Introduction

• Banks face potential future losses from the economic and financial impacts of climate change. Effective risk management must identify climate risk drivers, measure exposures and concentrations, and quantify financial risk metrics.

Methodological Considerations

• Mapping and measuring risk exposures are crucial for effective risk governance (BCBS, 2015). Climate-related financial risks require special consideration due to their unique properties, which pose challenges to traditional risk measurement. Financial institutions and supervisors encounter methodological concepts when mapping and measuring exposure to climate-related financial risks. They can be structured as follows –

1. Conceptual considerations in measuring climate-related financial risks
2. Unique data requirements of mapping exposures and measuring climate-related financial risks
3. Role and characteristics of climate risk classifications as an initial input into mapping climate-related risk
4. exposures (Not a part of learning objectives)
5. Various approaches that estimate climate-related financial risks
6. Conceptual considerations for scenario analyses and stress testing (Not a part of learning objectives)

a) Conceptual Considerations – Risk managers are presented with new ideas when evaluating financial risks related to climate change. These ideas play a significant role in determining the extent of exposure to climate risk and measuring the financial risks associated with it.

a1) Physical and transition risk measurement considerations – To measure climate-related financial risks, there are two types of drivers – physical and transition risks. Physical risks are hazards that can be linked to financial exposure using damage functions, which measure the impact of specific hazards on real assets and financial flows. The challenge is finding the empirical functions for all sectors, exposures, and hazards. Transition risks, on the other hand, can be estimated by linking specific transition risk drivers to economic factors that generate financial flows. Due to the interdependent nature of these risks, they may need to be considered together, as actions taken to mitigate physical risks may also increase the probability of transition
risks.
a2) Exposure granularity – Banks and other financial institutions are exposed to climate-related financial risks through their dealings with clients and counterparties. To estimate the impact of climate risk drivers, banks and supervisors must decide on the level of exposure granularity required for their risk assessments. This decision is influenced by factors such as the specific risk drivers, the availability of data, the risk management decision, and computational complexity. The chosen methodology will affect the usefulness of model outputs for risk management, with higher granularity approaches being more useful for underwriting and valuation but suffering from high computational intensity and a lack of standardized data. Conversely, less complex methodologies that tolerate sparse data may be more useful for strategic planning or portfolio allocation. Banks and supervisors face a trade-off in these methodological decisions.
a3) Top-down and bottom-up approaches – Both exposure mapping and risk estimation involve selecting either top-down or bottom-up approaches. Top-down approaches begin with a general or aggregated measure of risk and attribute it to component parts, while bottom-up approaches assess risk at the component level and aggregate it up to provide a consolidated view of risk. Top-down approaches assume that sector or industry averages can be applied to individual exposures, while bottom-up approaches assess relative risk exposure at the asset or counterparty level. The choice of approach should consider potential correlations among risk exposures that could amplify or diversify risk within a portfolio or bank.
a4) Incorporating risk mitigation and risk reduction – Banks assessing climate-related financial risks should estimate the effect of potential risk mitigation and distinguish between net and gross exposures, where net exposure incorporates offsetting strategies and gross exposure does not. Counterparties can also take measures to offset climate-related risks through insurance coverage or other adaptation strategies. Calculating gross
exposure is important to understand the present magnitude of climate-related risks and to account for the potential failure or changes to mitigants. Banks and supervisors should consider measuring the effects of gross risk exposure and mitigants separately to fully understand the costs, benefits, and efficacy of different risk management strategies.
a5) Heterogeneities – Banks must consider their unique climate-related financial risks, which can vary based on their clients’ exposure to geographic regions, markets, sectors, political environments, and technology. This heterogeneity affects the selection of measurement approaches and model types. For example, climate hazards vary regionally, and sectoral classification can mask differences in risk exposure within a sector. National and subnational policy and regulatory regimes also influence risk vulnerability.
a6) Sources of uncertainty – Measuring climate-related financial risks involves uncertainty and assumptions about the interactions among climate, anthropogenic activity, and economic activity, which can lead to misestimation of risks. Estimates of climate sensitivity have generally trended higher, indicating that climate-related financial risks may be underestimated. Historical data is limited in risk estimation or model calibration, and assumptions about the future behaviour of economic actors, policymakers, and technological advancements introduce a high level of uncertainty. Model uncertainty arises from potential non-linearity in impacts, interconnections among natural systems, and spatial heterogeneity. The long time horizons over which these climate impacts are expected to manifest require assumptions about discounting and timing of occurrence. Forward-looking scenarios are often evaluated to take these different drivers of uncertainty into account. Advances in modelling, data, and technical alternatives could help reduce uncertainty, but residual uncertainty may remain even with such advances.

b) Data needs – Assessing climate-related financial risks requires new types of data that may not meet traditional quality standards. Historical data may not be representative of future climate-economy or climate- financial relationships. Three broad data categories needed for risk assessment are:

b1) data describing physical and transition risk drivers, needed to translate climate risk drivers into economic risk factors (i.e. climate-adjusted economic risk factors)
b2) data describing the vulnerability of exposures, linking climate-adjusted economic risk factors to exposures
b3) financial exposure data, needed to translate climate-adjusted economic risk into financial risk

b1) Data describing physical and transition risk drivers – Climate-related risk assessment for banking exposures begins with data on physical and transition risk drivers, including climate and hazard information. Geographical data, cost and performance data for energy substitutes, and other data can be used to estimate energy price relationships, identify at-risk locations, and influence economic outcomes. Many types of climate risk driver data are supplied by government agencies, academic organizations, and commercial third parties.
Off-the-shelf data sets are available to identify areas exposed to individual physical hazards, and several third-party data providers offer climate analytics that calculate aggregate physical risk indices. Climate data may feature in models that generate optimal carbon prices or projections of economic activity, but only some institutions have the expertise and resources to estimate these risk factors in-house, and others may choose to
source them from external third parties.
b2) Data describing the vulnerability of exposures – Banks and supervisors need exposure vulnerability data to assess the impact of climate-adjusted economic risk factors on bank exposures. These data vary depending on the climate risk driver being considered. Physical risks can affect the economy in various ways, and geospatial data are crucial to assess their impact on bank lending exposure. Granular details about a borrower’s value chain and interconnectedness among borrowers may also be necessary to estimate the
potential impairment. Transition risk assessments require data about a corporation’s carbon sensitivity, including GHG emissions, productive capacity, and innovation capabilities. The price elasticity of energy demand and supply constraints are also crucial in assessing the profitability of economic sectors.
b3) Data to translate climate-adjusted economic risk factors into financial risk – Banks and supervisors need additional data to estimate financial loss from exposure vulnerability. They can use financial variables commonly used in risk measurement, such as cash flow projections, valuations, or prices. Portfolio composition and counterparty information are also necessary to estimate potential impacts on exposures and adjust credit valuation and revenue forecasts. To model bank liquidity impacts from climate risk, data on funding provider behaviours (depositors, bondholders, wholesale funds) in response to climate risk drivers are required.
c) Candidate methodologies – The early stage of climate-related financial risk estimation lacks a consensus on preferred modelling approaches. There are various conceptual modelling and risk measurement approaches that can be used for estimation.
c1) Models for assessing economic impacts of climate change – To quantify climate-related financial risks, banks and supervisors must first specify the economic variables underlying asset performance. Various modelling approaches exist, each with strengths and weaknesses related to complexity, validity of assumptions, transparency of mechanisms, data requirements, and computational burden.

1. Integrated assessment models (IAMs) are commonly used to link transition risk drivers and GHG emissions to economic growth impacts. However, IAMs may not accurately project acute physical risks and adaptation possibilities.
2. Input-output models quantify economic linkages among sectors and geographic areas, while macroeconomic modelling attempts to capture general equilibrium effects. However, greater complexity can lead to greater opacity. Underestimating tail events can have significant effects on banks’ resiliency, and it is difficult to quantify unknown uncertainty.
3. Computable general equilibrium (CGE) models allow for complex policy experiments, but have a black box aspect due to the complexity of behavioural interactions.
4. Dynamic stochastic general equilibrium (DSGE) models introduce further complexity and uncertainty.
5. Overlapping generation (OLG) models focus on intergenerational distribution of consumption.
6. Agent-based models (ABMs) are advocated for their ability to better reflect uncertainty and complexity, but have drawbacks such as high computational and data demands.

Outputs of these models can be used as inputs into stress testing approaches that link future climate paths and economic damages to risk in bank portfolios. However, these models often lack spatial granularity needed to
assess specific risk exposures, and a narrower scenario narrative may be needed for more precise assessments.

c2) Broad risk measurement approaches – Measuring climate-related financial risk is not fundamentally different from standard scenario analyses or stress tests, but uncertainties in impact, time horizon inconsistencies, and limited historical data make it complex and less reliable. Conventional risk measurements can be adapted, but current practices include:

1. Climate risk scores or ratings – Climate risk scores rate the climate risk of assets, companies, portfolios, or countries. They use a classification scheme and grading criteria to assign a quality score. The grading criteria can be qualitative or quantitative. These scores help banks and supervisors assess climate exposure. However, ratings may be biased due to limited data on smaller counterparties and short data histories.
2. Scenario analysis – Climate scenario analysis involves four steps: identifying physical and transition risk scenarios, linking impacts to financial risks, assessing sensitivities, and extrapolating impacts to calculate exposure and potential losses. It can be performed at different levels of granularity and can help quantify tail risks and uncertainties. This analysis is used to evaluate the potential implications of climate risk drivers on financial exposures over the long term.
3. Stress testing – Stress testing is a type of scenario analysis that evaluates a financial institution’s short-term resiliency to economic shocks. It can be macroprudential or microprudential, and can now include climate-related risks and scenarios. However, due to the uncertainty inherent in longer-dated assessments and limited predictive power of historical observations, estimates of capital shortfall may be less reliable compared to conventional stress tests.
4. Sensitivity analysis – Sensitivity analysis is used to evaluate the effect of a specific variable on economic outcomes. One parameter is altered across multiple scenario-runs to observe the range of outputs. It has been used in transition risk evaluation to assess potential effects of a specific climate-related policy on economic outcomes, particularly in research settings to evaluate the range of economic impacts from the implementation of a carbon tax. It can be a useful tool for risk decision-makers to understand the range of potential climate impacts given the uncertainties in scenario analyses.
5. Natural capital analysis – Natural capital analysis identifies how natural degradation affects a financial institution by assessing its dependencies on natural capital. It is conducted at a portfolio level and identifies relevant geographies, sectors, and assets, natural assets, potential disruptions, and entities most at risk. Natural capital analysis can also be adapted for transaction-level analysis. This analysis recognizes natural resources as finite and emphasizes their increasing cost as they become scarcer due to climate change.
6. Climate VaR – Climate VaR assessments apply traditional VaR to estimate how climate change will affect a financial institution’s balance sheets. These metrics use forward-looking, portfolio-level analysis to quantify the impact of climate change on the value of financial assets within a given time frame and probability, under particular climate scenarios

Measurement Methodologies

• Banks are developing methodologies to map and measure exposure to climate-related financial risks, but their quantification is in early stages. Supervisors are also evolving their methodologies, including adapting prudential data and developing scenario analysis and stress testing. Third parties have also developed approaches for assessing climate-related financial risks. Financial institutions and supervisors often rely on third-party methodology providers for individual metrics or scenarios

a) Exposure mapping and measurement –

a1) Bank-level methodologies – These include the following:

1. Portfolio and sectoral exposures – To assess risks, banks first identify the risk transmission channels based on their exposure profile, and develop indicators such as carbon intensity or physical risk vulnerability. In the absence of quantitative data, banks evaluate climate-related risks qualitatively and monitor concentration risks using heatmaps. Transition and physical risks are distinguished using indicators and metrics.
2. Transition risks – Banks analyze sectors that could be affected by a transition to a low-carbon economy, such as oil and gas, utilities, transportation, car manufacturing, metals and mining, and construction. They use various methods to measure transition risk, such as analyzing carbon-related assets or calculating the carbon footprint of assets to identify “pockets of risk”. Banks also use indicators related to the “greenness” of assets and real estate exposures as proxies for transition risk and to measure alignment with climate targets. Some banks assess the potential risk differential between “green” and “brown” activities, but strong conclusions on a risk differential have not been established yet. Availability of a clear risk classification system could enable more systematic tracking of risk profiles.
3. Physical risks – Banks use indicators and metrics to measure physical risk at the portfolio level. These aimto identify geographic risk concentrations, hazard type, probability, and potential severity. Location-basedphysical risk scores are developed to estimate sensitivity to various acute or chronic physical risks.Geospatial mapping is used to assess and monitor the extent to which exposures may be affected byphysical risks. For example, banks perform geospatial mapping of their portfolios to identify exposure toflood risk and assess their corporate credit exposure at risk of water stress.
4. Client or project ratings and scores – Banks are increasingly assessing climate-related financial risks toindividual counterparties, using qualitative and quantitative analyses to inform credit decisions. Theseassessments often include an analysis of climate-related opportunities and risks for companies that thebank finances, such as their carbon footprint and climate change adaptation solutions. Banks use rating orscoring approaches, which can be based on industry benchmarks and adjusted to account for company-specific aspects, to assign a climate risk rating for each client. Although climate risk ratings or scores aregenerally used to reflect climate-related factors when granting credit, banks are working towards integratinga climate-related score or rating into their counterparty credit assessment processes and risk managementframeworks. This might involve using a traffic light classification to differentiate clients according to theirrelative exposure to climate-related risks or adjusting the credit score based on the climate risk assessment.
   

a2) Supervisor-level methodologies – Supervisors use similar indicators and metrics as banks to assess exposure to climate-related financial risks, based on portfolio, sectoral, or geographical characteristics. They also consider
different sets of indicators for transition and physical risks.

1. Transition risks – Supervisors use similar indicators and metrics as banks to assess exposure to climate-related financial risks. They use either regulatory information or surveys to match indicators associated with transition risk to banks’ exposures. For corporate portfolios, supervisors often use indicators describing the emission intensity or sensitivity to climate policies of banks’ counterparties at the entity or sectoral level. They also use supervisory counterparty-level data combined with data on counterparties’ emissions or emission intensity to capture vulnerability within sectors. This data can be used for sensitivity analyses and scenario analysis to assess exposure to climate-related financial risks.
2. Physical risks – Supervisors use various methods to assess the impact of physical risks on banks’ exposures. Information from government agencies or commercial vendors is used to identify hazards and vulnerable regions. The assessment of exposures can be conducted at the counterparty, activity, or sectoral level. The output of such assessments provides the proportion of a portfolio located in an area with a higher level of physical risk or exposed to a particular climate event. The impact of these hazards on the value of the collateral held by banks can also be considered. While progress in analyzing exposures to physical risks varies, some supervisors have made progress in identifying hazards and mapping and measuring exposures.

b) Risk quantification: scenario analysis, stress testing and sensitivity analysis –

b1) Bank-level methodologies – Bank-level scenario analysis and stress testing methods may be used to quantify climate-related financial risks or to inform strategic planning. Currently, these tools are mostly focused on credit
risk or market risk analysis. Climate-related financial risk scenario analysis is applied with the goal of understanding the potential impact on selected portfolios, to refine methodologies and assess limitations and benefits. These
exercises are also used to identify counterparties which need to be engaged to support their transition. A crucial objective for banks is the translation of transition and physical risk drivers into financial risks, and in particular
their incorporation into internal models, which still seems challenging.

1. Banks analyze transition risk scenarios to evaluate impacts on credit parameters for counterparties in specific sectors, such as corporate exposures relevant to climate policies. For instance, a shadow price is used to measure the effect of adjustments to basic prices on counterparties. The objective is to anticipate potential market or policy changes and estimate the impact on the financial profile of a counterparty to assess adjusted credit risk. Although mostly used by institutions with more advanced climate risk analysis, other banks also signal their intent to use such methods in their planning or decision-making.
2. Banks mainly analyze physical risk related to corporate and household exposures, including mortgages. They assess the potential impact on counterparties’ credit quality, especially in sectors sensitive to weather
   patterns. Chronic risks are evaluated by estimating productivity changes, which translate into revenue changes for firms. Acute risks, on the other hand, are analyzed by evaluating potential impacts on the value of the bank’s real estate collateral and real estate exposures, including housing prices.

b2) Supervisor-level methodologies – Supervisors use scenario analysis and stress testing for microprudential supervision and macroprudential policies. For microprudential supervision, they quantify banks’ financial exposures to specific climate risk drivers and assess the vulnerability of banks’ business models. For macroprudential policies, they assess the size and distribution of climate risks in the financial system and whether these risks may be systemic. These analyses primarily focus on credit and market risks from financial institutions’ loan and investment portfolios. However, future applications may extend beyond these areas. Climate scenario analysis and stress testing are exploratory tools used by supervisors to better understand the impact of climate change on banks’ risk management and business strategy, and to raise awareness in the industry about these risks, unlike traditional supervisory stress testing, which aims to determine a bank’s resilience to
financial losses and may inform capital requirements. Different approaches are used to model the impact of climate-related risks at the macro, sector, and firm levels.

1. At the macro level, climate scenarios are translated into macroeconomic and financial market variables using multi-country macroeconometric models. The impact of these variables on credit risk parameters is then estimated.
2. Sector-level calibration may be introduced to differentiate the risk profile across sectors and enhance the granularity of the analysis.
3. Borrower-level analysis requires more granular data, including asset-level data such as location, emission profile, or physical hazard. Micro models can be used to assess the impact on a borrower’s creditworthiness. Examples include financial analysis of individual companies, modelling their cash flows and collateral values, and assessing their current mitigation and adaptation plans. Rating models can also be used to assess financial outcomes and identify the set of firms that exhibit the biggest increase or
   decrease in credit risk.

b3) Third-party approaches – Supervisors and banks sometimes use third-party methodologies or tools for climate scenario analysis and stress testing. These methodologies typically include exposure mapping, scenario selection, and the assessment of impacts on financial performance. They may provide various risk metrics such as climate value-at-risk, PD of firm, expected shortfall, and expected losses. Some methodologies also consider factors such as vulnerability to climate events, adaptive capacities, and alternative sources of production. Third-party methodologies may use climate risk scores or damage functions to assess physical risks, and can segment or aggregate exposures by type of loan, sector, or hazard.

Gaps and Challenges in Model Framework Design

a) Limitations of observed risk classification approaches –

The goal of risk classification is to group exposures based on criteria that reflect their risks. Publicly available information, such as air emissions accounts and vulnerability indices, can be used to identify sectors and countries sensitive to climate change. This can help banks monitor concentration risk and inform strategic planning and portfolio allocation. However, aggregate risk classification approaches have limitations, as identification criteria may not be granular enough to differentiate among counterparties. For example, some approaches assume that counterparties from the same geographical location share similar risk characteristics. However, the reality is that transition and adaptation capabilities of counterparties may vary depending on the specific physical or transition
risk driver under scrutiny. Moreover, an individual counterparty’s sensitivity to climate-related risks may not necessarily align with a transition-sensitive sector or geographical area. Additionally, without the ability to clearly identify and measure the hedging strategies employed by counterparties, it could be challenging to differentiate between gross and net exposure. Banks can complement these approaches by assigning a rating or score, but this is challenging due to issues with data availability and quality and the lack of an established link between climate-related information and financial risk parameters. Therefore, ratings are often considered separately from banks’
financial risk scoring.

b) Risk differentiation and comparability across banks and jurisdictions –

Risk classification systems aim to achieve comparability, but this objective has its drawbacks. Standardization and simplification elements are introduced to enable comparisons between banks’ exposures within or across jurisdictions, which can lead to a reduction in risk differentiation. A balance is needed between granularity/complexity and comparability. Obstacles to developing common standards include differences in risk profiles across jurisdictions and aggregation issues for large international banks. Local conditions may conflict with group-wide policies for climate risk assessments, and IT requirements for large climate databases may exacerbate existing deficiencies in risk data aggregation and reporting noted by supervisors.

c) Challenges in the availability of appropriate data –

Limited data and information availability hinder the development of climate risk measurement processes. External data providers may alleviate some of these limitations, but concerted efforts by banks and supervisors are likely necessary to overcome these challenges. This obstacle has been identified as a major impediment in recent years, and this observation has been confirmed by industry outreach seminars and a supervisory survey.

d) Data describing physical and transition risk drivers –

Climate data may be insufficient or not updated consistently with standard financial risk measurement frequencies, presenting methodological challenges. Climate model information may be presented in aggregated form, which may not accurately reflect the economic and financial impact analysis required for higher granularity. Moreover, the quality and availability of climate-related data may vary among jurisdictions, which may impede the implementation of comprehensive analyses.

e) Data describing the vulnerability of exposures –

1. Third-party rating information – External data providers can help banks by providing information to fill gaps in counterparty and exposure data related to climate risks. However, there are challenges to using this data, including the limited transparency and standardization of the underlying information. Climate-related ratings may be especially difficult to assess because of the lack of established standards and the proprietary nature of the models used to generate them. Comparability of ratings across different providers is also limited, and the data may not be representative of smaller companies. To overcome these challenges, banks and supervisors will need to conduct regular due diligence and work together to develop more reliable and standardized approaches to climate risk measurement.
2. Counterparty-level information – Banks rely on proprietary non-public information, statutory disclosures, and other data to evaluate the creditworthiness of counterparties. Obtaining client data can address some data gaps, but it may be incomplete and imprecise, particularly for small clients. Banks may also face limitations in updating data for existing exposures, which may create gaps in climate reporting. When proprietary data is unavailable, banks may rely on public information disclosed by borrowers, which is often limited in quantity and quality and may not be comparable across jurisdictions or firm sizes. The lack of generally accepted standards for climate-related disclosures further limits data comparability.
3. Supervisory reporting data – Supervisors can use their supervisory reports to obtain data needed for climate-related analyses on the banking system. These reports can provide information on banks’ asset portfolios, both from a macro and micro perspective. However, existing supervisory reporting may lack sufficient granularity to assess transition and physical risks. The geographic and sectoral breakdown of corporate exposures may not allow for sufficient risk differentiation, so supervisors may need to develop additional tools or amend current regulatory reports as appropriate.

e) Challenges in designing a modelling framework to capture climate-related financial risks –

1. Scenario design and complexity of climate-related financial risks – Climate-related financial risks are complex and uncertain, making it challenging to develop comprehensive scenarios that can be integrated into existing risk assessment processes. These risks vary across regions and sectors, and there are data gaps and model-based uncertainties that make it difficult to compute financial losses. Banks and supervisors report difficulties in modelling such scenarios, including uncertainty in projections of physical and transition risk drivers and non-linearities.

a. Uncertainty around climate risk drivers – Multiple uncertainties surround both transition and physical risk drivers, especially related to non-linear aspects of the climate system. There could be many combinations of transition drivers leading to the same CO2 concentration outcome, resulting in significant variance in economic and financial impact across and within sectors and geographies. Given this, accounting for uncertainty through alternative scenarios and different modelling approaches is necessary to make the analysis more robust. However, exploring multiple scenarios may be resource-intensive.
b. Capturing the specific impacts of climate scenarios – There is a need to develop modeling frameworks that can capture the impacts of climate scenarios, including stressed variants. However, currently available methodologies are not fully integrated or tractable, and banks and supervisors are using existing tools based on their relative strengths. Macroeconometric models are used to adjust for the macroeconomic impact of carbon price shocks when assessing transition risk, but they may not adequately capture climate scenario dynamics. Existing models that estimate
economic impacts from climate change usually do not capture the full range of potential climate change impacts, such as extreme weather events or potential future disruptive changes in climate.
c. Comprehensiveness of modelled impacts – Modeling challenges exist in capturing the impact of climate scenarios with a chosen risk classification and granularity. For physical risk, the lack of damage functions at the granular geographic level makes translating physical risk drivers into expected losses difficult. The assessment should also model second-round effects like the
propagation of policy or physical risk shocks and the adaptive and mitigation abilities of economic agents. Additionally, assessing the market power of individual firms to pass on carbon price increases to consumers and evolving insurance coverage or governments’ natural catastrophe schemes are also critical information to consider.

Supervisors may develop multiple models to address the complexities of climate-related risks. Some have added production network modules to macroeconometric models to capture sector-level transition impacts of carbon price shocks. However, connecting these disparate models with different theoretical foundations would require additional research.

2. Translating scenario outputs to financial risks – Financial models used to assess credit risk rely on historical statistical relationships between risk drivers and parameters such as PDs or LGDs. However, historical observations of climate-related risks are not good predictors for future patterns, making it challenging for financial models to derive empirical risk parameters. To address this, some banks and supervisors attempt to attribute the impact of stressed scenario variables on counterparties based on characteristics such as CO2 emissions, then infer the impact on risk parameters. However, this approach has limitations, including data availability challenges and simplified assumptions about the relationship between variables and financial impact.

3. Time horizon related challenges – Banks and supervisors face challenges in considering longer time horizons when assessing climate-related risks. These horizons can extend up to 2050, leading to increased uncertainty in economic and financial projections, as well as limitations in modelling. Banks may struggle to internalize negative feedback loops from their lending decisions due to risk frameworks geared towards short-term exposures. Reflecting on existing measurement approaches and potentially leveraging best practices in modelling approaches may help address these challenges. However, there is a need to evaluate the appropriateness of individual supervisory tools for various time horizons. Results of climate-related scenario exercises may challenge the traditional use of scenario analysis in stress testing, instead informing the strategic and business model resilience of institutions.
4. Operational complexity in the measurement of risk – The TFCR industry outreach and supervisory stocktake suggest that banks and supervisors need to continue building operational capabilities to assess climate-related risks. This is a difficult task that requires significant resources, including adequate systems infrastructure, relevant human resources, and a sophisticated organization. A bank’s ability to assess its overall exposure to climate risks will heavily rely on the quality of its IT systems and its ability to manage large amounts of data. Climate risk measurement also requires interdisciplinary expertise and may require more sophisticated modelling techniques for complex banking groups.