drag compounds.

Rating divergence. Providers disagree, so a "high ESG" portfolio depends on whose data built it. This is significant enough to warrant its own section below.

The rating agencies disagree with each other

Correlation across providers averages 0.54

Berg, Kölbel and Rigobon examined six major providers — KLD, Sustainalytics, Moody's ESG, S&P Global, Refinitiv and MSCI — and found an average pairwise correlation of 0.54, ranging from 0.38 to 0.71. By comparison, Moody's and S&P agree on credit ratings almost perfectly. On ESG, the same company can be a leader to one agency and a laggard to another.

Meta scores low risk and laggard at the same time

Sustainalytics places Meta Platforms as the lowest-risk company in this group. MSCI rates it a laggard. Two of the three best-known agencies put the same company at opposite ends of their scales.

The gap is structural rather than an error. MSCI rates companies relative to industry peers, so Meta is measured against other interactive media firms. Sustainalytics scores unmanaged risk on a scale designed to compare across industries. A company can carry little absolute unmanaged risk and still trail its own sector badly.

A tobacco company and an oil major get the same grade

MSCI rates Philip Morris and ExxonMobil identically, at BBB. Sustainalytics separates them by more than 20 points — Philip Morris in the Low Risk band, ExxonMobil in High Risk. One agency sees two average performers. The other sees one of the safest companies in this group and one of the riskiest.

Neither is scoring the product. Both are scoring how much ESG exposure threatens the company's value after management.

The disagreement is mechanical, not philosophical

Two agencies can both decide to score labor practices, then measure them differently — one counts reported injury rates, another counts whether a policy exists. That kind of measurement gap explains 56% of the disagreement. A further 38% comes from the agencies not scoring the same things at all: one includes lobbying activity, another leaves it out. Only 6% comes from weighting the same measurements differently.

Governance is the worst case. It is the dimension investors treat as most objective, and it is where the agencies agree least — a correlation of 0.30.

A company's ESG rating depends substantially on who is doing the rating.

How ESG investing got here

From institutional risk framework to retail product

ESG began as a risk framework for institutional investors and became a retail product category. Assets grew quickly through the late 2010s as fund managers applied ESG labels across existing strategies.

Greenwashing, politics, and a definitional problem

Greenwashing enforcement exposed a gap between what funds claimed and what they held, prompting regulators in the EU and US to tighten fund-naming and disclosure rules. Political opposition in the United States produced state-level restrictions on ESG considerations in public pension management, and several asset managers withdrew from climate alliances. And retail investors discovered what the ratings actually measure.

That third point is the durable one. An investor who buys an ESG fund to stop funding harm finds it does no such thing. ESG ratings score risk to the company, not harm to the world. The framework was built to answer a different question.

The largest ESG ETF sits on a 3.7°C global warming trajectory

The Paris Agreement's outer ceiling is 2°C.

What ESGU actually holds

ESGU is BlackRock's flagship ESG fund and the biggest in the category by assets, at roughly $17 billion. Its Global Warming Potential is 3.7°C on 2025 impact data. Around 4.5% of the portfolio sits in fossil fuel producers. Information technology is 37.9%. The five largest holdings are Apple, NVIDIA, Microsoft, Amazon and Alphabet.

Why the fund is built that way

None of this is hidden and none of it is a failure. BlackRock states the objective directly: track ESG considerations while seeking risk and return characteristics similar to the MSCI USA Index. The exclusions are narrow — civilian firearms, controversial weapons, tobacco, thermal coal and oil sands, most applied at revenue thresholds rather than categorically. The fund does what it says.

Where the gap opens

The gap is between what the fund says and what an investor assumes. Someone who bought ESGU to stop funding climate breakdown bought the US large-cap market with five categories trimmed, on a 3.7°C global warming trajectory. Global Warming Potential is measurable at holding level from ACA Ethos data — how Ziggma's impact data works sets out the methodology.

ESG investing, screening, and impact investing are not the same

ESG investing versus negative screening

These are different methods and the evidence separates them.

Negative screening removes categories from the investable universe — tobacco, controversial weapons, thermal coal. It is the oldest sustainable investing method and the most easily understood. It is also the approach the 2022 review finds performs worst among ESG methods, because exclusion narrows the opportunity set without adding information.

ESG integration uses ratings as an input alongside financial analysis rather than as a filter. It performs better in the literature, and it makes no claim about real-world outcomes.

Positive screening selects for measured performance rather than removing categories. It is the closest ESG method to impact investing.

ESG asks what the world does to a company

Impact investing asks what the company does to the world.

That is single versus double materiality, and it changes what gets measured. An ESG rating scores a tobacco company on governance quality and regulatory exposure management. An impact framework scores the product. The difference between climate-aligned and ESG investing follows the same line.

What impact investing requires

The Global Impact Investing Network defines impact investing by three commitments: intentional measurable benefit alongside financial return, measurement against stated targets, and additionality.

Additionality is harder in public markets

Buying shares transfers money to another investor, not to the company. The public-market mechanisms are indirect — cost of capital, proxy voting, and the signal that persistent reallocation sends about which business models are financeable. The case for public-market impact investing rests on those mechanisms.

What impact screening has that ESG lacks is holding-level outcome data: emissions intensity per unit of output, resource use, labor practices, controversy history.

The record of impact investing

Two datasets, both positive

The impact-screened evidence is narrower than the ESG literature and points more consistently positive.

The window decides the answer

The Clean200 is republished annually with returns restated from the same July 2016 inception. Depending on the edition, the same index has shown a gap of under four percentage points or over sixty.

Later editions state a sustainable-revenue-weighted basis; earlier ones do not specify. Almost all of the outperformance arrived after January 2023.

The authors of the strongest study disclaim it

Schroders warns that caution is needed in interpreting the magnitude of the excess returns, given survivorship bias inherent in the impact portfolio, and states that the tests do not support a causal reading. Further analysis is needed to establish whether the results come from intrinsic qualities or from systematic risk exposures.

Two studies is not a literature

The ESG evidence base runs to over 1,100 papers. The impact-screened evidence base is a fraction of that, and it is younger.

Why screening might produce a real edge

Operational discipline. The measures impact screening captures — emissions intensity per unit of output, resource efficiency, workforce stability, governance accountability — are measures of how well a company is run. The screen may be selecting for management quality using different instruments.

Cost-of-capital repricing. As capital reallocates toward companies meeting transition criteria, their cost of capital falls. That shows up as return during the transition and disappears once it completes. Real but finite.

Transition risk pricing in slowly. Carbon pricing, disclosure mandates and stranded-asset exposure are known risks that markets have priced gradually. Screening ahead of the repricing captures the difference.

None of these is established.

What this means for your portfolio

If you hold individual stocks

None of the fund-level evidence on this page is about your portfolio.

Morgan Stanley's figure describes the median of a 110,000-fund universe. It says nothing about the twenty or thirty positions you actually own. Neither does the Clean200, unless you hold its constituents in its weights.

The unit that matters is the holding

Every position has its own emissions intensity, its own labor record, its own governance history, and its own transition exposure. Those are measurable at the security level, and measuring them is the only way to know what your portfolio does.

Ziggma scores each holding on an Impact Score from 0 to 100, across four sub-scores: Climate Action, Resource Use, Fair Labor, and Accountability. Global Warming Potential expresses a holding's emissions trajectory as a temperature reading. The Controversy Score flags governance and conduct incidents. The Ziggma screener filters on all of them.

If you hold ETFs

The fund-level evidence is about you, and it is positive. Start there.

The difficulty is that the label carries less information than it appears to. Morningstar assigns the sustainable classification on what a fund states in its prospectus. Two ETFs holding materially different companies can both qualify.

Your regional exposure may not be what you think

Around 70% of sustainable funds invest globally or in Europe against 40% of traditional funds. For a fund holder that is not an explanation of past returns, it is a description of current exposure. If your sustainable allocation is doing the work of your international allocation, you may be less diversified than the labels suggest.

Thematic funds concentrate more than they appear to

A clean-energy ETF concentrates a single sector, and holding several can leave a portfolio more exposed than a glance at the ticker count implies. Ziggma's Portfolio Checkup measures this through the Herfindahl-Hirschman Index and reports a Beta Risk Factor alongside it.

Screen the holdings, not the label

Ziggma scores ETFs on the same basis as individual securities: Overall Impact, Climate Action, net-zero target, and Global Warming Potential.