What the 2027 EBA stress test's new climate module actually requires

Something changed in the last few months that most credit and risk teams haven't fully absorbed yet. The European Banking Authority has published draft methodology, templates, and template guidance for the 2027 EU-wide stress test, and for the first time in the exercise's history, it includes a dedicated climate module alongside the core macro-financial scenario. The exercise covers 63 banks across the EU and Norway, but on the Norwegian side it's almost certainly only DNB large enough to participate directly, consistent with previous rounds.

That doesn't make this only DNB's problem. Methodology set at EU level tends to filter down to the rest of the sector, through supervisory expectations and eventually through market practice. This is a published draft, currently open for consultation, that will shape how Norwegian and European banks think about climate exposure for years to come, regardless of who is formally in scope for the test itself.

Why this stress test is different from the ones before it

Previous EU-wide stress tests treated climate as either absent entirely or addressed through separate, standalone climate exercises run on their own timeline. The 2027 exercise is the first to build a climate module directly into the same package as the core macro-financial scenario. That's a meaningful structural shift, even though the climate module's results currently stay separate from the core solvency outcome banks are tested on. It signals that supervisors no longer see climate risk as a specialist add-on to traditional risk assessment, run on its own schedule, but as part of the same overall exercise.

For DNB, this means concrete work delivering data and models into the EBA's framework. For the rest of the Norwegian banking sector, the more useful way to read the 2027 exercise is as an early signal of where the Financial Supervisory Authority of Norway and the EBA are likely heading for them too, well before the requirements formally arrive.

What's actually required under the new climate module

Institutions in scope must apply climate transition and flood scenarios together with the adverse macro-financial scenario, over a three-year horizon. That combination matters: the exercise doesn't ask banks to model a flood in isolation, or a carbon price shock in isolation. It asks them to model both alongside a broader economic downturn, which is a much harder and more realistic test of how these risks interact and compound.

On the transition side, the scenarios incorporate carbon pricing, country-level greenhouse gas emissions pathways, energy price shocks, and sector-specific gross value added impacts. This is designed to capture how a shift in policy or energy costs moves through real economic activity, not just through a single company's balance sheet.

On the physical side, flood risk is the named hazard, which is notable given how many types of physical climate risk exist. The focus on flood specifically, rather than a broader physical risk category, reflects both data availability and the disproportionate financial materiality flood exposure carries for real estate-backed lending.

The methodology also concentrates on exposure to non-financial corporations and to real estate specifically, rather than modelling climate impact across every asset class a bank might hold. These are the segments where the transmission channels from climate hazard to financial loss are clearest and most material, which makes them the most useful place to test methodology first.

The simplification running alongside the new complexity

It would be easy to assume the 2027 exercise is simply "more requirements." It's more accurate to call it different requirements, some of which point in the opposite direction. The EBA is cutting the total number of required data points by around fifty-five percent compared with the previous exercise, largely by drawing on data banks already submit through regular supervisory reporting rather than requiring bespoke templates built solely for the stress test.

This matters for how the module should be read. The EBA isn't only asking for more granular climate data. It's also trying to reduce the operational burden of the exercise overall, which suggests the climate module was designed to fit within a leaner process rather than be layered on top of an already heavy one.

Why real estate sits at the centre of this

There's a reasonably direct explanation for why real estate carries so much weight in the new module. Real estate exposure is measurable in a way that many other forms of climate-related exposure aren't. A flood scenario needs asset-level data to produce a meaningful result: an address, a hazard map, an estimate of physical exposure. Real estate lending is one of the few places in a bank's book where that kind of granular data is realistically obtainable today.

That practical reality has a direct consequence for anyone managing real estate-backed loans: address-level flood data stops being a nice-to-have input and becomes something closer to a structural requirement of the exercise itself.

What this means for banks that aren't DNB

It's tempting for a smaller or mid-sized Norwegian bank to read this and conclude it's DNB's problem. That conclusion doesn't tend to hold up for long. Methodology set at EU level has a strong track record of filtering down to banks outside the direct exercise, through supervisory expectations, through internal model validation standards, and eventually through market practice becoming the assumed baseline.

A Norwegian bank not directly required to run the 2027 exercise should still expect, within a reasonably short window, that its own supervisor, its own board, or its own internal audit function will start asking a version of the same questions the exercise itself asks. Can you show flood exposure at the asset level for your real estate book. Can you demonstrate how a transition shock, such as a jump in energy prices, would move through your borrowers' cash flow. Can you document the assumptions behind both of these, rather than asserting them.

Building the underlying data capability to answer these questions now, well before being formally required to, is considerably easier than doing it under supervisory pressure once the expectation has already solidified.

What "climate module" means in practice for a risk team

Stripped of the regulatory language, the module is asking for three concrete things. First, flood exposure data at the level of an individual property, not a region or postcode, for the commercial real estate portfolio. Second, a credible way to translate a transition shock, whether that's a carbon price increase or an energy cost spike, into an effect on borrower cash flow and, ultimately, probability of default. Third, documentation of the scenario logic behind both, detailed enough that a supervisor or an internal model validator could follow the reasoning without needing it explained verbally.

None of these three things require a bank to become a climate science organisation. They require access to the right underlying data, structured in a way that a credit or risk team can actually use without a specialist translating it for them every time.

The bottom line

The 2027 methodology is in consultation now and will apply to the 2027 exercise itself, which sounds like it leaves plenty of time. In practice, building asset-level data infrastructure, testing it, and getting a risk team comfortable using it takes considerably longer than most institutions expect. Waiting for the final, locked methodology before starting that work leaves very little runway, both for DNB and for the rest of the Norwegian banking sector that will face the same expectations later.

Sources

• European Banking Authority, "The EBA launches early consultation on simplified EU-wide stress test, with climate risk integration," published 11 June 2026 — eba.europa.eu

• EBA, "2027 EU-Wide Stress Test – Draft Methodological Note," 11 June 2026 — eba.europa.eu

• ESG Today, "EBA Integrates Climate Risk into EU Banking Stress Test," June 2026 — esgtoday.com

• NordSIP, "EBA Brings Climate Risk Into 2027 Bank Stress Test," 18 June 2026 — nordsip.com