Climate risk and credit risk provisions

By Charlotte Boulogne, Nina Larkin, Antonie Jagga, Rishi Mulavineth, Nitthila Prathapar, Sam Bray

It is becoming increasingly important for banks to understand and assess the impact of climate risk on their lending book, including the impact on their credit losses and as a result the adequacy of their credit provisions. Given the number of areas to consider in the context of Expected Credit Losses (‘ECL’) (e.g. impact on credit ratings, macroeconomic scenarios, asset valuations, etc.), investing in this area now would be a ‘no-regrets’ exercise, providing value to broader critical processes of banks (e.g. operations including business continuity planning). In this article we highlight key considerations for banks to factor in when incorporating climate risk into their ECL processes, models and estimates.

Why is climate risk relevant to credit risk provisions?

With the growing investor and regulatory focus on the impacts of climate change on the measurement and reporting of climate risks, it is becoming increasingly important to consider climate risk in ECL modelling. In the context of credit risk, climate risk refers to the financial impacts arising from climate change, including physical and transition risks. Some considerations of physical and transition risks that may impact ECL are outlined below.

Risks arising from the physical effects of climate change and environmental degradation

Has potential impacts on all lending portfolios of a bank.

Deterioration in the value of collateral assets may increase Loss Given Default (‘LGD’), but current modelling may not capture the extent this is impacted by climate risk and may not be granular enough. For example, HPI forecasts may not necessarily incorporate the impact of updates to flood risk by local councils, irrespective of flood events being observed.
Climate events that cause a destruction of assets or interfere with business operations may lead to an upward shift in Probability of Default (‘PD’).

Risks arising from the transition to a low-carbon economy such as policy changes and technological innovation

Primarily affects corporate portfolios, particularly industries impacted by energy transition policies (e.g. mining, agriculture) and their employees (indirect impact on consumer lending).

In high-risk industries, certain assets such as manufacturing assets or reserves may become stranded, rendering them unusable. This can lead to a decrease in collateral value, resulting in an increase in LGD.
Industries impacted by transition risk may see decreasing profits over the medium to long term. Consequently, businesses that are unable to transition their operations will have a reduced ability to service their debt, increasing PD.

3 key questions for banks to consider:

What segments of their book will be most significantly impacted and require more accurate modelling?

Banks may choose to prioritise particular portfolios, sectors or asset classes that are most significantly impacted by climate risk when looking to incorporate climate risk considerations within credit provisioning. For example, if a bank has a substantial exposure to the mining industry, management may choose to quantify the impact of climate risk on the relevant exposures and take appropriate action accordingly, before incorporating climate risk modelling into other elements of the bank's portfolios.

When will the impact of climate risk be realised?

AASB 9 requires banks to consider the losses over the lifetime of a loan when calculating Stage 2 ECL. If climate risk is expected to materially impact a borrower's credit risk or asset value prior to a loan’s maturity, then banks need to consider the impact of climate risk within their credit provisioning. For example, climate-related government policies which affect specific industries (e.g. minimum energy efficiency requirements) and consequently impact their businesses operations, may result in changes to credit risk during the lifetime of the loans in these industries.

What is a cost effective solution to integrate climate risk into existing credit risk modelling?

While incorporating complex climate risk modelling into credit risk models may not be feasible in the short term, banks can consider prioritising the incorporation of factors related to climate change into their credit risk models, based on their impact on credit losses. Where synergies exist, banks could also leverage existing processes, people and infrastructure to achieve this. These strategies would allow banks to manage climate-related risks within their existing risk management frameworks, without incurring the additional costs and complexities of independent climate risk modelling.

How can banks incorporate climate risk considerations into their credit risk models?

Banks may consider various modelling approaches to capture the interactions between climate risk and credit provisioning:

Macroeconomic forecasts & scenarios (Leveraging Existing Model Structures)

Physical risk can impact house price values, however given potentially localised climate impacts, more localised factors may need to be considered.
Transition risk can impact the unemployment rate and gross value added for particular industries, such as mining and agriculture.
These macroeconomic factors can feed into forward-looking projection models or collateral values and thus impact PD, LGD and Staging assessments.

Credit risk ratings (Leveraging Existing Model Structures)

Credit risk ratings can be influenced by climate risk, particularly for corporate portfolios and to some extent all secured portfolios, where higher climate risk may result in a worse rating.
These ratings can be adjusted using a bottom-up (obligor-level) or a top-down (portfolio-level) approach. They can also be adjusted qualitatively, quantitatively or or through a hybrid approach.

Overlays and haircuts (Additional Analysis)

Overlays can be used as a short term solution to capture climate risk impacts on specific exposures. For example, the PD for Oil & Gas exposures may be stressed to capture the heightened transition risk.
Collateral haircuts can be applied to LGD estimates for specific properties based on localisation or asset type.
Any overlays or haircuts under consideration should be justified and based on data where possible.

Collaboration with third parties (Additional Analysis)

There may be opportunities to collaborate with insurance firms or other third parties, such as credit rating agencies.
For example, engaging with external service providers that specialise in climate risk analytics and data can help banks to identify and manage climate-related risks more effectively.

How do these considerations impact broader banks’ processes?

To enable a reliable integration of climate risk considerations into credit risk modelling (as per the examples above), additional internal and external data points will be required when building models. These new data points should be accurate and therefore collected through a controlled process. Given this, the identification and collection of climate-related data are essential and should be integrated into the broader data strategy (see our article here).

There is a trade-off between offering financing solutions to a large proportion of customers, regardless of how climate risk can impact their properties, and the risk of financing non-insurable properties. However, excluding customers based on the higher impact of climate risk on their credit risk may result in adverse selection and reputational risk.

The downside-scenario assumptions adjusted for climate risk in credit provisioning models should be consistent with other assumptions used in other climate risk scenario analysis (e.g. stress testing and/or climate vulnerability assessment).

Once the Australian Sustainability Reporting Standards (ASRS) are finalised, banks might be required to disclose material information about their climate-related risks and their risk management frameworks, in addition to quantifying the impact on their financial statements under different scenarios. Linking these scenario analyses to those used in credit risk modelling will ensure consistency across different use cases and avoid the need for multiple different models.

Providing board members with a deeper understanding of the implications of climate risk on credit risk will enhance strategic decision-making. This knowledge will empower leaders to set informed strategic objectives aligned with the evolving financial landscape shaped by climate-related challenges and opportunities.