Sustainability Atlas
Climate Finance & Markets·13 min read··...

Deep dive: Climate risk stress testing & scenario regulation — the fastest-moving subsegments to watch

What's working, what isn't, and what's next — with the trade-offs made explicit. Focus on data quality, standards alignment, and how to avoid measurement theater.

In May 2024, the Federal Reserve released findings from its pilot Climate Scenario Analysis exercise with six of the nation's largest banks, revealing that data gaps and modeling uncertainty remain the primary barriers to effective climate risk quantification. Meanwhile, the European Central Bank's 2025 EU-wide stress test found that climate transition risks could add 74 basis points to CET1 capital ratio depletion over 2025-2027 for exposed banks—a figure that crystallizes the material financial stakes of getting this right. As regulatory frameworks diverge across jurisdictions and methodological debates intensify around damage functions, understanding the fastest-moving subsegments in climate risk stress testing has become essential for financial institutions navigating this evolving landscape.

Why It Matters

Climate risk stress testing represents one of the most consequential developments in financial regulation since the 2008 crisis stress testing regime. The fundamental premise is straightforward: climate change creates both physical risks (extreme weather events, sea-level rise, chronic temperature shifts) and transition risks (policy changes, technology disruptions, market repricing) that can materially impair bank balance sheets. What makes this regulatory evolution distinct is its unprecedented time horizon—climate scenarios often extend 30 years or more—and its inherent uncertainty, given the non-linear dynamics of both climate systems and socioeconomic responses.

The Network for Greening the Financial System (NGFS), comprising over 130 central banks and supervisors, has established the global standard for climate scenarios. Their Phase V scenarios, released in November 2024, incorporate updated climate and economic data while introducing what has become a contentious damage function based on Kotz et al. (2024) research. This function, which attempts to translate temperature changes into GDP impacts, has faced significant academic critique, with Nature publishing a formal expression of concern in December 2024 regarding data reliability. This controversy illustrates a broader challenge: climate stress testing requires bridging climate science, economic modeling, and financial risk management—disciplines with fundamentally different epistemological frameworks and uncertainty tolerances.

The stakes extend beyond regulatory compliance. Banks with concentrated exposures to carbon-intensive sectors face potential stranded asset risks as transition policies accelerate. Insurers must price physical risks that may materially change within policy lifetimes. Asset managers confront fiduciary questions about climate-adjusted valuations. And all financial institutions face growing disclosure requirements under frameworks like the ISSB standards, making robust scenario analysis capabilities a competitive necessity.

Key Concepts

Understanding climate risk stress testing requires distinguishing several foundational concepts that are often conflated in practice.

Physical vs. Transition Risk: Physical risks encompass direct impacts from climate change—acute events like hurricanes and floods, and chronic shifts like rising temperatures and changing precipitation patterns. Transition risks arise from the economic adjustments required to decarbonize, including policy changes (carbon pricing, regulations), technological disruption (renewable energy cost declines), and market sentiment shifts (fossil fuel divestment). These risk categories interact in complex ways; aggressive transition policies reduce long-term physical risks but increase near-term transition risks.

Stress Testing vs. Scenario Analysis: While often used interchangeably, these techniques serve different purposes. Traditional stress tests assess capital adequacy under severe but plausible adverse conditions over short horizons (typically 1-3 years). Climate scenario analysis explores how different climate pathways—orderly transition, disorderly transition, hot house world—affect financial exposures over decades. The Basel Committee's April 2024 discussion paper emphasized this distinction, noting that climate analysis serves exploratory rather than capital-setting purposes in most jurisdictions.

Static vs. Dynamic Balance Sheet Assumptions: A critical methodological choice concerns whether to assume banks maintain their current exposures throughout the scenario horizon or adjust portfolios in response to changing conditions. Static assumptions simplify analysis but may overstate risks if banks would naturally de-risk. Dynamic assumptions are more realistic but introduce modeling complexity and management action assumptions that are difficult to validate.

Damage Functions: These mathematical relationships translate climate variables (temperature, precipitation, sea level) into economic impacts (GDP growth, asset values, default rates). The NGFS damage function controversy highlights their importance and limitations. Different damage functions can produce wildly different results, and none adequately capture potential tipping points or compound climate events.

KPI CategoryMetricTypical Range (2024-2025)Target Direction
Capital ImpactCET1 ratio depletion from climate scenarios50-150 basis pointsMinimize
Data CoveragePortfolio emissions coverage (Scope 1+2)40-80%>90%
Scenario BreadthNumber of scenarios analyzed3-6Context-dependent
GranularityCounterparty-level physical risk assessment20-50% of exposures>80%
Time HorizonMaximum scenario projection period10-30 yearsAligned with asset life

What's Working and What Isn't

What's Working

Standardized scenario frameworks are enabling comparability. The NGFS scenarios have achieved widespread adoption, providing a common reference point across jurisdictions. Whether analyzing a Net Zero 2050 pathway or a Delayed Transition scenario, regulators and institutions can now communicate using shared terminology and broadly aligned assumptions. This standardization has accelerated learning curves and enabled cross-institutional benchmarking.

Regulatory coordination is improving incrementally. The Basel Committee's 2024 discussion paper and the Financial Stability Board's ongoing roadmap work are building consensus on core principles, even as implementation varies. The June 2025 BCBS voluntary disclosure framework represents a pragmatic approach—establishing common expectations while respecting jurisdictional differences in regulatory mandates.

Physical risk analytics have matured significantly. Providers like Munich Re, Jupiter Intelligence, and Four Twenty Seven (now part of Moody's) offer increasingly granular hazard assessments. Location-based exposure analysis for flooding, wildfire, and extreme heat has progressed from experimental to operationally deployable for most asset classes with known geographic footprints.

Transition risk pathway analysis is becoming more sophisticated. Carbon price trajectories, technology adoption curves, and sector-specific transition pathways are increasingly well-defined within major scenarios. For high-emitting sectors like power generation, steel, and cement, the analytical frameworks for assessing transition exposure are reasonably mature.

What Isn't Working

Credit risk transmission mechanisms remain poorly modeled. While we can estimate a sector's carbon intensity or a facility's flood exposure, translating these into probability of default and loss given default estimates with any precision remains elusive. The Fed's pilot exercise explicitly noted that participating banks used widely varying approaches with limited ability to validate results. The fundamental challenge is that historical data provides limited guidance for unprecedented climate conditions.

Data gaps persist at the counterparty level. Most banks lack granular emissions data for the majority of their portfolios. Scope 3 emissions—often the largest component for financial institutions—depend on modeled estimates with significant uncertainty bands. Physical risk assessment requires precise asset locations, which many banks cannot reliably identify for commercial lending portfolios.

Time horizon mismatches create governance tensions. Climate risks materialize over decades, but bank strategic planning cycles operate in 3-5 year windows, executive tenures average 4-7 years, and equity analysts focus on quarterly results. This temporal mismatch undermines incentives for genuine integration of long-term climate risks into decision-making, encouraging "measurement theater" where compliance activities generate reports but don't inform strategy.

Jurisdictional divergence is widening rather than narrowing. The ECB is embedding climate into mandatory stress testing with potential capital implications. The Bank of England has pioneered exploratory scenarios but stopped short of capital requirements. And US federal banking regulators have recently exited the NGFS and withdrawn interagency guidance, signaling a pullback from climate risk supervision. This divergence creates operational complexity for global institutions and risks regulatory arbitrage.

Tipping points and compound risks are inadequately addressed. Current scenarios typically assume smooth, linear relationships between temperature and economic impacts. But climate science increasingly emphasizes threshold effects—ice sheet collapse, permafrost feedback loops, Amazon dieback—that could produce discontinuous and cascading impacts. No mainstream stress testing framework adequately captures these tail risks.

Key Players

Established Leaders

European Central Bank (ECB): The ECB has led global supervisory practice, conducting its 2022 climate stress test and integrating climate risks into the 2025 EU-wide stress test. Their Climate and Nature Plan 2024-2025, completed in January 2026, embedded climate considerations across supervision, monetary policy, and operations.

Bank of England: The PRA's Climate Biennial Exploratory Scenario (CBES) in 2021-2022 was the first comprehensive climate stress test for major UK banks and insurers. Their supervisory expectations continue to push institutions toward more sophisticated scenario capabilities.

Network for Greening the Financial System (NGFS): This coalition of central banks maintains the dominant scenario framework used globally. Their ongoing work on short-term scenarios (3-5 year horizons) and updated methodological guidance shapes practice worldwide.

Moody's Analytics: Through acquisitions including Four Twenty Seven and integration with credit models, Moody's offers end-to-end climate risk assessment capabilities spanning physical and transition risks with direct credit risk translation.

MSCI: Their climate risk tools, including Climate VaR and transition pathway assessment, have become standards for asset managers and institutional investors conducting portfolio-level climate analysis.

Emerging Startups

Cervest: Focuses on asset-level physical risk analytics using machine learning and satellite data, enabling granular assessment of climate hazards for real estate and infrastructure portfolios.

Watershed: Provides carbon accounting infrastructure that many financial institutions use as a foundation for transition risk assessment, with increasingly sophisticated scenario modeling capabilities.

Riskthinking.AI: Offers climate risk analytics specifically designed for financial applications, emphasizing integration with existing risk management frameworks and regulatory requirements.

Jupiter Intelligence: Specializes in high-resolution physical risk projections, providing asset-level hazard assessments that banks and insurers use for exposure analysis and pricing.

Risilience: Focuses on climate scenario analysis for corporate strategy, helping companies stress-test business models against transition pathways.

Key Investors & Funders

Breakthrough Energy Ventures: Bill Gates's climate fund has invested in several climate analytics startups, recognizing the infrastructure role of risk assessment tools.

Generation Investment Management: Al Gore's sustainability-focused fund has been an active investor in climate data and analytics companies.

European Investment Bank: Through various programs, the EIB has funded climate risk methodology development and capacity building for financial institutions.

UK Centre for Greening Finance and Investment (CGFI): Provides public research funding specifically for climate financial risk methodologies and tools.

Examples

1. ING Group's Terra Approach: The Dutch bank developed a proprietary methodology for assessing alignment of its lending portfolio with Paris Agreement temperature pathways. Terra analyzes nine high-emitting sectors representing 75% of global emissions, comparing client trajectories against sector decarbonization benchmarks. This approach enables ING to identify clients and sectors where transition risks are concentrated and to engage proactively on decarbonization strategies. The methodology has been open-sourced through the Partnership for Carbon Accounting Financials (PCAF), enabling broader adoption.

2. Swiss Re's CLIMADA Platform: The global reinsurer developed and open-sourced CLIMADA, a probabilistic natural catastrophe damage model that integrates climate scenarios with economic and insurance exposure data. This tool enables Swiss Re to stress-test its underwriting portfolio against various climate futures and has been adopted by research institutions and other financial firms. The platform explicitly handles uncertainty through Monte Carlo simulation, providing probability distributions rather than point estimates.

3. Bank of America's Internal Carbon Price: BofA implemented an internal carbon price mechanism that applies shadow carbon costs to lending and investment decisions. This transition risk tool forces explicit consideration of how different carbon price trajectories would affect client creditworthiness. While not a comprehensive stress test, it represents practical integration of climate scenarios into credit decisions, with the carbon price level updated periodically to reflect evolving regulatory expectations.

Action Checklist

  • Conduct a data inventory to identify gaps in counterparty emissions data, physical asset locations, and sector classifications essential for climate risk modeling
  • Evaluate current scenario analysis capabilities against NGFS Phase V scenarios and identify methodology upgrades needed for compliance with emerging regulatory expectations
  • Establish governance frameworks that link long-term climate scenario outputs to near-term strategic decisions, including sector exposure limits and client engagement priorities
  • Build internal expertise or vendor partnerships to translate climate hazard projections into financial risk metrics (PD/LGD impacts, asset value adjustments)
  • Develop a methodology for assessing compound and tail risks that current linear scenario frameworks inadequately capture
  • Create stakeholder communication materials that explain scenario analysis limitations and prevent over-interpretation of outputs as precise predictions

FAQ

Q: Will climate stress tests result in higher capital requirements for banks? A: Currently, most jurisdictions treat climate scenario analysis as exploratory rather than capital-setting. The ECB's 2025 stress test incorporates climate risks but within the existing capital framework. However, supervisory expectations are evolving. Banks with poor climate risk management practices may face Pillar 2 capital add-ons at individual supervisory discretion. The trajectory suggests gradual integration into capital requirements, but explicit climate capital buffers remain unlikely in the near term.

Q: How should organizations handle the significant uncertainty in climate scenarios? A: Rather than seeking false precision, best practice involves analyzing multiple scenarios representing genuinely different climate and policy pathways, focusing on strategic insights rather than specific numerical outputs, and communicating results with appropriate uncertainty bounds. The goal is decision-relevant insight—identifying exposures that are problematic across scenarios—rather than point estimates of capital impacts. Sensitivity analysis around key assumptions (damage functions, discount rates, transition timing) helps bound the range of plausible outcomes.

Q: What time horizon should climate stress tests cover? A: The answer depends on the purpose. For capital adequacy assessment, horizons aligned with traditional stress tests (3-5 years) may be appropriate, focusing on near-term transition risks from policy changes. For strategic planning, longer horizons (10-30 years) that span physical risk manifestation periods are necessary. The NGFS short-term scenario work, targeting 3-5 year horizons for systemic risk assessment, attempts to bridge this gap. Institutions should conduct analysis across multiple horizons rather than selecting a single "correct" timeframe.

Q: How do climate stress testing requirements differ across major jurisdictions? A: Significant divergence exists. The EU has the most prescriptive requirements, with the ECB integrating climate into supervisory stress tests and the EBA issuing guidelines on ESG scenario analysis. The UK emphasizes supervisory expectations without mandatory capital implications. Canada's OSFI is implementing phased standardized exercises. The US has pulled back from federal climate risk supervision, leaving state regulators and voluntary frameworks to fill gaps. Global institutions must navigate this patchwork, typically anchoring on EU requirements as the most stringent while adapting for local variations.

Q: What role should internal carbon prices play in stress testing? A: Internal carbon prices serve as a practical transition risk tool, forcing explicit consideration of future carbon costs in current decisions. They work best when calibrated to plausible regulatory carbon price trajectories (often derived from NGFS scenarios) and updated periodically. However, they complement rather than replace full scenario analysis—carbon pricing captures one transition risk mechanism but misses technology disruption, demand shifts, and physical risks entirely. Organizations should view internal carbon pricing as one component of a comprehensive climate risk framework.

Sources

Topics

riskstressIRRimpact measurementWACCadditionalityblended finance

Related Articles