The digital euro is now at its most consequential stage. Trilogue negotiations opened in Brussels on 13 July, with the Irish presidency targeting political agreement before year end. Two questions that were flagged as needing resolution first – holding limits and merchant compensation – go to the heart of who the design actually protects.
The clearest answer so far has come from the design’s own architects. In a February 2024 piece for the European Central Bank’s blog and a longer companion column for VoxEU, Ulrich Bindseil, Piero Cipollone and Jurgen Schaaf wrote plainly that central banks and legislators ‘have not endorsed’ views favouring bank disintermediation, but ‘instead defended the role of banks and have designed [central bank digital currencies] accordingly’.
On the business side, where the digital euro’s holding limit is currently set at zero, they were equally direct: merchants would be able to process the digital euro but not hold it, specifically to ‘help protect the corporate deposit base of the banking system.’ That is not an inference about the design’s purpose. It is the ECB’s own senior officials stating the purpose outright.
What do the numbers show?
The formal stability case is considerably thinner than that admission would suggest. Under the scenario the ECB calls most likely, ordinary use with no crisis involved, a holding limit of €3,000 for consumers is estimated to cost banks 8-18 basis points of net interest income, a rounding error against normal year-to-year swings in profitability.
Even under an extreme, historically unprecedented bank run scenario, only nine of roughly 2,000 banks assessed would risk breaching their liquidity buffers, and the ECB states none would be a significant institution. A companion study modelling bank responses in detail found that real concerns would only emerge above a limit of €5,000.
None of that proves the current figure is wrong. It does show the number sits well inside a margin the ECB’s own analysis treats as safe, which raises the question of what the margin is actually protecting, if it is not ‘financial stability’.
Why the ‘bank run’ fear is overstated
The deposit flight argument does not hold up as cleanly as the political rhetoric assumes. When money moves from a bank account into the digital euro, it does not leave the financial system, it moves onto the ECB’s own balance sheet, and the ECB can lend that money straight back to the bank that lost it. The Centre for Economic Policy Research’s 2026 Digital Money report formalises this as a neutrality result: if the central bank recycles displaced funding back to banks on terms comparable to what deposits offered, bank lending capacity is largely unaffected. Only the price of that funding changes.
That result depends entirely on the central bank actually choosing to recycle funds, and the ECB has a direct, large-scale precedent for doing exactly that. TLTRO III, the ECB’s targeted lending programme, offered euro area banks three-year loans at rates as low as minus 1%, conditional on banks meeting a real-economy lending benchmark, and the programme peaked at roughly €2.2tn outstanding.
This was not a hypothetical mechanism. The ECB has already demonstrated, at real scale, that it will price funding specifically to sustain bank lending through a shock. TLTRO III is the clearest evidence available that a digital euro-induced deposit shift need not translate into reduced credit supply, provided the ECB is willing to act the way it already has.
Who’s actually exposed, and who actually benefits?
If the underlying risk is this contained, and the ECB already has the tools to manage it, why does political attention concentrate on a €3,000 number rather than on these tools? Part of the answer is which banks are actually exposed.
The ECB’s own extreme scenario identifies a narrow population, small, unusually deposit-reliant institutions without meaningful capital markets access, similar in profile to community banks in the US. The US addresses that specific vulnerability with a narrow instrument, a ban on stablecoin remuneration, rather than a blanket cap on all users. Europe’s equivalent population may be larger than the ECB’s sample captures.
Germany alone has a substantial number of small savings banks and co-operative institutions that are deposit funded and locally rooted in much the same way, and the ‘nine of 2,000’ figure is drawn from the ECB’s assessment sample rather than the full European banking population. The exposure this argument is actually protecting against may sit outside the number most commonly cited.
The other part of the answer is what the limit costs in profit terms, not stability terms. CEPR’s Digital Money report puts a figure on it: using US data, the portion of bank net interest margin exposed to this kind of competitive pressure, the part attributable to market power rather than ordinary competition, averages roughly 0.2% of GDP over the past 50 years. That is a real, calculable cost to a specific industry. It is not a threat to the financial system as a whole, and the report’s own conclusion is direct: bank pressure for tight limits looks like a defensive move to protect that profit margin, not a stability measure.
The zero limit for businesses deserves separate scrutiny on these grounds. Martijn van der Linden has argued that a digital euro constrained too tightly cannot achieve the adoption needed to become a credible alternative to foreign payment networks and dollar-denominated stablecoins, a strategic autonomy argument for a higher limit, not a lower one.
The same logic applies with more force to the business case specifically. A business barred from holding a digital euro cannot use it as working capital or for direct supplier payment; every euro received must re-enter conventional banking rails immediately. That’s a real constraint on any future ambition to develop the digital euro into a genuine cross-border and Treasury instrument rather than a purely domestic retail one.
The bigger question hiding underneath
There is a further implication worth stating plainly, given how directly TLTRO III illustrates it. That programme’s cheapest rate was conditional on banks meeting a lending benchmark, which means the ECB was not simply offering neutral funding on request, it was already steering where credit flowed preferentially. Extend that mechanism to a digital euro-induced recycling operation and the same logic follows: the ECB gains real, discretionary influence over the terms on which credit reaches the economy. That is not a hypothetical extension of the neutrality result. It is what the neutrality result implies once it is actually put into practice.
This isn’t radical or unprecedented. Governments already do versions of it quietly. A government-backed mortgage guarantee makes borrowing cheaper below a certain home price. There’s a legal ban on financing cluster munitions. Tax breaks steer money towards renewable energy. None of this gets called credit guidance, but that’s exactly what it is: public authorities nudging where money flows, using tools they already have. And as long as elected officials do this with public oversight, that’s fine. Whether that means the European parliament setting any future criteria and the ECB handling technical execution under its secondary mandate, or some other arrangement, is a separate question.
There’s precedent for doing this more openly, too. From 1945 until the early 1980s, France’s central bank actively steered credit towards reconstruction priorities, farming, rail, electricity and exporters, while leaving everyday lending decisions to ordinary banks. Germany, Italy and Belgium ran similar systems. It is worth being clear about where this shouldn’t go – China’s version today is far more centrally directed, and that isn’t the model in question here. The useful precedent is the lighter post-war European one: better terms for priority activities, not government orders about who gets a loan.
What follows from that, how such influence should be exercised and whether it requires democratic oversight beyond the ECB itself, is a large enough question to deserve its own treatment rather than a closing paragraph here. For now, the more immediate point stands on its own. The digital euro’s holding limit has been debated almost entirely as a matter of crisis prevention. The ECB’s own numbers, its own admissions and its own precedent for recycling liquidity at scale all suggest the real stakes lie elsewhere, in a contest over a specific, calculable bank privilege that the stability framing has largely kept out of view.
Paul Helmich is an independent financial economist. This piece draws on his working paper ‘Whose interest does the digital euro’s holding limits serve?’, available on SSRN and forthcoming with Rosa & Roubini.
OMFIF will be in Bangkok for the International Monetary Fund-World Bank meetings to discuss the evolving role of digital public money. Register to attend here.
Interested in this topic? Subscribe to OMFIF’s newsletter for more.