Non-binding is not the same as ineffective: what the ECB concedes while tidying up

On 26 June 2026 the European Central Bank (ECB) announced that it had reviewed its supervisory publications and discontinued around 40 of roughly 130 documents. The coverage read it as a cut in red tape. Taken on its own it is nothing of the sort: retiring papers that were already outdated or superseded changes no expectation currently in force. The story sits elsewhere. It sits in a subordinate clause from Frank Elderson, and it describes a problem the supervisor has never conceded quite this plainly.

Whether this belongs in the filing cabinet or in the next board paper turns on one distinction: housekeeping or course correction. It is both, though not in the proportions the coverage suggested.

The sentence that matters

Frank Elderson starts with the textbook position: "However, supervisory guidance does not create new legal requirements beyond the applicable prudential framework. Binding obligations come only from directly applicable European regulations or directives as transposed into national law." Guidance creates no legal duty. Only a regulation, or a directive once transposed, binds anyone.

Then comes the line that sets the law against the lived experience: "The fact that a guide is non-binding does not mean that it is irrelevant in practice."

Anyone who has defended an audit finding or worked a Supervisory Review and Evaluation Process (SREP) dialogue knows the mechanism behind that line. Formally a guide is the supervisor's opinion. Practically it is the yardstick a bank is measured against. Departing from it is not wrong, but it does need explaining, and explaining costs more in a supervisory dialogue than compliance does. Out of that asymmetry grows a binding force that no legal act ever conferred.

Frank Elderson names that gap as the job to be done. One question guiding the review was whether the ECB could make it clearer that supervisory expectations are not legally binding. A related one was how day-to-day supervision could be brought into line with that non-binding character. This is not a communications exercise. It is an instruction to the supervisor's own examiners.

Where the substance actually sits

Anyone hunting for a tangible outcome will not find it among the 40 retired documents. It sits in a single clarification.

The revised guide to the Internal Capital Adequacy Assessment Process (ICAAP) settles the status of the management buffer, and Frank Elderson puts it with unusual force. The revised guide, he writes, makes it "crystal clear that the management buffer reflects banks' own forward-looking capital planning above the minimum requirement and that it does not constitute an additional capital requirement. It is not a supervisory add-on."

For capital planning and regulatory affairs, that is the sentence of the whole exercise. The management buffer has been a running sore in the dialogue for years: banks planned it as their own margin of safety and then watched it treated in practice as though it were an extra requirement. The supervisor's rejecting that reading changes nothing in the law. It changes the negotiating position, which in this field amounts to the same thing as money.

A guide becomes a report

The second piece of substance tends to get overlooked, although it carries the more consequential decision.

In July 2024 the ECB put a draft guide on governance and risk culture out for consultation. More than 1,000 comments came back. The result is not a revised draft but a change of format: the draft will be replaced by a report on good practices, planned for the first quarter of 2027. The ECB attributes this to the feedback and to the ongoing review of the European Banking Authority (EBA) guidelines on internal governance.

This is more than a delay. A guide states expectations; a report on good practices records observations. It drops one rung on the ladder from binding to advisory, and it drops in a field the supervisor itself treats as central. Frank Elderson leaves no doubt on that score: it is weak governance, he argues, that lets risks build unchecked until they strike capital and liquidity, which makes bad governance the earliest and most reliable warning that an institution is heading for trouble.

A supervisor that calls a topic its earliest warning signal, and then deliberately publishes a collection of good practices rather than a rulebook on it, is in earnest about the non-binding part. Whether that increases the effect or reduces it remains an open question.

What is coming, and what that means for the calendar

Precision matters here, because the coverage has muddled publication dates with consultations.

Three guides are being substantially revised: the guide to licence applications, the guide to on-site inspections and the guidance on leveraged transactions. Those revisions are under way and are due to conclude by the end of this year. For the licensing guide the ECB names the third quarter of 2026 as the publication date, and for the guide on risk data aggregation and risk reporting the fourth quarter of 2026. These are dates on which finished texts appear, not consultation windows. Consultation is promised only where substantial changes are needed, and that promise is general rather than tied to the guides above.

So anyone keeping a calendar should book publications for the second half of 2026, not deadlines for comment. The one substantial consultation in this story already happened.

Looking further out, the ECB says it will work towards a single guidebook consolidating its supervisory publications by theme. That would be the real structural break. No date has been put against it.

The counter-argument

Two objections can be raised against the weight of all this, and the first is strong.

First, this is the supervisor's own account of itself. An authority announcing that it will henceforth press less firmly than its papers sound has not thereby changed a single examiner. The de facto binding force is made in meetings, not in blog posts. Whether the joint supervisory teams behave differently will show up in the next SREP cycle at the earliest, and it will be hard to measure.

Second, and this is a matter of fairness: the flood of paper was not visited upon the banks; it was ordered by them. Frank Elderson records that the documents were issued for good reason, often at the request of banks, to improve transparency and consistency. Anyone complaining today about the volume helped create it. More clarity about the guidance's non-binding status necessarily means less instruction, and that shifts the burden of interpretation back onto the institution.

From which follows a risk the ECB does not address: a bank that took its bearings from a now-retired guide is left without a reference. The texts stay online marked as discontinued, but an archived document makes a poor justification. That gap goes unmentioned in the published material, and it is a question rather than a proven criticism.

Recommendations for practice

For regulatory affairs, compliance and capital planning, four areas deserve attention.