2025 EU Stress Test Results

On 1 August 2025, the European Banking Authority published the results of the 2025 EU-wide stress test. The headline finding: European banks are resilient. Despite absorbing €547 billion in cumulative losses over a three-year adverse scenario driven by geopolitical escalation and trade fragmentation, the EU banking system maintained a CET1 capital ratio above 12% — comfortably above supervisory minimums.

But the aggregate numbers tell only part of the story. Underneath the system-wide resilience lies significant variation in how individual banks performed, which risk types dominated the losses, and what the results mean for capital planning and supervisory expectations going forward.

Headline Results: Stronger Than 2023 Despite Harder Stress

The 64 banks in the sample started the exercise with an aggregate CET1 capital ratio of 15.76% at end-2024 — the strongest starting position in the history of the exercise. Under the adverse scenario, this ratio fell to 12.06% by end-2027, representing a depletion of 370 basis points. Under the baseline scenario, the ratio actually rose to 16.9%, reflecting the strong profitability environment.

The 370 basis-point depletion marks a significant improvement over the 479 basis points recorded in the 2023 exercise — even though the 2025 adverse scenario was slightly more severe in GDP terms. The primary explanation is not that banks faced lighter losses; rather, they entered the exercise with substantially stronger income-generating capacity. Return on equity across the sample had reached its highest level in over a decade, providing a larger buffer to absorb scenario-driven losses.

Where the Losses Came From: The CET1 Waterfall

The stress test results can be decomposed into a “waterfall” showing how different factors contributed to the overall capital depletion. Net interest income provided the single largest positive contribution, more than offsetting the combined impact of all three risk types. Credit risk was by far the dominant source of losses.

The chart below shows the simplified capital walk from the 15.76% starting ratio to the 12.06% ending ratio. Net income from all sources (NII, fees, commissions, and trading — net of administrative expenses) contributed approximately +511 basis points. Of this, net interest income alone accounted for +1,047 basis points at the gross level, reflecting the higher interest rate environment. Against this income buffer, credit risk losses subtracted 437 basis points, market risk 108 basis points, and operational risk 61 basis points. Dividend distributions and other items (including taxes) accounted for a further 159 basis points of depletion, while the increase in risk exposure amounts reduced the ratio by an additional 116 basis points.

Credit Risk: The Dominant Loss Driver

Credit risk losses totalled €394 billion across the sample — up from €347 billion in 2023. These losses represent 1.9% of total exposures at end-2024 and account for more than two-thirds of total risk-type losses. The increased credit losses reflect the scenario’s severe GDP contraction and the corresponding rise in corporate and household defaults.

The dispersion across individual banks was substantial: credit risk depletion ranged from just 10 basis points to 828 basis points. Losses were concentrated in three portfolio segments: non-financial corporate (NFC) exposures accounted for approximately half of total credit losses, with particularly high loss rates in the SME, consumer lending, and commercial real estate segments.

Bank-Level Dispersion: A Wide Range of Outcomes

While the aggregate figures suggest comfortable resilience, the individual bank results tell a more nuanced story. The CET1 depletion under the adverse scenario ranged from +106 basis points (a bank that actually strengthened its capital position under stress) to −1,263 basis points at the most affected institution. The median CET1 ratio at the end of the adverse scenario was 13.0%, with an interquartile range of 11.0% to 15.5%.

For most banks (33 out of 64), maximum depletion occurred in the third year of the horizon. However, 25 banks experienced their worst point in 2025 — the first year — reflecting concentrated market risk losses from the initial asset price shock. Six banks hit their trough in 2026. This variation in timing depends on three factors: the severity of the market risk hit in year one, the speed at which income sources recover, and the trajectory of risk exposure amounts over the three-year horizon.

The results also revealed that geographically diversified banks tended to cluster closer to the median depletion, with less extreme outcomes. Banks with higher starting CET1 ratios generally showed larger depletions, as their additional initial capitalisation covered risks that materialised during the scenario. And banks with higher proportions of Stage 2 or Stage 3 loans at the starting point produced larger credit risk losses over the horizon — a logical but important confirmation of the link between pre-existing asset quality and stress test outcomes.

The NII Surprise: Income as the Primary Shock Absorber

The most striking feature of the 2025 results is the sheer scale of net interest income’s contribution. At +1,047 basis points, NII accounted for more positive CET1 impact than any single loss category contributed in negative terms. This represented a significant increase relative to the 2023 exercise and was the single most important factor explaining the lower overall depletion.

Two structural factors drive this result. First, banks entered the exercise with significantly higher net interest margins than in previous cycles, reflecting the post-2022 rate-hiking environment. Second, the adverse scenario’s interest rate path — which features modestly higher short-term rates before a gradual decline — creates a favourable repricing dynamic for banks with typical maturity structures, where assets tend to reprice more slowly than liabilities.

The centralisation of NII projections in 2025 is worth noting. For the first time, the EBA prescribed binding formulas for how asset and liability rates evolve under each scenario. While this improves cross-bank comparability, it also means that individual banks’ NII results reflect the EBA’s models rather than their own interest rate risk assessments — a change that may affect how the market interprets bank-level income projections.

Market Risk and Operational Risk

Market risk losses contributed 108 basis points to the capital depletion. Credit spreads were the main driver, accounting for approximately 45% of market risk losses — particularly for banks with the highest fair-value exposures. Losses from sovereign bond holdings in the market risk portfolio amounted to €55 billion, with 64% flowing through capital via the fair-value-through-other-comprehensive-income (FVOCI) channel. Interest rate, equity, and fund-related losses were more evenly distributed.

Operational risk losses totalled €54.8 billion (61 basis points), broadly in line with the 2023 exercise. A notable shift in composition was observed: banks projected declining losses from major conduct risk events but increasing losses from other operational risk categories, consistent with the evolving risk landscape around cyber threats and process failures.

The CRR3 Factor: First Stress Test Under the New Rules

The 2025 exercise was the first to incorporate the CRR3/CRD VI banking package. On a transitional basis, the impact was largely neutral — the introduction of new credit risk standardised approach rules broadly offset decreases in market and operational risk capital requirements.

This distinction between transitional and fully loaded results matters because CRR3’s transitional arrangements will be phased out gradually through 2033. Banks with large internal-model portfolios — particularly in real estate lending — will need to plan for steadily rising capital requirements as the output floor takes full effect.

What the Results Mean for Bank Capital Planning

The stress test results feed directly into supervisory capital decisions through two channels. First, each bank’s maximum CET1 depletion under the adverse scenario is the starting point for setting Pillar 2 Guidance (P2G). The ECB applies a two-step approach: banks are placed into a depletion “bucket” in step one, then supervisors adjust within the bucket range based on bank-specific factors in step two. Banks with steeper depletions can generally expect higher P2G.

Second, the qualitative findings — including data quality, modelling capability, and governance — feed into the broader SREP assessment and can influence Pillar 2 Requirements (P2R). The ECB’s introduction of on-site inspections during the 2025 exercise signals that process quality is now under closer scrutiny than ever before.

Despite the positive aggregate results, the EBA cautioned against complacency. Banks with concentrated exposures to vulnerable sectors or countries, weak sectoral modelling capability, or high CRR3 output-floor sensitivity face meaningful headwinds in the supervisory dialogue ahead.

For a detailed view of the macroeconomic scenarios that drove these results, use our EBA 2025 Stress Test Scenario Explorer. To understand which economic sectors were most affected, see our sectoral risk analysis of the 2025 stress test.