What Is Climate Value-at-Risk and How Is It Used?

Climate value-at-risk (Climate VaR) is a framework for expressing the potential financial impact of climate scenarios on asset values in monetary terms — translating scenario analysis from a qualitative exercise into a quantitative estimate of portfolio value change. Like conventional Value-at-Risk (VaR), which estimates potential losses over a given time period and confidence level, Climate VaR provides an estimate of how much a portfolio might lose under adverse climate scenarios. Unlike conventional VaR, which is based on historical return distributions, Climate VaR is forward-looking, scenario-based, and explicitly incorporates both the physical risks of climate change and the transition risks of decarbonization policy.

Quick definition: Climate Value-at-Risk is a forward-looking risk metric that estimates the potential impact on portfolio valuations from climate-related risks — physical damage from climate change and transition costs of decarbonization — expressed as a percentage or dollar amount of portfolio value under defined climate scenarios.

Key takeaways

• Climate VaR models two channels simultaneously: transition risk (the financial cost of decarbonization policies — carbon pricing, emission standards, technology disruption) and physical risk (the financial cost of climate change damages — extreme weather, sea level rise, productivity losses).
• The most widely used Climate VaR methodology is MSCI's, which estimates transition costs (additional capex and operating costs for decarbonization compliance) and physical costs (damage from climate hazards) as present values, then expresses their sum as a percentage of company enterprise value.
• Key inputs to Climate VaR models: climate scenarios (which determine how much physical risk vs. transition risk materializes); emission intensity of portfolio companies (high-emission companies face higher transition costs); asset location data (physical risk depends on where assets are located); and carbon price trajectory (the cost of decarbonization compliance depends on how fast carbon prices rise).
• NGFS (Network for Greening the Financial System) provides the most widely used scenario framework for financial sector climate stress testing, offering scenarios ranging from orderly net-zero transition (low physical risk, moderate transition risk) to high-damage physical scenarios (high physical risk, low transition risk) and "hot house world" (very high physical risk, minimal transition action).
• Significant limitations: Climate VaR models require subjective scenario choices, contain substantial model uncertainty (particularly for physical risk at the asset level), and may systematically underestimate tail risks from climate tipping points or policy surprises.

How Climate VaR Is Calculated

The Two-Channel Framework

Climate VaR aggregates two types of estimated financial impacts:

Transition cost channel: For each company in the portfolio, estimate the net present value of additional costs imposed by the transition to a low-carbon economy under a given scenario:

• Additional capital expenditure to decarbonize operations (fuel switching, efficiency improvement, carbon capture)
• Additional operating costs from carbon pricing (direct compliance costs for covered emissions)
• Reduced revenue from declining demand for carbon-intensive products (in scenarios where demand falls)
• Less: value of transition opportunities (companies providing clean technology solutions)

Physical damage channel: For each company (and for each asset within the company if location data is available), estimate the net present value of climate-related physical damage costs:

• Increased insurance costs and uninsured losses from more frequent and intense extreme weather
• Asset write-downs from sea level rise, flooding, or wildfire risk
• Revenue losses from heat stress on labor productivity, water stress on operations, and supply chain disruption
• Adaptation costs (building flood defenses, relocating facilities)

Aggregation: Each channel produces a distribution of outcomes, which are then combined into a total Climate VaR estimate expressed as a percentage reduction in portfolio value under the specified scenario.

The MSCI Climate VaR Methodology

MSCI's Climate VaR, one of the most widely adopted commercial methodologies, uses:

Transition VaR: Based on company GHG emissions and emission intensity, MSCI estimates required decarbonization costs under each scenario using sector-specific marginal abatement cost curves and carbon price trajectories. For a coal power company in an orderly net-zero scenario, this might represent a large percentage of enterprise value.

Physical VaR: Using climate hazard models and asset location data (where available), MSCI estimates the present value of physical damage costs. Physical VaR is expressed as a percentage of company assets at risk from specific hazards.

Combined Climate VaR: The sum of transition VaR and physical VaR, offset by any estimated transition opportunities, expressed as a percentage of portfolio value.

A portfolio Climate VaR of -10% would indicate that, under the specified climate scenario, the portfolio is expected to lose approximately 10% of its value relative to a no-climate-change baseline.

NGFS Scenarios for Financial Sector Stress Testing

The Network for Greening the Financial System, a group of over 130 central banks and financial supervisors, has developed a suite of climate scenarios specifically designed for financial sector risk assessment:

Orderly Scenarios (Low Physical, Variable Transition Risk)

Net Zero 2050 (Orderly): Strong policy action starts immediately and smoothly; carbon pricing rises gradually; technology deployment accelerates; global warming limited to 1.5°C. Relatively low physical risk; moderate transition risk for high-carbon sectors; transition opportunities for clean technology.

Below 2°C (Orderly): Less ambitious policy than Net Zero 2050; some overshoot; warming limited to below 2°C by 2100. Lower transition risk than NZ 2050; moderate physical risk.

Disorderly Scenarios (Moderate Physical, High Transition Risk)

Divergent Net Zero: Achieves 1.5°C but through higher-cost, disorderly transition — policy action is delayed then abrupt; significant economic disruption. Higher transition risk and economic disruption than orderly NZ 2050.

Delayed Transition: Policy action delayed until 2030, then implemented rapidly to achieve below 2°C. High transition risk (rapid change compresses into shorter timeframe); moderate physical risk.

High Physical Risk Scenarios (Hot House World)

Nationally Determined Contributions (NDC): Current policy commitments only; warming reaches approximately 2.5°C by 2100. Moderate physical risk beginning to materialize; lower transition risk (less policy action).

Current Policies: No new climate policy; warming reaches 3°C+ by 2100. High chronic physical risk; high acute extreme weather risk; low transition risk.

Hothouse World: Minimal climate action; warming of 4°C or more; severe and widespread physical risk. Catastrophic physical scenarios with profound economic disruption.

NGFS scenario risk matrix

Regulatory Applications: Climate Stress Testing

Central banks and financial supervisors are increasingly using Climate VaR-type methodologies for macroprudential risk assessment:

Bank of England Climate Biennial Exploratory Scenario (CBES, 2021-2022): The BoE required major UK banks and insurers to assess their balance sheet exposure to three climate scenarios (Early Action — orderly transition; Late Action — disorderly delayed; No Additional Action — hot house) across a 30-year horizon. The results, published in May 2022, found that the UK financial system faces meaningful losses under all three scenarios, with the late action (delayed, then rapid transition) scenario generating the highest financial system stress.

ECB Supervisory Climate Stress Test (2022): The European Central Bank conducted a climate stress test of 104 directly supervised banks, assessing climate risk exposures over three scenarios (orderly, disorderly, hot house) over short (3-year) and long (30-year) horizons. The ECB found that banks significantly underestimate climate risk in their internal models and have insufficient data to properly assess physical risk at the asset level.

Federal Reserve climate scenario analysis (2023-2025): The US Federal Reserve has engaged in climate scenario analysis pilots with major US banks, though the US regulatory climate stress testing framework is less advanced than the UK and EU equivalents.

Practical Investment Applications

Portfolio-Level Climate Stress Testing

Asset managers conducting Climate VaR analysis can:

1. Select scenarios (typically NGFS scenarios for methodological consistency with regulatory expectations)
2. Run portfolio climate risk through a Climate VaR model
3. Identify the top 10-20 portfolio holdings by contribution to climate risk
4. Assess whether high-risk holdings have adequate risk compensation (higher returns) or represent mispriced risks
5. Use results in TCFD scenario analysis reporting

Relative Comparison of Fund Climate Risk

Climate VaR enables comparison of climate risk across different fund strategies:

• An ESG low-carbon fund should show lower transition VaR than a conventional benchmark (due to lower exposure to high-carbon sectors)
• A renewable energy fund may show lower transition VaR but potentially higher physical VaR (infrastructure exposed to weather events)
• A broadly diversified fund shows climate VaR contributions from multiple sectors and hazard types

Integration with Financial Risk Management

Climate VaR can be incorporated into portfolio risk management frameworks alongside conventional VaR:

• Scenario analysis results inform position sizing and concentration limits
• Physical risk scores inform real estate and infrastructure exposure limits
• Carbon pricing sensitivity analysis informs exposure limits for high-carbon equity or credit holdings

Limitations of Climate VaR

Model uncertainty: Climate VaR models require assumptions about climate-economy coupling, carbon price trajectories, physical damage functions, and technology cost curves that are deeply uncertain. Small changes in assumptions can produce large differences in VaR estimates.

Tipping points and tail risk: Standard Climate VaR models typically use relatively smooth, continuous damage functions. Climate tipping points (Amazon disinclusion, permafrost carbon release, ice sheet collapse) could produce non-linear, much larger economic impacts that are not captured in conventional modeling approaches.

Systemic risk underestimation: Individual company or portfolio Climate VaR may underestimate systemic risk effects — where climate impacts affect multiple sectors simultaneously, causing correlated losses and second-order economic effects that are not captured in bottom-up company-by-company modeling.

Discount rate sensitivity: Physical climate damages that materialize in 30-50 years are heavily discounted in present-value calculations. Using higher discount rates substantially reduces Climate VaR for chronic physical risks, while using lower discount rates (arguably more appropriate for systemic long-run risks) substantially increases it.

Common mistakes

Treating Climate VaR output as precise: Climate VaR estimates have wide confidence intervals. A model estimate of -8% portfolio impact should be interpreted as "potentially in the range of -4% to -15% depending on scenario and model assumptions" — not as a precise projection. Point estimates give false precision.

Using Climate VaR only for transition risk: Early Climate VaR implementations focused primarily on transition risk (better data, more tractable models). Physical risk, while harder to model, may be more financially significant over longer horizons. Balanced analysis requires both channels.

FAQ

How different are Climate VaR estimates from different providers?

Significantly different. Studies comparing Climate VaR estimates from different providers for the same portfolio have found variation of 50-100% or more in estimated impacts for the same scenario. This reflects different model architectures, scenario parameterizations, and physical risk methodologies. Climate VaR should be used directionally and comparatively rather than as an absolute number.

Is Climate VaR used in investment mandates?

Increasingly, yes. Some institutional investors include Climate VaR limits in investment management agreements or use Climate VaR to set concentration limits on climate-risk exposures. TCFD guidance recommends quantitative scenario analysis, and Climate VaR is the primary commercial implementation of this recommendation. Regulatory requirements for climate stress testing in banks are creating pressure for equivalent capabilities in asset management.

Related concepts

• Physical Climate Risk
• Transition Risk
• Stranded Assets
• TCFD Framework
• Net-Zero Alignment
• ESG Glossary

Summary

Climate Value-at-Risk translates climate scenario analysis into portfolio impact estimates by quantifying both transition costs (decarbonization compliance) and physical damages (climate hazards) as present values under defined scenarios. NGFS scenarios provide a standard framework ranging from orderly transition (low physical, moderate transition risk) to hot house world (high physical, low transition risk). Central bank climate stress tests (BoE, ECB) are progressively requiring financial institutions to assess climate scenario impacts on balance sheets. Climate VaR is subject to substantial model uncertainty, tipping-point underestimation, and discount-rate sensitivity — estimates should be used directionally and with wide uncertainty bands rather than as precise projections. The metric is most valuable for relative comparison across portfolios, identification of concentrated climate risk positions, and TCFD scenario analysis reporting.

Next

→ Net-Zero Alignment