The frame
The European Central Bank’s 2026 thematic stress test is not a forward exercise. On 12 December 2025 the ECB told the 110 banks it directly supervises that, this year, supervisors will prescribe the outcome and the banks will have to describe the scenario that gets there. Each institution must define a geopolitical scenario that would deplete its Common Equity Tier 1 (CET1) ratio by at least 300 basis points, and explain — in operational, balance-sheet terms — how that scenario plays out across its book.
For a German reader the headline supervisor matters as much as the methodology. The 2026 reverse test sits on the ECB side of the German bank supervisory architecture, not the BaFin side. Last week’s German story was BaFin and the Bundesbank tightening control-function supervision under the national programme. This week’s story is the ECB driving the methodology on a different risk class, with Bundesbank acting in its standing role as the German national competent authority inside the Single Supervisory Mechanism (SSM) joint supervisory teams. The two stories share the same German banks. They do not share the same supervisor in the lead seat.
What a reverse stress test actually is
Conventional stress tests prescribe a scenario — a macro path, a market shock, a liquidity squeeze — and ask each bank to compute capital, liquidity and profit impacts. The 2026 thematic test inverts that. The ECB’s wording is direct: “In a reverse stress test, a pre-determined outcome is prescribed and each bank defines the scenario in which that outcome would materialise.” The prescribed outcome here is a depletion of at least 300 basis points of CET1.
That switch is not cosmetic. Forward scenarios stop where the supervisor’s imagination stops, and they tend to anchor on shocks that resemble past crises. A reverse test forces a bank to start at the painful end and reason backwards into the channels that get it there — which counterparties, which exposures, which operational dependencies, which jurisdictions. The exercise is designed to surface the tail risks that forward tests rarely reach, and to test whether banks can articulate them at all.
For supervisors, the reverse design also pulls a different kind of evidence out of banks. A forward test produces a number; a reverse test produces a narrative the bank has to defend. That narrative reveals how good the bank’s internal stress-testing capability is, how it aggregates risk data across silos, and whether geopolitical exposure shows up in the same risk taxonomy as credit or market risk.
Geopolitical risk as its own category
The ECB is treating geopolitical risk as a cross-cutting risk driver, not as a sub-flavour of market or credit risk. The 12 December release puts it precisely: “Geopolitical risk is a cross-cutting risk driver that can have an impact on banks’ traditional risk categories, as it cuts across credit, market, liquidity, business model, governance and operational risks.”
That recategorisation matters because it tells banks where the supervisory expectation lands. Banks cannot satisfy the test by ticking the geopolitical box inside an existing market-risk scenario set. They have to show that their risk materiality framework can carry geopolitical shocks as their own object, and that their data aggregation across credit, market, liquidity and operational books can be re-cut along a geopolitical line.
The analytical backbone for that recategorisation is the ECB-ESRB joint report on financial stability risks from geoeconomic fragmentation, published on 22 January 2026. The report sets out a monitoring framework that integrates geopolitical indicators into financial stability analysis and names geoeconomic fragmentation and geopolitical risk as key sources of financial stability stress. The reverse stress test is not a stand-alone supervisory whim; it is the bank-supervision instrument that the financial stability arm has been building the case for.
ICAAP integration, SREP integration, no P2G impact this cycle
The ECB has folded the reverse geopolitical risk test into the 2026 internal capital adequacy assessment process (ICAAP). That keeps the data-collection cost manageable — banks reuse existing supervisory templates rather than running a parallel exercise — and it embeds the geopolitical scenario inside the document each bank already owes the supervisor on its capital adequacy.
The downstream link is the Supervisory Review and Evaluation Process (SREP). The ECB will use the outputs to “inform and complement” SREP, but on the qualitative side. The release is explicit that the test “is not intended to have any implications for Pillar 2 Guidance (P2G).” For 2026, in other words, this does not become a capital add-on. The number a bank produces does not feed a Pillar 2 ratchet.
That distinction needs to be read carefully. Qualitative SREP integration still has teeth: it shapes the supervisor’s narrative on a bank’s risk management, stress-testing framework and data quality, and those narratives drive the supervisory dialogue, the on-site inspection plan and — in subsequent cycles — the possibility of quantitative consequences. The 2026 exercise is the methodological foundation. The capital channel is for later cycles, if the ECB chooses to open it.
What lands for Germany
Every directly supervised German bank is in scope: Deutsche Bank and Commerzbank at the top, DZ BANK at the cooperative apex, Helaba, LBBW, BayernLB and NORD/LB across the Landesbank tier, and the long tail of German significant institutions that the SSM list of supervised entities catalogues. Each of them owes the ECB a scenario by the 2026 ICAAP cycle, and each will have its narrative read inside SREP.
The Bundesbank sits on the supervisor side of that exercise. Under the SSM, the joint supervisory team for every German directly supervised bank is jointly staffed by the ECB and the Bundesbank, and Bundesbank supervisors are the day-to-day interlocutors for those banks. The reverse-test submissions land in joint hands. The 12 December release does not name Bundesbank because it does not need to — the SSM division of labour is the standing architecture into which this thematic test plugs.
For the German banks the operational work is the construction of the scenario, not the size of the CET1 hit. A 300-basis-point depletion is severe — for Deutsche Bank or Commerzbank that is a multi-billion-euro CET1 swing — but it is fixed by the ECB. What the bank actually has to defend is the path: which exposures, which counterparties, which currency and trade-route assumptions, which operational and cyber dependencies, which sovereign exposures translate which geopolitical shock into the prescribed outcome.
The aggregate conclusions land in the summer of 2026. The supervisory consequence for any individual German bank is the SREP letter that follows. The methodology change is the news this week.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.