May 2025
The 2025 stress test is well advanced, and at the end of May some banks will receive on–site visits from the ECB for the first time. CRR III is creating additional work, and the ECB has raised its expectations for BCBS 239 and conservative assumptions. Continued efforts is key to optimising banks’ SREP scores and minimising qualitative findings.
The EBA-ECB stress test of 2025 is well underway. The EBA is testing a sample of 64 banks across the EU and Norway, with the ECB extending the stress test to a total of 96 banks under its direct supervision.
As in prior years, the ECB’s assessment of the test results will influence banks’ SREP scores and supervisory capital requirements. Quantitative results (capital depletion) will determine Pillar 2 guidance (P2G), while qualitative findings, including on banks’ data quality and governance, are likely to affect Pillar 2 requirements (P2R).
As always, banks completing their stress test submissions face a range of obstacles around data collection and impact estimation. Based on our conversations with clients, we expect banks to find the following four features of the 2025 stress test cycle particularly challenging.
1. CRR 3
The integration of the EU’s third Capital Requirements Regulation (CRR III) into banks’ starting points and projections has been the biggest innovation of the 2025 stress test. Although CRR III took effect from January 2025, transitional arrangements mean that banks do not need to implement it fully until 2030. However, the stress test requires banks to restate their end-of-2024 data and also to apply CRR 3 on a fully phased-in basis.
This means that banks need two different sets of RWA calculations: One for the stress test and one for their quarterly regulatory reporting. Within the stress test, the introduction of the new output floor also requires institutions with approved internal models to complete an additional RWA calculation using standardised approaches. The latter is particularly challenging with regards to credit risk - in terms of the effort and the expected impact on the results.
Anticipating this additional workload, many banks began preparing early for the stress test, with larger and better-resourced teams (including consultants) than before. Our conversations with clients suggest that many institutions have found 2025’s CRR 3 elements to be stretching. Banks could also face additional questions from the ECB once their stress test submissions have been compared with June’s quarterly COREP.
2. On-site visits
At the end of May, the ECB will make on-site visits to a selection of banks within its sample. This will be the first time that the ECB has focused on the processes banks use to generate EBA-ECB stress test results. The chosen banks will receive just one week’s notice of a visit. The ECB will also evaluate banks’ compliance with BCBS 239 principles (see below) as part of the EBA-ECB stress test process.
This kind of inspection will be a novel experience for the affected banks, adding a ‘mini-audit’ to the stress test process. Banks judged to have weak processes can expect to see this impact their P2R, and may be presented with findings to remediate. That creates significant additional pressure, given the ad-hoc processes that many institutions rely on to complete the stress test.
3. BCBS 239
It has been more than a decade since the publication of the BCBS 239 principles for effective risk data aggregation and risk reporting. ECB-supervised banks are accustomed to ensuring that their quarterly risk reporting complies with BCBS 239, but in the past banks have not typically applied the same standards to their stress test data gathering activities.
According to ECB, this needs to be changed. The ECB’s guide on risk data aggregation and risk reporting recently stated that it expects BCBS 239 principles to be applied to all regulatory reporting, including stress tests. The ECB has also emphasised that the quality of banks’ data submissions will be a particular focus during the 2025 cycle.
This is likely to leave many banks in a race to strengthen their stress test processes, given that assembling the necessary estimates and data usually falls outside the capabilities of quarterly reporting frameworks.
4. Conservative expectations
As in prior years, the stress test methodology is based on a ‘constrained bottom-up approach’ with banks compiling their own impact assessments within a defined set of parameters. Although the test also contains several top-down elements, the bottom-up approach creates the possibility for banks to present their submissions in an excessively favourable way.
This year, the ECB has gone further than before to encourage banks to take a conservative approach to their submissions, stating that “…the ECB will therefore strengthen its review of insufficiently prudent submissions”[1]. The ECB has also been explicit that any banks displaying insufficient prudence will face additional scrutiny, and that some could receive on-site inspections after the conclusion of the stress test “to identify structural weaknesses”.
Illustrating this strict new approach, the EBA and ECB will conduct a top-down simulation of Net Interest Income based on historic data submitted by the banks In March, banks’ access to its model is temporarily suspended, potentially to prevent any ‘gaming’ of their submissions.
In conclusion, although most banks are already well advanced with their data collection submissions, the 2025 stress test is far from over. In particular, banks could still face follow-up questions or on-site visits. As always, planning and preparation are key. Banks should therefore ask themselves:
- Are we appropriately prepared for any ECB on-site visit or inspection?
- How well prepared are we for inspection?
- Do we have any outstanding CRR 3 or BCBS 239 issues to resolve?
- What additional steps can we take to strengthen our data aggregation and verification?
