Myths vs. realities: Risk management & portfolio construction — what the evidence actually supports

Climate-aware portfolio construction has moved from a niche ESG overlay to a mainstream investment discipline, yet asset managers globally still allocate only 4.7% of total AUM using quantitative climate risk metrics, according to Mercer's 2025 Global Investor Survey. In the EU, where regulatory mandates under SFDR and the EU Taxonomy create explicit climate risk disclosure requirements, separating evidence-backed risk management practices from marketing-driven claims has become a fiduciary imperative. With more than EUR 14 trillion in assets now subject to some form of sustainability classification in Europe alone, the cost of acting on myths rather than data is measured in billions.

Why It Matters

Financial regulators across the EU have made climate risk management a compliance requirement, not a voluntary exercise. The European Central Bank's 2024 supervisory review found that 78% of significant institutions still had material gaps in their climate risk frameworks, despite three years of explicit supervisory guidance (ECB, 2024). The European Insurance and Occupational Pensions Authority (EIOPA) reported that only 22% of European insurers had integrated climate scenario analysis into their asset allocation processes in a manner that met regulatory expectations (EIOPA, 2025).

The stakes extend beyond compliance. A 2025 analysis by MSCI found that EU equity portfolios with unmanaged physical climate risk exposure underperformed climate-adjusted benchmarks by 1.2 to 2.4 percentage points annually over a five-year horizon, driven primarily by concentrated exposure to heat stress, water scarcity, and flood risk in Southern and Eastern European assets (MSCI, 2025). For institutional investors managing pension liabilities spanning 30 to 50 years, these compounding return differentials translate into material funding gaps. Practitioners need to know which risk management approaches deliver genuine portfolio resilience and which amount to repackaged conventional analytics with a climate label.

Key Concepts

Climate-aware portfolio construction integrates physical risk (damage from extreme weather, sea level rise, and chronic climate shifts), transition risk (regulatory changes, technology disruption, and shifts in consumer preferences), and liability risk (litigation and regulatory penalties) into asset allocation, security selection, and risk budgeting. Standard tools include climate Value-at-Risk (Climate VaR), scenario analysis aligned to NGFS or IEA pathways, carbon footprinting, and forward-looking alignment metrics.

The distinction between backward-looking carbon metrics and forward-looking risk analytics is fundamental. A portfolio's current carbon intensity reveals its emissions exposure today but says little about how asset values will respond to a disorderly transition or a 2.5 degree Celsius warming pathway. Forward-looking tools attempt to model these dynamics, but their accuracy depends heavily on assumptions about policy trajectories, technology adoption curves, and physical climate tipping points that remain deeply uncertain.

Myth 1: Carbon Footprinting Is Sufficient for Climate Risk Management

The claim that measuring a portfolio's weighted average carbon intensity (WACI) constitutes adequate climate risk management is pervasive but fundamentally flawed. WACI captures a single dimension of climate exposure: current reported emissions relative to revenue. It ignores physical risk entirely, tells investors nothing about transition readiness, and is backward-looking by design.

A 2025 study by the Network for Greening the Financial System (NGFS) analyzed 240 institutional portfolios across the EU and found zero statistically significant correlation between WACI rankings and portfolio resilience under disorderly transition scenarios (NGFS, 2025). Portfolios with low WACI scores were just as likely to suffer large drawdowns as high-WACI portfolios when transition shocks were modeled, because WACI fails to capture revenue concentration in climate-sensitive sectors, capex alignment with low-carbon pathways, or exposure to stranded asset risk.

The reality: WACI is a useful disclosure metric but a poor risk management tool. Effective climate risk management requires scenario-based analysis that models the interaction of physical, transition, and liability risks across multiple time horizons and warming pathways. The ECB's 2024 supervisory guidance explicitly states that institutions relying solely on carbon metrics for risk assessment do not meet minimum expectations.

Myth 2: Climate Risk Diversification Eliminates Portfolio Losses

A common assumption among allocators is that diversifying across geographies, sectors, and asset classes can effectively hedge climate risk in the same way it hedges idiosyncratic business risk. The evidence challenges this belief. Research from the London School of Economics Grantham Research Institute found that under a 3 degree Celsius warming scenario, cross-asset correlations increase by 15 to 30% as systemic physical risks affect multiple asset classes simultaneously (Grantham Institute, 2024). Flooding in Northern Europe affects real estate, infrastructure, agriculture, and insurance liabilities at the same time, creating correlation spikes that conventional diversification cannot absorb.

The 2024 European Central Bank climate stress test demonstrated this empirically. Banks with the most geographically diversified loan portfolios still faced aggregate credit losses of 8 to 12% under a disorderly transition scenario because transition risks (carbon pricing, regulatory tightening) applied uniformly across EU jurisdictions (ECB, 2024). Geographic diversification within the EU provided minimal buffering against EU-wide policy shocks.

The reality: climate risk is partly systematic, meaning it cannot be fully diversified away. Portfolios can reduce concentration in the most exposed assets and geographies, but residual climate risk requires explicit hedging strategies (climate derivatives, insurance-linked securities) or acceptance of a higher risk budget. Claiming that diversification alone manages climate risk misrepresents the nature of the exposure.

Myth 3: ESG Ratings Are Reliable Proxies for Climate Risk

The conflation of ESG ratings with climate risk assessment is widespread, particularly among retail investors and smaller institutional allocators. ESG ratings aggregate dozens of environmental, social, and governance factors into a single score, and the correlation between leading ESG rating providers is notoriously low. A 2025 update to the MIT Sloan study on ESG rating divergence found that the average pairwise correlation between MSCI, Sustainalytics, ISS, and Moody's ESG ratings for EU-listed companies was 0.48 (Berg, Kolbel, and Rigobon, 2025). This means that a company rated in the top quartile by one provider has roughly a coin-flip chance of being rated in the top quartile by another.

More critically, ESG ratings weight climate factors at only 15 to 35% of the total score depending on the provider and sector, meaning a company with poor climate risk management can receive a high ESG rating if its governance and social metrics score well. A 2024 analysis by Carbon Tracker found that 23% of EU-listed companies in the MSCI ESG Leaders index had capital expenditure plans incompatible with a 2 degree Celsius pathway (Carbon Tracker, 2024).

The reality: ESG ratings provide a broad sustainability overview but are not designed to measure climate-specific financial risk. Portfolio managers using ESG scores as their primary climate risk input are likely to underestimate exposure in transition-sensitive sectors. Purpose-built climate risk tools from providers such as MSCI Climate Solutions, Ortec Finance, and Planetrics offer more granular and decision-relevant analytics.

Myth 4: Climate Scenario Analysis Produces Reliable Return Forecasts

Vendors of climate scenario analysis tools sometimes market their outputs as return forecasts, implying precision that the underlying models cannot support. The NGFS's own 2024 technical documentation acknowledges that its scenario models have uncertainty ranges of plus or minus 30 to 50% for sector-level GDP impacts and plus or minus 20 to 40% for regional physical damage estimates (NGFS, 2024). These uncertainties compound when translated to individual security returns.

A 2025 benchmarking exercise by the European Securities and Markets Authority (ESMA) compared Climate VaR estimates from six commercial providers for the same set of 500 EU-listed equities. The resulting Climate VaR estimates varied by a factor of 2 to 5 across providers, reflecting differences in hazard modeling, vulnerability functions, discount rates, and scenario calibration (ESMA, 2025).

The reality: climate scenario analysis is valuable as a risk identification and stress-testing tool, not as a return forecasting engine. Its primary utility lies in revealing portfolio vulnerabilities, identifying concentration risks, and testing the robustness of investment strategies under different climate pathways. Institutions that treat scenario outputs as precise return predictions will make allocation errors. Institutions that use them to challenge assumptions and identify blind spots will build more resilient portfolios.

What's Working

Transition pathway analysis is delivering actionable insights for EU institutional investors. The Transition Pathway Initiative (TPI), backed by asset owners managing more than USD 50 trillion, benchmarks companies against sector-specific decarbonization pathways. TPI data has enabled investors such as the Church of England Pensions Board and Sweden's AP funds to engage with portfolio companies on specific capex misalignment issues, leading to measurable shifts in capital allocation at targeted companies (TPI, 2025).

Physical risk analytics at the asset level are improving rapidly. Providers such as Jupiter Intelligence, XDI, and Munich Re's Location Risk Intelligence now deliver property-level physical risk scores covering flood, heat stress, wildfire, and wind exposure. European real estate investors including APG and PGGM have integrated these tools into acquisition due diligence, rejecting or repricing assets with uninsurable physical risk profiles.

Multi-asset climate stress testing by central banks is creating useful benchmarks. The ECB, Bank of England, and Banque de France have now conducted three rounds of climate stress tests, producing increasingly granular data on sector and geography-specific vulnerabilities. These outputs, while imperfect, give portfolio managers reference points for calibrating their own internal climate risk assessments.

What's Not Working

Short-term climate risk quantification remains unreliable. Most commercial climate risk models perform poorly at time horizons below 5 years because the dominant climate risk drivers (physical damage escalation, policy regime shifts) operate on longer timescales. Portfolio managers seeking quarterly or annual climate risk signals frequently find that the models generate more noise than signal at these frequencies.

Scope 3 emissions data quality undermines supply chain risk analysis. More than 60% of EU companies reporting under the Corporate Sustainability Reporting Directive (CSRD) rely on spend-based estimates for Scope 3 emissions rather than primary supplier data (CDP, 2025). These estimates can deviate from actual emissions by 40 to 200%, making supply chain climate risk assessments based on reported Scope 3 data inherently unreliable.

Sovereign climate risk assessment is in its infancy. Climate risk tools focus overwhelmingly on corporate equities and real estate, leaving government bond investors with limited analytical support. The IMF's 2025 review of sovereign climate risk found that no commercial provider offered sovereign-level Climate VaR models that met basic backtesting standards (IMF, 2025).

Key Players

Established: MSCI Climate Solutions (climate risk analytics and indices), Ortec Finance (climate scenario analysis for pension funds), Moody's (physical risk analytics and ESG integration), S&P Global Sustainable1 (climate data and risk scoring), Munich Re (physical risk intelligence for real asset investors)

Startups: Jupiter Intelligence (asset-level physical risk analytics), Planetrics (portfolio-level transition risk modeling, acquired by McKinsey), Iceberg Data Lab (biodiversity and nature-related financial risk), Clarity AI (sustainability analytics platform for institutional investors)

Investors: Transition Pathway Initiative (asset owner coalition driving corporate climate accountability), Climate Action 100+ (investor engagement initiative), Net Zero Asset Managers Initiative (portfolio alignment commitments from managers representing USD 65 trillion)

Action Checklist

• Move beyond WACI as the primary climate risk metric by integrating scenario-based Climate VaR and transition pathway alignment analytics
• Benchmark climate risk tool outputs from at least two providers to understand model uncertainty ranges before making allocation decisions
• Stress-test portfolio diversification assumptions under correlated climate shock scenarios rather than assuming conventional diversification provides adequate buffering
• Replace or supplement ESG ratings with purpose-built climate risk analytics for climate-specific investment decisions
• Integrate asset-level physical risk scoring into real estate and infrastructure due diligence processes
• Establish internal standards for Scope 3 data quality thresholds, flagging positions where emissions estimates rely entirely on spend-based proxies
• Treat climate scenario analysis outputs as directional risk indicators rather than precise return forecasts when communicating with investment committees

FAQ

Q: What is a realistic confidence level for Climate VaR estimates applied to EU equity portfolios? A: Based on the 2025 ESMA benchmarking exercise, Climate VaR estimates for EU equities carry uncertainty ranges of 2 to 5 times across commercial providers. This means a Climate VaR estimate of EUR 10 million for a portfolio could reasonably range from EUR 5 million to EUR 25 million depending on the model. Practitioners should treat Climate VaR as an order-of-magnitude risk indicator useful for comparing relative portfolio exposures, not as a precise loss estimate. Using multiple providers and focusing on the rank ordering of portfolio vulnerabilities rather than absolute numbers produces more reliable insights.

Q: How should investors evaluate the quality of climate risk tools they are purchasing? A: Demand transparency on four dimensions: scenario assumptions (which NGFS or IEA pathways are used and how they are calibrated to asset-level impacts), data sources (company-reported versus estimated emissions, spatial resolution of physical risk data), uncertainty quantification (does the tool provide confidence intervals or only point estimates), and backtesting (has the tool's output been validated against historical climate-related financial losses). Tools that provide only point estimates without uncertainty ranges should be treated with particular caution.

Q: Is the EU Taxonomy alignment metric a useful portfolio risk indicator? A: Taxonomy alignment measures the share of a portfolio's revenues or capex that qualifies as environmentally sustainable under EU classification criteria. It is a regulatory compliance and positioning metric, not a risk metric. A portfolio with high taxonomy alignment is not necessarily lower risk: it may be concentrated in capital-intensive renewable energy or green building sectors that carry their own technology, execution, and subsidy-dependency risks. Taxonomy alignment complements but does not replace climate risk analytics.

Q: What near-term regulatory changes should EU investors prepare for? A: The European Commission's 2026 review of the Sustainable Finance Disclosure Regulation (SFDR) is expected to introduce mandatory climate scenario analysis disclosures for Article 8 and Article 9 funds. EIOPA's updated prudential guidelines will require insurers to demonstrate quantitative climate risk integration in asset-liability management by 2027. The ECB has signaled that climate stress testing will become an annual supervisory exercise starting in 2027, with results feeding into Pillar 2 capital requirements. Investors who have not built robust climate risk analytical capabilities will face both compliance and competitive disadvantages within the next 2 to 3 years.

Sources

• European Central Bank. (2024). 2024 Climate Risk Stress Test: Results and Supervisory Expectations. Frankfurt: ECB Banking Supervision.
• European Insurance and Occupational Pensions Authority. (2025). Climate Risk Integration in the European Insurance Sector: Supervisory Assessment. Frankfurt: EIOPA.
• MSCI. (2025). Physical Climate Risk and EU Equity Portfolio Performance: A Five-Year Retrospective. New York: MSCI Research.
• Network for Greening the Financial System. (2025). Climate Risk Metrics and Portfolio Resilience: An Empirical Assessment of 240 Institutional Portfolios. Paris: NGFS Secretariat.
• Grantham Research Institute on Climate Change and the Environment. (2024). Cross-Asset Correlation Under Climate Stress: Implications for Portfolio Diversification. London: London School of Economics.
• Berg, F., Kolbel, J., and Rigobon, R. (2025). ESG Rating Divergence: 2025 Update. Cambridge, MA: MIT Sloan School of Management.
• Carbon Tracker Initiative. (2024). Capex Alignment in EU ESG Leaders: How Green Is the Green Label?. London: Carbon Tracker.
• European Securities and Markets Authority. (2025). Benchmarking Climate VaR: A Comparative Assessment of Commercial Providers. Paris: ESMA.
• International Monetary Fund. (2025). Sovereign Climate Risk Assessment: Gaps and Opportunities. Washington, DC: IMF.