After the hard work of producing climate-stressed economic pathways, financial institutions face the challenge of converting these into projected financial impacts on their own book.

For lenders, credit risk is generally perceived to be the most material transmission channel. To model this, the impact on borrowers, and hence of default, needs to be assessed. We now look at the two most material cases for retail banks.


Corporate Loans

Certain corporates are particularly exposed to transition risk, including firms in the energy, transport, extraction and manufacturing sectors. Corporate loan stress testing typically relies on projecting counterparty-specific financial statement data, and thence deriving stressed probabilities of default.

Key Challenge: Time Horizons

The time horizon can extend more than 30 years into the future, presenting challenges around modelling the evolution of company financials. Complex behaviour is seen in components of the P&L such as dividend issuance and the drawdown and repayment of debt. So careful calibration is required to model plausible firm behaviour under stress.

Key Challenge: Double Counting

Climate change scenario providers capture physical risk within their macro-economic scenarios. However, it is not always clear whether this physical risk assessment is limited to systemic impacts, or whether it also includes microeconomic channels, such as blended price impacts arising from physical hazards on individual properties. A detailed Integrated Assessment Model, such as that run by KPMG, gives full control and transparency around which risks are included or excluded.


Residential Mortgages

Residential mortgage portfolios are likely to be subject to significant physical and transition risks, impacting both the obligor and the value of the collateral.

The policies required to achieve a reduction in emissions will result in impacts on borrowers. For example, regulations around energy performance standards drive retrofitting costs. Broader macroeconomic effects through disorderly transition scenarios can cause general economic malaise, resulting in increased unemployment.

Pre-eminent physical threats vary by geography. Chronic risks such as subsidence and coastal inundation are liable to increase under ‘hot house world’ scenarios. And extreme weather events such as flooding, wildfires and storms are likely to increase in both frequency and severity.

Key Challenge: Forward-Looking Physical Risk

Physical hazard events can be modelled as shocks to borrowers’ income. However, as borrowers start to build a more forward-looking view of physical risks, these additional costs are likely to start being reflected in the value of a property. Contemporaneous consideration of physical risks may therefore be insufficient to capture impacts on collateral value. One solution is to incorporate these adaptive expectations into the projection of physical risk impacts on collateral value.

Key Challenge: Missing Data

Retrofitting costs are a key aspect of transition risk for mortgage holders. To model these effectively it is necessary to know current energy efficiency levels. In the EU and the UK, EPC (Energy Performance Certificate) ratings are used as a measure of energy efficiency, while in Switzerland the GEAK (Gebäude­energieausweis der Kantone) is the predominant standard. Completeness of data on the housing stock varies by country, and data imputation techniques are usually required to produce a useful dataset.

In summary, climate scenario analysis is still a young science, with industry consensus yet to be reached on several important topics. However, the pressure on banks to solve the problems discussed above, among others, will only increase as the climate emergency deepens.


Guest contribution by Owen Matthews, Director Financial Services, KPMG