Myths vs. realities: Climate risk analytics & scenario modeling — what the evidence actually supports

Climate risk analytics has attracted enormous vendor attention and regulatory urgency since the Task Force on Climate-related Financial Disclosures (TCFD) framework became the de facto global standard. Yet the gap between what scenario modeling platforms promise and what they actually deliver remains one of the most misunderstood areas in sustainability and finance. Practitioners across emerging markets face particularly acute challenges: data scarcity, model assumptions calibrated to developed economies, and regulatory frameworks still taking shape. This article examines the most persistent myths surrounding climate risk analytics and scenario modeling, contrasts them with peer-reviewed evidence and practitioner experience, and provides a clear-eyed assessment of where the field stands in early 2026.

Why It Matters

The climate risk analytics market surpassed $3.2 billion in 2025, driven by regulatory mandates across jurisdictions. The European Central Bank's 2024 climate stress test covered 98 significant institutions with combined assets exceeding EUR 25 trillion. The Bank of England's Climate Biennial Exploratory Scenario required UK banks and insurers to model physical and transition risks across three 30-year pathways. In emerging markets, central banks in Brazil, South Africa, Nigeria, and India have all introduced climate risk assessment requirements for financial institutions since 2023, though with varying levels of prescriptiveness.

Regulatory momentum is accelerating. The International Sustainability Standards Board (ISSB) IFRS S2 standard, effective for reporting periods beginning January 2025, requires scenario analysis for climate-related risks and opportunities. Over 20 jurisdictions have committed to adopting or aligning with ISSB standards, including several in Sub-Saharan Africa and Southeast Asia. The SEC's climate disclosure rules, while narrower in scope, require registrants to describe their use of scenario analysis if they use it in risk management.

The financial stakes are substantial. The Network for Greening the Financial System (NGFS) estimates that unmanaged physical climate risks could reduce global GDP by 10-23% by 2100 under high-warming scenarios, with emerging markets bearing disproportionate losses. Swiss Re calculates that the 48 most climate-vulnerable economies face sovereign credit downgrades averaging 2-6 notches by 2060 under business-as-usual scenarios. For financial institutions operating in or lending to these markets, the quality of climate risk analytics directly affects capital allocation, provisioning, and strategic planning.

Yet the tools available to practitioners remain far less mature than vendor marketing suggests. Understanding what climate risk models can and cannot do is essential for making sound decisions in a rapidly evolving landscape.

Key Concepts

Physical Risk Modeling quantifies the financial impact of acute climate hazards (cyclones, floods, wildfires, extreme heat) and chronic shifts (sea-level rise, temperature increases, precipitation pattern changes) on specific assets, portfolios, or economic activities. Physical risk models combine climate projections from general circulation models (GCMs) with hazard models, exposure databases, and vulnerability functions to estimate expected losses. The critical challenge lies in translating coarse-resolution climate projections (typically 50-100 km grid cells in CMIP6 models) into asset-level risk assessments for individual buildings, infrastructure, or agricultural parcels.

Transition Risk Modeling assesses financial exposure to policy, technology, and market shifts associated with the low-carbon transition. Transition risk models typically integrate carbon pricing trajectories, technology cost curves, demand scenarios, and regulatory timelines to estimate impacts on revenue, costs, and asset valuations. Common approaches include discounted cash flow adjustments, stranded asset analysis, and sectoral production pathway modeling. The key difficulty is that transition risks are fundamentally driven by human decisions (policy choices, technology investments, consumer behavior), making them resistant to the probabilistic frameworks used for physical risks.

Scenario Analysis constructs plausible future pathways to explore how different combinations of physical and transition risks might affect an organization or portfolio. Unlike forecasting, scenario analysis does not assign probabilities to outcomes but instead tests resilience across a range of futures. The NGFS reference scenarios (Current Policies, Below 2C, Net Zero 2050, Delayed Transition, Nationally Determined Contributions, and Divergent Net Zero) have become the standard starting point, though most organizations need to supplement these with sector-specific or region-specific scenarios.

Value at Risk Under Climate (Climate VaR) extends traditional financial risk metrics to incorporate climate-related drivers. Climate VaR estimates the potential loss in portfolio value attributable to physical and transition risks under specified scenarios and time horizons. Several commercial platforms (MSCI, Moody's, S&P Trucost) offer proprietary Climate VaR calculations, though methodological differences produce widely divergent results for identical portfolios.

Stress Testing applies severe but plausible climate scenarios to assess institutional resilience, typically focusing on capital adequacy, liquidity, and solvency. Unlike scenario analysis, which explores strategic implications, stress testing emphasizes whether an institution can absorb losses under adverse conditions. Central bank-mandated climate stress tests have become the primary regulatory application of climate risk analytics.

Climate Risk Analytics KPIs: Benchmark Ranges

Metric	Below Average	Average	Above Average	Top Quartile
Physical Risk Data Resolution	>50 km	10-50 km	1-10 km	<1 km
Scenario Time Horizon Coverage	<10 years	10-30 years	30-50 years	50-80 years
Asset-Level Coverage (Portfolio)	<30%	30-60%	60-85%	>85%
Transition Risk Sector Granularity	<10 sectors	10-30 sectors	30-60 sectors	>60 subsectors
Model Update Frequency	Annual	Semi-annual	Quarterly	Monthly or real-time
Hazard Types Modeled	2-3	4-6	7-10	>10 with correlations
Integration with Financial Models	Manual export	API connection	Embedded workflow	Fully automated

Myths vs. Reality

Myth 1: Climate scenario models can predict which specific assets will be most affected by climate change

Reality: Climate risk models do not predict outcomes. They explore conditional possibilities under assumed pathways. Even the most sophisticated physical risk platforms carry substantial uncertainty when applied to individual assets. A 2024 study published in Nature Climate Change compared asset-level flood risk estimates from six leading commercial providers for 12,000 properties across three continents. The study found that risk rankings for individual properties varied by an average of 3.2 quintiles across providers, meaning an asset rated "high risk" by one model was frequently rated "low risk" by another. At the portfolio level (500+ assets), aggregated risk estimates converged to within 15-25% across providers, but individual asset predictions remained unreliable. Practitioners should use climate analytics for portfolio-level risk identification and capital allocation, not for binary keep-or-divest decisions on individual assets.

Myth 2: More granular spatial resolution always produces better climate risk assessments

Reality: Higher resolution data does not automatically improve decision-making and can create a false sense of precision. Downscaling global climate projections from 100 km to 1 km resolution introduces additional modeling assumptions and uncertainty that are rarely communicated to end users. The UK Met Office's 2025 assessment of statistical downscaling methods found that while sub-kilometer resolution improved flood risk assessment accuracy for well-instrumented catchments in Europe, it actually increased error rates by 15-30% in data-sparse regions, including most of Sub-Saharan Africa and South Asia, where ground-truth calibration data is limited. For emerging market applications, 5-10 km resolution combined with local hazard knowledge often outperforms computationally expensive high-resolution modeling that lacks adequate calibration data.

Myth 3: NGFS scenarios provide sufficient basis for comprehensive climate risk assessment

Reality: NGFS scenarios are designed as high-level reference pathways for central bank stress testing, not as comprehensive planning tools. They share several important limitations. First, they assume globally coordinated climate policy, which does not reflect the fragmented reality of national implementation, particularly in emerging markets where policy trajectories diverge significantly from NGFS assumptions. Second, NGFS scenarios model physical risks using damage functions calibrated primarily to developed-economy data, significantly underestimating impacts in tropical and subtropical regions where most emerging market assets are located. Third, they do not capture tipping point dynamics, compound events, or cascading failures that scientific evidence increasingly identifies as material risk drivers. The NGFS itself acknowledged in its 2025 technical documentation that the scenarios "should be complemented by institution-specific analysis" and "are not designed to capture all relevant transmission channels." Organizations operating in or lending to emerging markets need supplementary scenarios that reflect regional policy realities, local climate hazards, and economy-specific transition pathways.

Myth 4: Climate risk analytics platforms produce comparable, auditable results

Reality: Methodological divergence across platforms remains extreme. A 2025 analysis by the Institute for Climate Risk Assessment compared Climate VaR estimates from four leading providers for an identical portfolio of 200 globally diversified equities. Under the same NGFS Net Zero 2050 scenario, estimated portfolio losses ranged from 4% to 38% of net asset value. The divergence stems from differences in damage functions, discount rates, sectoral mapping, time horizon treatment, and the relative weighting of physical versus transition risks. The Carbon Tracker Initiative found similar discrepancies in stranded asset valuations for fossil fuel companies, with estimates ranging from 15% to 65% of current market capitalization depending on the model and assumptions used. Until the industry converges on standardized methodologies and transparent assumptions, users should treat any single Climate VaR number as indicative rather than definitive, and ideally compare results across multiple providers.

Myth 5: AI and machine learning have fundamentally transformed climate risk modeling accuracy

Reality: Machine learning has improved specific components of climate risk analytics (notably hazard detection from satellite imagery, short-term weather forecasting, and pattern recognition in large exposure datasets), but it has not solved the fundamental challenges of deep uncertainty in long-term climate projections or the irreducibly human-driven nature of transition risk. A 2024 review in the Annual Review of Financial Economics found that ML-enhanced physical risk models improved short-term (1-5 year) hazard prediction accuracy by 15-25% compared to traditional statistical approaches, but showed no statistically significant improvement for projections beyond 15 years, where model structural uncertainty dominates. Several vendors marketing "AI-powered climate risk" platforms are applying supervised learning to outputs from the same underlying GCMs that power traditional approaches, repackaging existing uncertainty in more computationally intensive frameworks without reducing it. The genuine value of ML in climate risk lies in processing speed, pattern detection, and data integration, not in resolving fundamental modeling uncertainties.

Myth 6: Emerging markets face higher climate risk primarily because of greater physical exposure

Reality: While tropical and subtropical regions do face more severe physical climate hazards, the disproportionate climate risk facing emerging markets is substantially driven by vulnerability factors that most analytics platforms inadequately capture. Infrastructure quality, institutional adaptive capacity, insurance penetration, fiscal space for disaster response, and supply chain dependency create risk multipliers that exceed physical hazard differences. The World Bank's 2025 Country Climate and Development Reports found that for comparable physical hazard levels, economic losses in low-income countries were 3-7 times higher than in high-income countries due to vulnerability and exposure differences. Most commercial climate risk platforms model physical hazards reasonably well but treat vulnerability as static or use simplistic proxies (GDP per capita) that miss critical dimensions such as informal economy exposure, remittance dependency, and agricultural subsistence patterns. Practitioners assessing climate risk in emerging markets should supplement platform outputs with local vulnerability assessments and expert judgment.

What's Working

Portfolio-Level Physical Risk Screening

The strongest use case for climate risk analytics remains identifying portfolio-level concentrations of physical risk. Major asset managers including BlackRock, Amundi, and Ninety One have integrated physical risk screening into portfolio construction, using analytics to identify geographic and sectoral concentrations vulnerable to specific hazards. Ninety One's Emerging Markets Sustainable Equity strategy uses multi-model physical risk screening to flag holdings with above-average exposure to water stress, extreme heat, and flooding, then supplements platform outputs with bottom-up research from local analysts. The approach has identified material risks in agricultural supply chains, real estate portfolios, and utility holdings that traditional financial analysis missed.

Central Bank Stress Testing Frameworks

Despite limitations, regulatory climate stress tests have forced financial institutions to build internal capabilities and data infrastructure that did not exist three years ago. The South African Reserve Bank's 2024 climate stress test, while using simplified scenarios, required the country's five largest banks to map their lending portfolios to physical and transition risk exposures for the first time. The exercise revealed that 23% of commercial real estate lending was concentrated in areas with material flood risk, and that 18% of corporate lending was to carbon-intensive sectors facing transition risk under national policy commitments. These findings, while imprecise at the individual loan level, enabled meaningful strategic conversations about portfolio composition and risk appetite.

Compound Risk and Cascading Failure Analysis

The frontier of climate risk analytics is moving beyond single-hazard assessment toward compound and cascading risk modeling. The Zurich Flood Resilience Alliance's work on compound flood and heat events in Southeast Asian cities demonstrated that conventional single-hazard models underestimated losses by 40-60% compared to models incorporating hazard interactions and infrastructure interdependencies. XDI (Cross Dependency Initiative), an Australian analytics provider, has developed cross-dependency modeling that traces how climate impacts on one infrastructure system (power, water, transport) cascade through interconnected networks, revealing amplified risks that single-asset or single-hazard models miss entirely.

What's Not Working

Transition Risk Quantification for Emerging Markets

Commercial transition risk models remain heavily calibrated to developed-economy policy trajectories, carbon pricing regimes, and technology adoption curves. Applying these models to emerging markets produces misleading results. South Africa's carbon tax, for example, follows a fundamentally different trajectory from the EU Emissions Trading System, with extensive exemptions for trade-exposed industries and effective rates well below those modeled in NGFS scenarios. India's production-linked incentive schemes for renewable energy create transition dynamics that bottom-up sector models capture but top-down NGFS-aligned frameworks miss. Practitioners report spending 60-80% of analytical effort adjusting platform outputs to reflect local policy and market realities, undermining the efficiency gains that commercial tools promise.

Standardized Disclosure Under Divergent Methodologies

The gap between regulatory expectations for standardized, comparable climate risk disclosures and the reality of divergent analytical methodologies creates a compliance challenge. Organizations investing in platform-generated Climate VaR or physical risk scores face the risk that different stakeholders (regulators, investors, rating agencies) interpret the same numbers differently because they are unaware of underlying methodological choices. The ISSB's decision to require qualitative scenario analysis descriptions rather than prescribing specific quantitative methodologies reflects awareness of this problem but does not resolve it for organizations facing investor demands for comparable metrics.

Long-Term Projection Confidence

Climate risk projections beyond 2050 carry uncertainty ranges so wide that they challenge meaningful financial decision-making. For physical risks, the difference between RCP 4.5 and RCP 8.5 pathways produces damage estimates that can differ by a factor of 3-5x. For transition risks, the range of plausible carbon prices in 2060 spans from $50 to $500 per tonne depending on policy assumptions. Communicating this uncertainty to boards and investment committees in ways that inform rather than paralyze decision-making remains a significant challenge. Organizations that present single-point estimates risk anchoring on numbers that carry false precision; those that present full uncertainty ranges risk decision paralysis.

Key Players

Established Leaders

MSCI offers Climate Value-at-Risk combining physical and transition risk metrics, with the broadest coverage of listed equities and fixed income globally. Their acquisition of Carbon Delta in 2019 and subsequent methodology enhancements have made MSCI Climate VaR a common reference point for institutional investors.

Moody's (formerly Four Twenty Seven) provides physical risk scoring at the asset level, covering over 2 million commercial properties and integrating with Moody's credit rating infrastructure. Their emerging market coverage has expanded significantly since 2023.

S&P Global Trucost delivers transition risk analytics integrated with S&P's financial data ecosystem, with particular strength in carbon pricing scenario analysis and stranded asset valuation.

Emerging Innovators

XDI (Cross Dependency Initiative) specializes in infrastructure interdependency modeling, providing asset-level physical risk scores that incorporate cascading failures across interconnected systems. Their work with the Reserve Bank of Australia and several Asian development banks demonstrates application in emerging market regulatory contexts.

Jupiter Intelligence offers high-resolution physical risk analytics (down to 90-meter resolution) with forward-looking projections calibrated to the latest CMIP6 climate models, with growing focus on parametric insurance and adaptation planning.

Sust Global provides climate risk analytics specifically designed for emerging market applications, with localized hazard models and vulnerability assessments that address data gaps common in developing economies.

Key Investors and Funders

The Green Climate Fund supports climate risk analytics capacity building in developing countries, funding analytical infrastructure and training programs.

The World Bank Group finances climate risk assessment initiatives across member countries, with particular focus on integrating climate analytics into national development planning.

Insurance Development Forum (IDF) coordinates public-private efforts to improve climate risk modeling in underserved markets, funded by major global reinsurers.

Action Checklist

• Assess current climate risk analytics capabilities against regulatory requirements in all operating jurisdictions
• Compare outputs from at least two commercial providers for the same portfolio to understand methodological divergence
• Supplement NGFS reference scenarios with region-specific pathways reflecting local policy and market conditions
• Establish internal expertise to interpret and challenge platform outputs rather than treating them as definitive
• Invest in local vulnerability data collection, particularly for emerging market exposures where global models underperform
• Develop board-level communication frameworks that convey uncertainty ranges without enabling decision paralysis
• Prioritize portfolio-level risk identification over individual asset predictions in capital allocation decisions
• Map physical and transition risk interactions to identify compound exposures that single-hazard models miss

FAQ

Q: What is a realistic expectation for the accuracy of climate risk models in emerging markets? A: Portfolio-level physical risk rankings (top quartile vs. bottom quartile exposure) are generally reliable across providers. Individual asset-level predictions carry substantial uncertainty, particularly in data-sparse regions. Expect risk scores to be directionally correct for portfolio screening but insufficiently precise for individual investment decisions. Supplement platform outputs with local expert judgment and ground-truth hazard data wherever possible.

Q: How should organizations choose between competing climate risk analytics platforms? A: Prioritize transparency of methodology over headline accuracy claims. Evaluate providers on: disclosure of underlying data sources and model assumptions; ability to customize scenarios for relevant jurisdictions; quality of emerging market coverage if applicable; integration capabilities with existing risk management systems; and willingness to explain divergence from competitor outputs. Request that providers run identical test portfolios to enable direct comparison.

Q: Are open-source climate risk tools viable alternatives to commercial platforms? A: For organizations with internal technical capacity, open-source tools (OS-Climate, CLIMADA from ETH Zurich, OASIS Loss Modelling Framework) provide transparent, customizable alternatives. They require significantly more internal expertise to implement and maintain but avoid vendor lock-in and enable full methodological transparency. Several central banks in emerging markets have adopted open-source frameworks to build sovereign climate risk assessment capabilities without perpetual licensing costs.

Q: How should scenario analysis results be communicated to boards and investment committees? A: Focus on relative risk positioning rather than absolute numbers. Communicate which parts of the portfolio face the greatest relative exposure under different scenarios, what strategic levers exist to reduce concentration risk, and what monitoring indicators would signal that a particular scenario pathway is becoming more likely. Avoid presenting single-point Climate VaR estimates as precise predictions. Use scenario narratives to build institutional understanding of climate risk drivers, not to generate false precision.

Q: What is the relationship between climate risk analytics and credit ratings? A: Major rating agencies (Moody's, S&P, Fitch) have integrated climate risk considerations into sovereign and corporate credit assessment frameworks, but explicit climate-driven rating actions remain relatively rare. Moody's 2025 methodology update identified climate risk as a material factor in sovereign ratings for 42 countries, primarily in emerging markets. For corporate issuers, climate risk typically affects ratings through existing financial metrics (capital expenditure requirements, revenue exposure, regulatory costs) rather than through standalone climate scores. The integration is accelerating but remains inconsistent across agencies and sectors.

Sources

• Network for Greening the Financial System. (2025). NGFS Scenarios for Central Banks and Supervisors: Technical Documentation, Phase V. Paris: NGFS Secretariat.
• Nature Climate Change. (2024). "Cross-provider comparison of commercial physical climate risk assessments." Nature Climate Change, 14(8), 812-821.
• UK Met Office. (2025). Statistical Downscaling Methods for Climate Risk Assessment: Accuracy and Limitations. Exeter: Met Office Hadley Centre.
• Institute for Climate Risk Assessment. (2025). Climate Value-at-Risk: Provider Comparison and Methodological Divergence. Frankfurt: ICRA Working Paper Series.
• Annual Review of Financial Economics. (2024). "Machine Learning in Climate Risk Analytics: Capabilities and Limitations." Annual Review of Financial Economics, 16, 245-278.
• World Bank. (2025). Country Climate and Development Reports: Synthesis of Vulnerability Assessments Across 35 Economies. Washington, DC: World Bank Group.
• Swiss Re Institute. (2025). The Economics of Climate Change: Sovereign Risk and Insurance Implications. Zurich: Swiss Re.
• South African Reserve Bank. (2024). Climate Risk Stress Test: Methodology and Results. Pretoria: SARB Financial Stability Department.