In June 2026 the Eurosystem set out the milestones for its Integrated Reporting Framework, IReF: a one-year pilot from the second quarter of 2030 and go-live in the second quarter of 2031, followed by a period of parallel reporting. It brings banks' separate statistical reporting requirements into one framework. Five years can sound like a long way off. For anyone who has run a reporting change, it is not.
What the ECB actually decided
The decision is about timing and commitment. IReF moves from design to delivery, with dates that banks, national central banks and software providers can plan against. The ECB says the goal is a lighter reporting load for banks over time, by integrating requirements that are reported separately today (ECB). Underneath that is a bigger shift: from templates designed around questions to granular data from which many questions can be answered.
Why it matters outside the euro area
IReF is a euro area project, but the direction is global. The Bank of England's February 2026 discussion paper on Future Banking Data reviews what the PRA collects and how often (Bank of England), and authorities in Asia and the Americas have their own granular programmes. When the largest supervisors move from templates to data, vendors, standards and expectations follow. Smaller authorities will feel the pull whether or not they plan for it.
What changes for supervisors
Three things. The data model becomes the asset that matters, so it needs owners, versioning and governance like any other supervisory instrument. Validation has to work at a much larger scale and as data arrives, not in a batch after the deadline. And analysts move from reading templates to querying data, which needs different tools and skills. None of these waits for 2031. Each can start with one dataset.
A practical path for smaller authorities
You do not need IReF's scale to get its benefits. Pick one question your templates answer badly, collect the smallest granular dataset that answers it, run it alongside the existing return and prove the numbers reconcile. Our guide to granular data reporting sets out the steps.
- 1Name an owner for your supervisory data model, separate from any single return.
- 2List the figures you collect more than once across returns, and pick one to define once.
- 3Choose one granular dataset to pilot alongside an existing template.
- 4Make sure your validation runs as data arrives, not after the deadline.
- 5Ask your platform provider how a move to granular collection would work, and how long it would take.
Source: ECB, press release, 8 June 2026. For the standards behind it, see data standards, explained.
For supervisors
New guides and Insights, by email
One short email when we publish something worth your time, from Jess Luper. No spam, unsubscribe any time.