By Luis de Guindos and Frank Elderson
Stress tests are key to ensuring that banks can withstand strong macroeconomic headwinds. At the Europen Central Bank (ECB), we use annual stress tests to inform our decisions about how much capital banks need to have. Yet for stress tests to provide a meaningful picture of banks’ resilience, the banks’ own projections need to be prudent which was not always the case in previous stress test exercises.
This behaviour falls short of our supervisory expectations and requires supervisors to conduct additional quality assurance to ensure that stress test results are credible. Despite the ECB’s extensive quality assurance, using top-down stress test models, peer benchmarking and supervisory scrutiny, this behaviour also makes it more likely that risks are underestimated. In this blog post, we explain the steps we are taking to address this issue and ensure that stress test results remain credible.
The 2025 stress test and why it matters
Put simply, stress tests look at how banks’ balance sheets would fare if the economy took a turn for the worse. This means that supervisors examine how banks’ capital positions evolve under severe – but plausible – scenarios.
Currently, bank stress tests in the EU are not a pass or fail exercise. Instead, the results guide various aspects of day-to-day supervision and the assessment of financial stability. For instance, the quantitative impact on banks’ capital positions in the stress test scenarios is used as a starting point to set the Pillar 2 guidance and the leverage ratio Pillar 2 guidance. Qualitative findings also feed into the Supervisory Review and Evaluation Process (SREP) and the ECB’s process for identifying financial stability risks.
Moreover, stress tests have become a vital tool for macroprudential analysis. The results of these stress tests – performed using banks’ projections – provide a starting point for the ECB’s macroprudential stress test, which we carry out without the banks’ involvement. Stress tests also foster transparency and market discipline. They provide crucial insights into the resilience of the European banking system as a whole and, in turn, European financial stability.
Today, the European Banking Authority (EBA) is launching the 2025 EU-wide stress test. As part of this exercise, the ECB will stress test 51 of the euro area’s largest banks as well as a further 45 banks not included in the EBA sample. We will examine how each bank’s capital position would evolve over a three-year horizon up to the end of 2027, using two projection scenarios: a baseline and an adverse scenario. In the adverse scenario a worsening of geopolitical tensions leads to severe supply and demand shocks that result in the European economy shrinking by 6.3% between 2024 and 2027.
Rising uncertainty and higher trade barriers lead to a drastic reduction in global trade, with a detrimental impact on the European economy. While the drop in GDP in this year’s adverse scenario is slightly larger than in the 2023 EBA exercise, when assessed across a broader range of macroeconomic and financial indicators including unemployment, real estate prices, interest rates and risk premia, the overall severity of the scenario is broadly comparable to the 2023 EBA exercise and earlier ones. The baseline, meanwhile, is based on the December 2024 Eurosystem staff macroeconomic projections for the euro area. This means that it is a forecast of how the macro-financial landscape may evolve using our assessment of its most likely path under current conditions.
As in previous stress tests, banks will use their own models to generate projections of how their capital would be depleted in these scenarios. As always, we will carefully review their projections, following a thorough quality assurance process, to ensure that the results are credible. But why is this quality assurance process so important? And is it really necessary?
Concerns about banks’ projections
Prudent projections are the backbone of sound stress testing. If banks submit projections that are overly optimistic and these are not challenged during the quality assurance process, stress test results would overestimate banks’ resilience to adverse macroeconomic shocks. That’s why it is crucial for banks to provide sufficiently prudent projections during stress tests.
However, we have found that in previous stress test exercises some banks submitted projections that did not fully reflect the impact of the stress test scenario and methodology on their risk profile. Therefore, the projections were not sufficiently prudent. This behaviour diverges considerably from our supervisory expectations and entails significant additional supervisory quality assurance to ensure that the outcome of the stress test is prudent and reliable.
Banks have an inherent incentive to reduce their projected losses in the adverse scenario and, in turn, influence their capital buffers and – through the public disclosure of results – how they are perceived by market participants. This increases the risk that banks and supervisors fail to identify and address vulnerabilities before shocks causing potentially large losses materialise. Furthermore, to remedy this behaviour we need to carry out robust supervisory quality assurance to ensure that data are reliable and projections are sufficiently prudent. This increases costs for supervisors and banks alike.
Disincentivising unwanted behaviour
In line with the supervisory priorities for 2025-27, we have put measures in place to disincentivise unwanted behaviour and address issues related to low data quality.
First, in the collection of bank balance sheet data for 2024, banks must submit their credit risk information without having already received the ECB’s credit risk benchmarks. These benchmarks contain the ECB’s projections of credit risk parameters for the exercise and are used by ECB staff to challenge banks’ projections. A delayed distribution of these benchmarks avoids disclosing information about forthcoming challenges at a too early stage.
Second, banks that submit overly optimistic projections will face additional scrutiny during the quality assurance phase of the stress test. This may include on-site visits, during which supervisors will look at internal reporting and governance frameworks as well as identified quality assurance issues that are relevant to the stress test.
Third, based on the insights gathered during these visits and banks’ behaviour during our exercises, some banks may be subject to additional on-site inspections as a follow-up to the stress test. These will focus on structural weaknesses in their stress-testing capabilities.
If banks repeatedly fail to remedy shortcomings in their stress-testing frameworks, they could end up facing additional measures forming part of an escalation ladder of increasing severity which is intended to incentivise banks to take prompt and effective action.
Increased focus on the data quality of banks’ stress test submissions
For banks to manage their risks, they first need to know what they are. This means banks must be able to effectively aggregate risk-related data and information across the institution. To reflect the importance of risk data aggregation and reporting capabilities for banks’ resilience, this year’s stress test will place even greater focus on assessing the quality of the data banks provide. As a consequence, findings concerning banks’ stress-testing practices may also feed into banks’ SREP scores relating to risk data aggregation capabilities and could ultimately affect their Pillar 2 requirements.
Making our stress tests even more insightful
In addition, in parallel to this year’s stress test, the ECB will also conduct an exploratory scenario analysis on counterparty credit risk (CCR) for a sample of 15 banks, with a view to strengthening the supervisory assessment of banks’ ability to model CCR under a range of stress conditions. This is a follow-up to a recent targeted review which identified some material deficiencies in how banks manage CCR and it will help us better understand the vulnerabilities stemming from interlinkages with the non-bank financial intermediation sector. This sector has grown exponentially over the past few years but its opaque nature makes it hard to measure its inherent risks and interconnectedness with banks.
This is why non-bank financial intermediaries (NBFIs) are a source of increasing regulatory and supervisory concern globally. This is reflected in our supervisory priorities for 2024-26 and 2025-27 and consistent with concerns identified through our monitoring of financial stability. The analysis will improve our understanding of banks’ modelling capabilities, of how to tackle vulnerabilities within CCR portfolios and of risks stemming from interlinkages with NBFIs. The scenario analysis is distinct from the stress test and will not have a bearing on the Pillar 2 guidance.
Moving forward, it is essential that our stress tests remain rigorous and insightful, given their importance in safeguarding the resilience of European banks. Systemic risks like those posed by climate change and nature degradation are often not fully captured in traditional models, underscoring the need to continuously adapt our stress-testing frameworks. We will incorporate such elements in our future stress test exercises to ensure that they remain reliable and that banks are well-equipped to weather even the strongest of headwinds. This will help protect the stability of our financial system.
Credible projections mean better stress tests
Sound and credible stress tests are a vital part of the supervisory toolkit and help ensure that banks can withstand tough macroeconomic conditions. Banks’ input and stress test results are also a crucial component in system-wide resilience assessments and macroprudential analysis. In today’s uncertain world, marked by an increasingly volatile external risk landscape and geopolitical shocks, having an accurate picture of banks’ resilience is more important than ever. This is why the ECB is ensuring that banks have strong incentives to provide projections that are sufficiently prudent and credible.
The post Why prudent projections are the backbone of sound stress testing appeared first on Caribbean News Global.
