Skip to main content

Climate Stress Tests: Onwards and Upwards

The last four years have seen central banks run a series of pilot climate scenario analyses or stress tests. This has been the case in many jurisdictions. Supervisory authorities clearly feel they can assess the climate readiness of banks through this lens. 2022 saw another such effort. The European Central Bank (ECB) completed a climate stress test (“ECB CST”) for 104 significant financial institutions. On the back of it, it published a report at the end of the year that highlighted what it felt were a range of good climate stress testing practices for banks to follow.

The ECB takes a very considered view in this report. It reflects on the challenges banks
continue to face as they undertake robust climate risk assessment – data being a key one. It also takes into account the varying approaches participating banks have adopted in the recent stress test and notes the progress made. But it then goes on to summarise what it considers to be good practices for each element in the anatomy of a climate stress test, from climate scenario design to emission and EPC data to Probability of Default (PD) and Loss Given Default (LGD) modelling. Coupled with ECB’s 2022 thematic review on climate risk, the report leaves little doubt that the ECB intends to raise the bar for future stress test submissions and supervisory reviews. While ECB-led, it is an expectation that could also be matched by supervisory authorities in other jurisdictions going forward.

This write-up builds on the findings of the ECB report by framing maturity models for different components of a climate stress test. It recognises that each bank is at a different stage of its climate risk management journey. Progress across banks will therefore vary for each element of a climate stress test -- be it scenario selection or generation, climate data enhancement or advancement in climate-related default risk modelling, impact assessment or embedding the results. One bank may be stronger in its scenario generation, another may be stronger in impact assessment. By providing a spectrum of practices, this write-up aims to help banks benchmark where they currently are on the journey and, against ECB-suggested best practice, pace their progress incrementally. To be clear, progress is expected on all fronts, but it is helpful to have a reference point for what basic looks like and what leading looks like for each element of a stress test.

This write-up is the first in a series. As there are at least six major elements to a climate stress test, to maintain brevity, this first write-up of this series focuses on two of the six elements: climate data and climate scenario generation. Maturity models are provided for these two elements. Future write-ups in the series will aim to cover other elements, such as assessment of business impact (modelling) and embedding stress tests in the business.

While the perspective shared here is primarily relevant to banks in-scope of ongoing ECB supervision, executives at other banks (and branches) operating in other regulatory jurisdictions would also benefit. Banks yet to fall in scope of regulatory climate stress tests, or those that have already been targeted for regulatory climate stress tests but are now looking to respond to stakeholder expectation to run further internal climate stress tests and, hence, moving beyond their first-generation efforts, should also find the maturity models here a helpful benchmark.