Myths vs. realities: Climate risk stress testing & scenario regulation — what the evidence actually supports

A 2025 survey by the Network for Greening the Financial System (NGFS) found that 46 central banks and financial supervisors across 28 jurisdictions have now conducted or mandated climate stress tests, up from just 8 in 2020. Yet the Bank for International Settlements estimates that fewer than 15% of these exercises have led to measurable changes in capital allocation or risk pricing within the institutions tested (BIS, 2025). The gap between regulatory ambition and practical impact has created fertile ground for myths about what climate stress testing can and cannot achieve, particularly in emerging markets where supervisory capacity and data infrastructure lag behind the pace of regulation.

Why It Matters

Climate risk stress testing sits at the intersection of financial regulation, climate science, and corporate strategy. Central banks from Brazil to South Africa to the Philippines have launched or announced climate stress testing frameworks in the past two years, influenced by the European Central Bank's 2022 pioneering exercise and the Bank of England's 2021 Climate Biennial Exploratory Scenario (CBES). For executives in emerging market financial institutions, understanding which claims about stress testing are grounded in evidence and which are overstated is essential for allocating compliance budgets, interpreting supervisory feedback, and integrating climate risk into strategic planning.

The stakes are significant. The International Monetary Fund estimates that climate-related financial losses in emerging markets could reach $1.3 trillion annually by 2030 under a delayed-transition scenario, with banking sector losses concentrated in fossil fuel exposures, agriculture, and real estate (IMF, 2025). Stress testing is one of the primary tools regulators are deploying to force institutions to quantify these exposures. Getting the exercise right matters for financial stability; getting it wrong wastes resources and creates false confidence.

Key Concepts

Climate stress testing applies forward-looking climate scenarios (typically physical risk and transition risk pathways) to a financial institution's balance sheet to estimate potential losses over horizons of 5 to 30 years. Unlike traditional stress tests that model 1 to 3 year horizons with established statistical distributions, climate stress tests must grapple with deep uncertainty, non-linear tipping points, and time horizons that exceed the maturity of most financial assets.

Scenario analysis uses a set of plausible future states (commonly aligned with NGFS scenarios such as Net Zero 2050, Delayed Transition, and Current Policies) to explore how different climate pathways would affect portfolios. Scenario analysis is exploratory rather than predictive: it does not assign probabilities to outcomes.

Transition risk captures the financial impact of policy changes, technology shifts, and market repricing during the shift to a low-carbon economy. Physical risk captures the direct impact of climate hazards (floods, droughts, extreme heat, sea-level rise) on asset values and creditworthiness.

Double materiality is the principle, embedded in the EU's Corporate Sustainability Reporting Directive (CSRD), that institutions should assess both how climate risks affect the institution and how the institution's activities affect the climate.

Myth vs. Reality Breakdown

Myth	Reality	Evidence
Climate stress tests produce precise loss estimates	Outputs are order-of-magnitude indicators, not precise forecasts	ECB 2022 results showed a 3x range across banks with similar portfolios
Passing a climate stress test means the institution is climate-resilient	Current tests assess a narrow set of exposures and miss systemic interactions	Bank of England CBES found significant gaps in second-order effects modeling
Emerging markets can adopt EU/UK frameworks directly	Local climate hazards, data gaps, and economic structures require adaptation	Banco Central do Brasil redesigned NGFS scenarios for tropical agriculture exposures
Longer time horizons always produce more useful results	Beyond 10 years, compound uncertainty often renders results uninformative	NGFS 2025 guidance recommends focusing on 5-10 year decision-relevant horizons
AI and machine learning solve the data gap problem	ML models amplify existing biases when trained on incomplete climate data	BIS research showed ML-based credit models underperformed simple heuristics for physical risk

What's Working

Brazil's central bank (Banco Central do Brasil, BCB) has emerged as a model for emerging market climate stress testing. In 2024, BCB completed its second climate stress test covering 95% of banking system assets. Rather than directly importing NGFS scenarios, BCB developed Brazil-specific transition pathways incorporating deforestation-linked credit risk, soy and cattle commodity price shocks, and physical risk from Amazon basin drought patterns. The exercise identified R$120 billion (approximately $24 billion) in potentially impaired agricultural credit under a disorderly transition scenario, prompting 12 major banks to revise their agricultural lending criteria (BCB, 2025).

The Monetary Authority of Singapore (MAS) conducted its inaugural Industry-Wide Stress Test in 2024, covering 15 major banks and insurers. MAS distinguished itself by requiring institutions to model both physical and transition risks simultaneously rather than in isolation, reflecting the reality that in tropical emerging markets, these risks compound. The exercise revealed that combined physical-transition scenarios generated losses 40 to 60% higher than the sum of individual risk assessments, demonstrating the importance of integrated analysis (MAS, 2024).

South Africa's Prudential Authority completed pilot stress tests in 2023 with a focus on just transition risks, recognizing that the country's coal-dependent energy sector creates transition risks that interact with social stability and sovereign credit risk. The exercise incorporated employment data from coal-dependent regions, modeling second-round effects where mining job losses cascade into mortgage defaults and small business failures. This approach, unusual among global supervisors, connected climate transition risk to household-level financial stress in a way that resonated with bank risk managers (South African Reserve Bank, 2024).

Open-source scenario tools from the NGFS, particularly the REMIND-MAgPIE and GCAM integrated assessment models, have reduced the barrier to entry for emerging market regulators. The Philippines' Bangko Sentral ng Pilipinas used NGFS scenarios with locally calibrated typhoon damage functions to run its first climate stress test in 2025 at a fraction of the cost that would have been required to develop bespoke scenarios from scratch.

What's Not Working

Data quality remains the binding constraint. The ECB's 2022 climate stress test found that 60% of participating banks could not identify the geographic location of their corporate borrowers' physical assets, making physical risk assessment impossible for those exposures. In emerging markets the problem is more acute: a 2025 IFC survey found that only 22% of banks in Sub-Saharan Africa and 31% in Southeast Asia could map loan portfolios to emissions-intensive sectors at the NACE 4-digit level required for granular transition risk analysis (IFC, 2025).

Time horizon mismatches undermine relevance. Most bank risk frameworks operate on 1 to 3 year horizons aligned with credit cycles and regulatory capital requirements. Climate stress tests typically use 10 to 30 year horizons that exceed loan maturities, making it difficult to translate scenario outputs into actionable risk management decisions. The Bank of England acknowledged in its CBES review that participating banks struggled to connect 30-year scenario outputs to near-term capital planning (Bank of England, 2023).

Static balance sheet assumptions distort results. Most climate stress tests assume institutions hold their current portfolios unchanged over the scenario horizon, ignoring the reality that banks will actively manage exposures. This creates a systematic upward bias in loss estimates for transition risk (banks will reduce fossil fuel exposure as policy tightens) and potentially a downward bias for physical risk (banks may increase exposure to climate-vulnerable real estate if they fail to price the risk). The ECB has acknowledged this limitation but has not yet developed a dynamic balance sheet approach that regulators agree is consistent across institutions.

Greenwashing through compliance theater is emerging as a concern. Some institutions treat climate stress testing as a checkbox exercise, running scenarios with minimal internal engagement and producing reports that satisfy regulators without influencing actual risk management. A 2025 survey by the Institute of International Finance found that only 28% of banks in emerging markets reported that climate stress test results had influenced lending policies, pricing, or capital allocation decisions (IIF, 2025).

Key Players

Established Institutions

• European Central Bank: Conducted the largest supervisory climate stress test to date covering 104 significant institutions across the euro area
• Bank of England: Pioneered the Climate Biennial Exploratory Scenario approach and published detailed methodology for public use
• Network for Greening the Financial System (NGFS): Provides open-source climate scenarios used by 130+ central banks and supervisors globally
• Banco Central do Brasil: Developed the most advanced emerging market climate stress testing framework with locally calibrated scenarios
• Monetary Authority of Singapore: Introduced integrated physical-transition risk stress testing methodology for tropical economies

Startups and Technology Providers

• Planetrics (McKinsey): Offers scenario-based climate risk analytics platform used by banks and asset managers for stress testing
• Moody's Analytics: Provides climate-adjusted credit risk models integrating physical and transition risk scenarios
• Ortec Finance: Delivers climate scenario analysis tools used by pension funds and insurers for regulatory stress testing
• ClimateAlpha: Specializes in physical risk analytics using satellite data and machine learning for asset-level climate exposure mapping

Investors and Standard-Setters

• International Monetary Fund: Integrates climate stress testing into Financial Sector Assessment Programs for member countries
• World Bank Group/IFC: Funds climate stress testing capacity building in emerging market regulators and financial institutions
• Financial Stability Board: Oversees global coordination of climate-related financial risk assessment frameworks

Action Checklist

• Conduct a data readiness assessment: map portfolio exposures to geographic location, sector classification, and emissions intensity before beginning scenario analysis
• Select scenarios appropriate to local risk profile rather than defaulting to NGFS scenarios without calibration for regional climate hazards and economic structure
• Focus initial stress tests on 5 to 10 year horizons where scenario uncertainty is manageable and results can connect to existing risk management cycles
• Model physical and transition risks in combination rather than in isolation to capture compounding effects
• Establish internal governance processes that require stress test results to be reviewed by board risk committees and integrated into strategic planning
• Invest in staff training: climate risk analysis requires skills in climate science, scenario design, and financial modeling that most risk teams do not yet possess
• Engage with supervisors early and proactively to understand expectations and contribute to methodology development
• Document limitations transparently: stress test reports should clearly state data gaps, modeling assumptions, and the uncertainty range of results

FAQ

Q: How should emerging market banks prioritize between physical risk and transition risk in their first climate stress test? A: Start with the risk that is most material to the local economy and portfolio. For banks in tropical and coastal emerging markets (Philippines, Bangladesh, Caribbean nations), physical risk from typhoons, flooding, and sea-level rise will typically dominate. For banks in fossil fuel-exporting economies (Nigeria, Colombia, Indonesia), transition risk from commodity price declines and stranded assets is the priority. Brazil's experience shows that integrating both risks is ideal but that starting with the locally dominant risk produces more actionable first results than attempting comprehensive analysis with inadequate data.

Q: What is the minimum data infrastructure needed to run a credible climate stress test? A: At minimum, institutions need: sector classification of corporate exposures at the 2-digit level (4-digit preferred), geographic location of borrower operations or collateral (at least country and sub-national region), and basic financial projections for major borrowers under stress. For physical risk, asset-level geocoding (latitude/longitude) is ideal but postal code or district-level location can support initial analysis. The IFC estimates that a mid-sized emerging market bank can achieve minimum data readiness within 6 to 12 months with dedicated resources (IFC, 2025).

Q: Are climate stress test results comparable across institutions? A: Not meaningfully, at present. Differences in scenario calibration, modeling methodology, exposure classification, and assumption choices create variations that dwarf underlying risk differences. The ECB found that loss estimates for similar portfolios varied by a factor of 3 across participating banks, primarily due to methodological differences rather than genuine risk variation. Supervisors are aware of this limitation, which is why most climate stress tests to date have been described as "learning exercises" rather than pass/fail capital adequacy assessments.

Q: Will climate stress testing eventually affect bank capital requirements? A: Regulators are moving in this direction but cautiously. The ECB has stated that climate risk may be incorporated into Pillar 2 capital add-ons, and the Basel Committee on Banking Supervision published a 2024 consultation on integrating climate risk into the Basel framework. However, most supervisors acknowledge that current methodologies are not mature enough to support binding capital requirements. The most likely near-term impact is through supervisory expectations that influence risk management practices rather than formal capital charges.

Q: How do climate stress tests interact with TCFD and ISSB disclosure requirements? A: Climate stress testing and climate disclosure are complementary but distinct. TCFD and ISSB (IFRS S2) require companies to disclose climate-related risks and opportunities, including scenario analysis results. Climate stress tests conducted by regulators provide one input to these disclosures but typically use standardized scenarios and methodologies that differ from institution-specific TCFD/ISSB analysis. Institutions that align their internal scenario analysis with both regulatory stress test requirements and disclosure frameworks can reduce duplication, but should be aware that regulatory scenarios may not cover all risks material to their specific business.

Sources

• Network for Greening the Financial System. (2025). Climate Scenario Analysis: Progress Report and Updated Guidance for Central Banks and Supervisors. Paris: NGFS Secretariat.
• Bank for International Settlements. (2025). Climate-Related Financial Risk: Assessment Methodologies and Supervisory Practices. Basel: BIS.
• International Monetary Fund. (2025). Global Financial Stability Report: Climate Risk and Financial Sector Resilience in Emerging Markets. Washington, DC: IMF.
• Banco Central do Brasil. (2025). Financial Stability Report: Climate Stress Test Results and Methodology. Brasilia: BCB.
• Monetary Authority of Singapore. (2024). Industry-Wide Stress Test 2024: Climate Risk Scenarios and Results. Singapore: MAS.
• South African Reserve Bank. (2024). Climate Risk Stress Testing: Pilot Exercise Results and Lessons Learned. Pretoria: SARB Prudential Authority.
• Bank of England. (2023). Results of the 2021 Climate Biennial Exploratory Scenario: Follow-Up Review. London: Bank of England.
• International Finance Corporation. (2025). Climate Risk Data Readiness in Emerging Market Banking: Survey Results and Recommendations. Washington, DC: IFC.
• Institute of International Finance. (2025). Climate Stress Testing Implementation Survey: Global Results. Washington, DC: IIF.