Assessment of climate-related credit risk

This report explores the methodologies used by banks to evaluate climate risks, drawing on a global survey of 32 banks across five regions. The survey reveals that core credit books, including large corporates (90%), commercial real estate (87%), and residential real estate (73%), are the exposure classes most frequently assessed. Banks are increasingly transitioning from qualitative expert-based judgement towards quantitative data-driven approaches, with most institutions having commenced these assessments within the last one to two years.

Use of results from climate risk assessments

Climate risk assessment outcomes are primarily utilised for climate risk reporting (62%), key risk indicators, and credit decisioning. While 62% of banks use these results for regulatory compliance, internal business objectives also drive adoption. Approximately two-thirds of surveyed banks use the findings to inform client engagement, typically through moderate interventions such as adjusting client ratings, requesting time-bound action plans, or aligning loan pricing with climate risk exposure.

Integration of climate into credit risk modelling processes

The integration of climate factors into traditional credit risk models is still in the early stages. Currently, 61% of respondents incorporate climate risk into Probability of Default (PD) modelling, while 43% include it in Loss Given Default (LGD) measurements. Integration into more advanced frameworks remains limited, with only 18% of banks incorporating climate factors into internal ratings-based (IRB) modelling and rank-ordering models. Expert judgement continues to be the dominant method for overriding or adjusting existing model outcomes.

Materiality assessment of climate risks

Banks employ structured materiality assessment practices to prioritise climate risks, frequently using shorter time horizons of 0–3 and 3–10 years for both physical and transition risks. Common practices include exposure analysis and qualitative scenario analysis to identify transmission channels. Over 70% of respondents have taken follow-up actions based on these assessments, such as informing policy and strategy development or updating risk inventories.

Transition risk assessment methodologies

Transition risks are predominantly assessed through metrics such as emission reduction targets, financed emissions, and fossil fuel exposure. For Scope 3 data collection, 77% of banks rely on client-reported emissions, while 77% also use industry averages. The Partnership for Carbon Accounting Financials (PCAF) is the preferred methodology for 78% of respondents, although the quality of collected Scope 3 data remains a significant challenge.

Physical risk assessment methodologies

While all surveyed banks incorporate first-order impacts like direct property damage, only 30% model second-order impacts such as labour productivity declines or supply chain disruptions. Hazard inputs frequently include rising sea levels, extreme weather events, and heatwaves. No banks have fully integrated climate tipping points into their assessments, reflecting a lack of standardised methodologies and available climate scenarios.

Sector-specific approaches for assessing physical and transition risks

Specific sectors receive varying levels of scrutiny; real estate and construction (90%) is the most frequently assessed sector for physical risk, while oil and gas and energy supply are priorities for transition risk. Common underlying factors included in methodologies range from fuel mix and carbon pricing for transportation to water availability for the agriculture sector.

Climate-related collateral value adjustments

Adjusting collateral values for future climate risks remains uncommon. Only 12% of banks currently adjust valuations for physical risks, and just 4% do so for transition risks. The report suggests that banks should ensure independent market valuators embed these risks in appraisal reports as a primary point of integration.

Climate scenario analysis

Network for Greening the Financial System (NGFS) reference scenarios are the most widely adopted, used by 85% of banks for transition risks and 56% for physical risks. Most banks expand these scenarios using internal teams to add sectoral and geographical detail rather than relying on third parties. Backtesting remains limited, with only 4% of banks performing it regularly.

Approaches and integration of ESG scores for risk assessment

More than half of the surveyed banks have an internal ESG risk scoring methodology. While 23% have a fully developed internal framework, 36% of banks partially or fully integrate these scores into credit ratings. Banks have yet to converge on a preferred method, with some aggregating environmental, social, and governance factors into a single score and others keeping them separate.

Data collection and governance

Public data (82% for physical risk) and external data providers are the most common sources. Banks are adapting procedures by introducing new questionnaires and staff training to improve data collection from clients. However, data granularity remains a challenge, particularly for project finance and small and medium-sized enterprise (SME) exposures.

Quantitative impact on key metrics

Quantification of climate risk impacts on capital requirements is largely in its infancy. Most banks reported only small changes in expected credit loss (ECL), typically between 0–2.5%. For metrics such as risk-weighted assets (RWA) and economic capital (ECAP), the majority of respondents selected “not assessed,” highlighting significant gaps in current analytical capabilities.