On 11 June 2026 the EBA published the draft methodology, templates and template guidance for the 2027 EU-wide stress test. Two headline moves: a 55 % cut in required data points anchored to regular supervisory reporting, and the first-ever climate module alongside the macro-financial scenario. The formal EBA consultation closes on 10 July 2026; industry associations consolidate member responses through August 2026. What CROs file in the next eight weeks will define the SREP cycle that follows the results.
The 2025 exercise ran on templates that duplicated large parts of COREP, FINREP and AnaCredit. Banks fielded dedicated stress-test teams whose primary job was reconciling numbers that already existed in supervisory returns. The EBA has read that signal. The 2027 methodology cuts required data points by 55 % — not by reducing scope, but by drawing the bulk of the input from regular supervisory reporting and eliminating dedicated stress-test datapoints that overlapped with existing returns.
Reference balance sheet, exposure classes, NPE flows and P&L building blocks are pulled from COREP, FINREP and AnaCredit rather than re-collected. Consequence: data lineage becomes the audit surface, not the stress-test template.
Transition and physical risks are integrated alongside the macro-financial scenario, in a dedicated module that does not affect core stress-test results at this stage. The EBA is measuring, not sanctioning — yet. That distinction matters for CFO capital-planning narrative.
Stress-test-specific definitions of PD, LGD, NII floors, credit-risk transitions and market-risk sensitivities have been trimmed and re-mapped to the regulatory taxonomy where possible. Effect: less internal translation work; more visibility for supervisors on where bank-specific assumptions still bite.
Draft methodology is out roughly seven months earlier in the cycle than 2023 or 2025 publications. The EBA has stated this reflects industry feedback from the May 2026 pre-consultation. Workshops with banks are scheduled through Q3.
Bank-side read: the exercise is easier to file, harder to hide behind. When the supervisor gets the same input via COREP and via the stress test, discrepancies are visible without a reconciliation exercise. The value of a clean data lineage rose materially on 11 June.
The 2027 exercise integrates a dedicated climate module — the first time transition and physical risks sit alongside the macro-financial scenario inside the EU-wide stress test. The EBA has been explicit that at this stage the climate module does not affect the core stress-test result. Read that as: measurement year one, not sanction year one.
That framing is helpful for the CFO capital-planning narrative in 2027. It is not a reason to under-invest in the module. Three points every CRO should have in their board pack by September 2026:
The module you populate in 2027 sets the definitional baseline the EBA will re-use, harden and eventually sanction against. Under-investing today means re-doing the work in 2029 under supervisory pressure. Better to shape the taxonomy while it is still consultative.
Transition risk maps loosely onto existing sector-level exposures. Physical risk requires geo-referenced collateral data and hazard-model outputs. Most European banks have neither at production quality. Sourcing a credible physical-risk dataset by Q1 2027 is the practical blocker.
The climate module outputs will be read by supervisors as capital-planning inputs, not as ESG disclosure. If the sustainability team owns the narrative alone, the JST will ask questions the bank cannot answer without the CFO’s and Risk’s desks in the room.
Consultation responses that shift EBA landing zones share a pattern: they are specific, sourced against the draft methodological note, and address a numbered paragraph. Generic pushback (“too burdensome”, “too prescriptive”) does not move the pen. Below are the six positions we are seeing draft into industry letters this cycle — each mapped to a paragraph of the draft note.
Draft methodology anchors calibration to a fixed lookback. If the window excludes the Covid-19 and rate-shock periods, transition intensities under-represent tail credit dynamics. Position: seek an explicit tail-inclusion window or a supplementary calibration overlay in Section 3 of the methodological note.
The 2025 exercise imposed prescribed beta floors that materially penalised NII-heavy business models in rate-cut scenarios. The 2027 note appears to preserve the same architecture. Position: seek country- or portfolio-level differentiation, evidenced by AnaCredit-derived actuals, replacing the flat floor.
The draft climate module prescribes a sector taxonomy. Many bank exposures sit across two NACE codes (agri-industrial, transport-logistics, mining-refining). Position: request explicit hybrid classification rules to avoid arbitrary primary-code allocation that would materially move transition-risk exposures.
The methodology gestures at FRTB alignment but leaves gaps in how banks not yet on FRTB IMA should populate market-risk sensitivities. Position: request explicit hierarchy of fall-back sensitivities (FRTB SA, current IMA, historical simulation) with weightings.
The methodology should not re-import loss-data assumptions that CRR3 explicitly retired. Position: confirm that operational-risk stress projections track the BIC-based SA methodology under CRR3 (ILM fixed at 1.0 in EU transposition), not a hybrid.
EBA notes explicitly that QA is out of scope for this consultation. That leaves banks without a public process for challenging JST-imposed overlays (top-down calibration adjustments). Position: request a documented, appealable QA challenge process, with a defined SLA for JST responses.
The response letter that lands is the one where each position is (a) tied to a paragraph number in the methodological note, (b) evidenced by a bank’s own COREP / FINREP / AnaCredit numbers or a published EBA / ECB source, and (c) offers an alternative wording the EBA can adopt. Anything less is background noise.
Because stress-test inputs now overlap with supervisory reporting, the SREP 2028 conversation will not be a debate about which number is right — the JST will already know. The debate will be about interpretation, capital planning implications, and the credibility of the bank’s narrative. Three shifts worth planning around.
If a bank’s stress-test PD, LGD or NPE stock does not tie back to COREP and AnaCredit within tolerance, that is now a first-order finding — not a template-reconciliation exercise. The BCBS 239 lineage evidence a bank had to file for its own governance now sits in the JST’s hands.
Even though climate does not affect the core result in 2027, EBA-level results feeding SREP 2028 will contain a climate-risk read for every participating bank. That read is a natural anchor for P2G recalibration in the 2028 – 2029 cycles. The banks that surface their climate story clearly in the module have a stronger negotiating floor when P2G is set.
With definitions harmonised, deviations from the common baseline stand out. Every internal overlay (business-model uplift, one-off adjustment, jurisdiction-specific credit assumption) becomes a JST question. The playbook is not to abandon overlays; it is to document them cleanly and defend them by paragraph reference to the methodological note.
Do not wait for the January 2027 reference date to discover that COREP, FINREP and AnaCredit do not tie internally. Run a mock T−0 reconciliation on 31 March 2026 data by end-Q3. If gaps exceed tolerance, six months is a comfortable remediation window. Three is not.
Physical-risk data spine, transition-scenario mapping, sustainability × Risk × Finance ownership matrix — all belong on the CRO desk by September, not December. First-year measurement quality sets the anchor for future sanctioned exercises.
Every internal overlay you plan to apply against the common methodology needs a one-page defence: business rationale, quantitative evidence, methodological-note paragraph reference. The pack is not for the EBA. It is for the JST challenge session in Q1 2027. Build it now while memory is fresh.
Ezelman runs the T−0 reconciliation, climate-module stand-up and overlay defence pack as a single mandate under the ANCHOR methodology. Ninety days, senior-only, delivered to the CRO’s ExCo — not to the working group.
Dates, paragraph numbers and document versions verified at publication. Where consultations are still open, we flag draft status in text. Corrections: research@ezelman.com.
Partner-led, senior-only, no audit conflict. If the question in this piece is live at your institution, the first conversation is with a partner — not a BD team.
Talk to a partner →