Trend AnalysisEconomics & Finance

ESG Investing Under Scrutiny: Does Climate Risk Management Actually Reduce Risk?

ESG investing has captured $35 trillion in assetsโ€”but does it actually protect portfolios from climate risk? A meta-analysis of over 2,200 studies reveals the uncomfortable truth: the ESG-performance relationship is positive but vanishingly small. We examine what this means for sustainable finance.

By Sean K.S. Shin
This blog summarizes research trends based on published paper abstracts. Specific numbers or findings may contain inaccuracies. For scholarly rigor, always consult the original papers cited in each post.

Environmental, Social, and Governance (ESG) investing has achieved what few financial innovations manage: cultural ubiquity. With over $35 trillion in assets under management globally, ESG has moved from the periphery of socially responsible investing to the mainstream of institutional finance. BlackRock's Larry Fink writes annual letters about stakeholder capitalism. The EU's Sustainable Finance Disclosure Regulation (SFDR) mandates ESG reporting for asset managers. Rating agenciesโ€”MSCI, Sustainalytics, Bloombergโ€”have built billion-dollar businesses scoring companies on ESG criteria. And yet, a fundamental question remains stubbornly unresolved: does ESG investing actually do what it claimsโ€”protect portfolios from climate risk while generating competitive returns?

The evidence, when examined rigorously, is far more equivocal than either ESG advocates or ESG skeptics admit.

The Research Landscape: Thirty Years of Conflicting Evidence

The most ambitious synthesis of ESG-performance evidence to dateโ€”Friede, Busch & the researchers' meta-analysis of over 2,200 studies (note: Krueger, Chrisanctus, Adeoye & Okoye (2024) is a separate survey of institutional investor attitudes, not this meta-analysis)โ€”found that the correlation between ESG and financial performance is generally positive. But the magnitude tells a different story: the average effect size is smallโ€”statistically significant given the enormous sample but economically modest. For a portfolio manager, this is close to noise.

The meta-analytic literature reveals massive heterogeneity: the relationship between ESG and performance depends critically on:

  • Which E, S, or G pillar is measured: Environmental scores show the weakest financial correlation; governance shows the strongest.
  • Which geography: ESG-performance links are stronger in Europe (where regulation creates compliance value) and weaker in the US (where ESG is more voluntary).
  • Which time period: The relationship was positive in 2010โ€“2019 but has weakened since 2020, possibly because ESG alpha has been arbitraged away as AUM scaled.
  • Which ESG rating provider: The same company can receive an "A" from MSCI and a "C" from Sustainalyticsโ€”a divergence that Berg, Kรถlbel & Rigobon (2022) quantify as a correlation of only r = 0.54 between major ESG rating providers.

The Rating Divergence Problem

Berg, Kรถlbel & Rigobon (2022) provide the most rigorous analysis of ESG rating disagreement and its consequences. Using machine learning to detect patterns in rating divergence, they identify three primary sources:

  • Measurement divergence (56% of disagreement): The same category (e.g., "water usage") is operationalized differentlyโ€”absolute usage vs. usage per unit revenue vs. trajectory.
  • Scope divergence (38%): Raters include different categoriesโ€”MSCI weights carbon emissions heavily; Sustainalytics emphasizes labor practices.
  • Weight divergence (6%): Raters assign different importance to the same metrics, reflecting implicit value judgments about which ESG issues matter most.
  • The practical consequence: a portfolio manager who uses ESG ratings for climate risk management is making an implicit bet on which rating agency's methodology best captures actual riskโ€”a meta-decision that most investors never explicitly make.

    Methodological Approaches

    Survey methodology (Ilhan, Krueger et al., 2020): Surveying institutional investors on their views and preferences regarding climate risk disclosure. This approach captures investor demand-side attitudes directly, revealing that institutional investors consider climate risk financially material โ€” though survey responses may overstate actual portfolio behavior.

    ESG rating analysis (Berg, Kรถlbel & Rigobon, 2022): Quantifying the divergence between ESG rating providers by computing inter-rater correlations across major agencies. The low correlation (r = 0.54) challenges the assumption that ESG ratings measure a single, coherent construct โ€” a finding with profound implications for portfolio construction and regulatory reliance on ESG scores.

    AI-enhanced ESG analysis (Chrisanctus, Adeoye & Okoye, 2024): Examining how artificial intelligence can improve ESG investing through enhanced portfolio management, natural language processing of corporate disclosures, and pattern recognition across large-scale ESG datasets.

    Institutional site visits (Song & Xian, 2024): Investigating whether institutional investors' corporate site visits increase firm-level climate change risk disclosure โ€” providing evidence on the mechanisms through which institutional engagement translates into corporate transparency.

    Critical Analysis: Claims and Evidence

    <
    ClaimEvidenceVerdict
    ESG investing generates superior financial returnsFriede, Busch & Bassen (2015) meta-analysis of 2,200+ studies: small positive effectโš ๏ธ Uncertain โ€” statistically positive but economically modest
    Institutional investors consider climate risk materialIlhan, Krueger et al. (2020) survey of institutional investorsโœ… Supported
    ESG rating providers agree on company assessmentsBerg, Kรถlbel & Rigobon (2022): r = 0.54 inter-rater correlationโŒ Refuted โ€” substantial disagreement
    AI can improve ESG analysisChrisanctus et al. (2024): AI-enhanced portfolio managementโœ… Supported โ€” emerging evidence
    ESG investing drives real-world environmental improvementLimited evidence of portfolio-to-impact causal chainโš ๏ธ Uncertain โ€” the mechanism is unclear

    The Uncomfortable Question: Is ESG a Product or a Policy?

    The deepest challenge in this literature is not methodologicalโ€”it is conceptual. ESG investing conflates two fundamentally different objectives: financial risk management (avoiding firms exposed to climate transition or physical risk) and impact generation (channeling capital toward environmentally beneficial activities). These objectives sometimes align and sometimes conflict.

    A fossil fuel company with excellent governance, strong labor practices, and transparent climate risk disclosure may score well on ESG metrics while operating a business model that is fundamentally incompatible with a 1.5ยฐC pathway. Conversely, a clean energy startup may score poorly on governance (weak board oversight, concentrated ownership) while making a genuine contribution to decarbonization. The ESG framework, as currently constructed, cannot distinguish between these casesโ€”and this ambiguity is not a bug but a feature, because it allows the financial industry to market ESG as simultaneously profitable, ethical, and risk-reducing.

    Open Questions and Future Directions

  • Disentangling E, S, and G: Should environmental, social, and governance factors be scored separately rather than aggregated? The evidence suggests the three pillars have differentโ€”sometimes oppositeโ€”relationships with financial performance.
  • Physical risk integration: Current ESG ratings heavily weight transition risk (regulatory, reputational) at the expense of physical risk (flooding, heat stress, water scarcity). Can geospatial climate models improve physical risk assessment?
  • Impact measurement: Can we move beyond portfolio-level ESG scores to measure actual environmental impact per dollar invested? This requires outcome data (tonnes COโ‚‚ avoided, biodiversity preserved) rather than input data (policies adopted, disclosures made).
  • Emerging market ESG: Most ESG data and frameworks reflect OECD institutional environments. How should ESG be adapted for markets where data availability, regulatory enforcement, and cultural expectations differ fundamentally?
  • The active vs. passive debate: ESG engagement (active ownership, shareholder resolutions) may generate more impact than ESG screening (portfolio exclusion). But the evidence base for engagement effectiveness is thin.
  • Implications for Researchers and Investors

    For institutional investors, the evidence demands intellectual honesty: ESG ratings are not reliable predictors of climate risk exposure, and the financial premium associated with ESG is near zero after correcting for publication bias. This does not mean ESG is uselessโ€”but it means investors should specify what they expect ESG to do rather than treating it as an all-purpose solution for risk, return, and impact simultaneously.

    For regulators, the Berg, Kรถlbel & Rigobon findings on rating divergence make a compelling case for standardizationโ€”not of ESG scores themselves (which involve legitimate value judgments) but of the underlying data. If all companies reported the same environmental metrics in the same format, rating divergence driven by measurement differences would shrink substantially.

    For researchers, the most productive direction is the causal chain from investment decision to real-world environmental outcome. The uncomfortable possibility is that ESG investingโ€”as currently practicedโ€”is primarily a signaling mechanism (companies signal virtue to attract capital) rather than a transmission mechanism (capital flows that cause environmental improvement). Proving or disproving this hypothesis would reshape sustainable finance.

    References (4)

    [1] Chrisanctus, O., Adeoye, O.B. & Okoye, C. (2024). Artificial Intelligence in ESG Investing: Enhancing Portfolio Management and Performance. International Journal of Science and Research Archive, 11(1), 0305.
    [2] Berg, F., Kรถlbel, J.F. & Rigobon, R. (2022). Aggregate Confusion: The Divergence of ESG Ratings. Review of Finance, 26(6), 1315โ€“1344.
    [3] Song, Y. & Xian, R. (2024). Institutional Investors' Corporate Site Visits and Firm-Level Climate Change Risk Disclosure. International Review of Financial Analysis, 92, 103145.
    [4] Ilhan, E., Krueger, P. et al. (2020). Institutional Investors' Views and Preferences on Climate Risk Disclosure. SSRN Working Paper.

    Explore this topic deeper

    Search 290M+ papers, detect research gaps, and find what hasn't been studied yet.

    Click to remove unwanted keywords

    Search 8 keywords โ†’