A high digital euro holding limit may encourage larger balances, experiment suggests

A high holding limit may lead some users to keep more money in a digital wallet, even if the limit does not require them to change their balance, according to an experiment published by the European Central Bank (ECB) on 15 October 2026. The research shows that designing a possible digital euro must also take people’s response to the communicated limit into account, alongside the mechanical effect of the cap.

In brief The experiment involved 761 people from France, Germany, Italy and Spain, who allocated tokens between two payment instruments. The high limit increased the average balances of participants whose initial choices were not constrained by the limit. With automatic funding from a linked account, more participants allocated money to the digital instrument, but the average balance fell. The study does not establish a euro amount for the future limit. Legislation is still being negotiated, and the ECB has not decided to issue a digital euro.

The paper, authored by Thomas Garcia, Claudia Lambert, Cosimo Pancaro, Claire Rimbaud and Alice Soldà, was published in the ECB’s Working Paper series as No. 3300. The results and interpretations belong to the authors and do not necessarily represent the institution’s official position.

A digital euro would be an electronic form of central bank money intended for everyday payments and available alongside cash. Unlike a deposit at a commercial bank, it would represent a direct claim on the central bank. The current project envisages a non-interest-bearing instrument geared towards payments.

The possibility of moving money from bank deposits into such a wallet raises questions about bank funding. The holding limit is intended to reduce the accumulation of large amounts in digital euro. It would apply to the balance of that wallet, not to a person’s total savings or money in bank accounts.

What the researchers tested

The experiment was conducted online through the Prolific platform in November 2025. The 761 participants made incentivised decisions, allocating ten tokens between two payment instruments presented under neutral names. One offered compensation for allocated amounts but was not accepted by all simulated sellers. The other, inspired by a central bank digital currency, offered no interest and was accepted by all sellers in the experiment.

Participants were randomly assigned to two groups, one with a limit of three tokens and the other with a limit of eight tokens. Each went through, in the same order, the no-limit situation, the situation with a limit, and the situation in which the limit was accompanied by automatic funding. The tokens and limits tested are not euro amounts approved for the future system.

The researchers compared the initial choices with those made after the rules were introduced. They focused particularly on participants whose initial balances did not exceed the limit, in order to distinguish the direct constraint from the change produced simply by communicating a limit.

A limit can become an anchor

With the eight-token limit, average balances increased among participants who were not constrained by the limit. With the three-token limit, attraction and rejection effects offset each other in this group. The authors interpret the result as evidence that a sufficiently high limit can become an anchor towards which some users’ choices are drawn.

This does not mean that every participant chose a higher balance or that any high limit would automatically result in additional withdrawals from banks. The result concerns allocations made in an experimental setting. The effects on balances and deposits after a possible launch of a real digital currency must be assessed separately.

How automatic funding changes the balance choice

The study also examines a feature called reverse waterfall, which would automatically cover the amount needed for a payment from a linked bank account. In an explanatory example, if the wallet contains €10 and the payment is €30, the €20 difference could come from the bank account if the user has activated the link and has sufficient funds. These amounts illustrate the mechanism and are not the values tested in the experiment.

The researchers defined adoption of the digital instrument as allocating at least one token to it. With automatic funding, the proportion of participants making this choice rose from 85.2% to 92.1%, an increase of 6.9 percentage points. At the same time, the average balance for the full sample fell from 2.35 to 2.05 tokens. The indicator reflects the choice to hold money in the experimental instrument, not the frequency of actual payments.

The waterfall feature would operate in the opposite direction, transferring to the linked bank account the portion of a receipt that exceeded the digital wallet’s limit. The European Commission’s June 2023 proposal includes both the transfer of excess funds upon receipt and the topping up of the balance for payments. Thus, a limit on the amount held in the wallet would not automatically become a limit on the value of each transaction.

In the authors’ interpretation, the limit and automatic funding must be analysed together. The possibility of covering a payment from the linked account may reduce the need to transfer money in advance into a separate wallet. In the experiment, the mechanism supported adoption without the accumulation of larger balances.

What remains to be decided for the digital euro

The sample recruited through Prolific should not be treated as a representative survey of the entire euro area. The fixed order of the conditions is a methodological limitation acknowledged by the authors, who postponed feedback until the end to reduce learning between stages. The experiment did not include participants from Romania or the Republic of Moldova and does not allow direct conclusions about financial behaviour in those countries.

The legislative process is continuing separately from the research. The EU Council agreed its negotiating mandate on 19 December 2025, while the plenary of the European Parliament confirmed its mandate for interinstitutional negotiations on 9 July 2026. According to the ECB, trilogue negotiations are under way. These stages do not amount to the final adoption of the regulation.

In its project presentation of 6 October 2026, the ECB states that a decision to issue the digital euro will be taken only after the legislation is adopted. The objective of being ready for a possible first issuance in 2029 depends on completing the legislative process by the end of 2026 and on technical preparations. The study contributes to the design discussion without establishing the final limit or a guaranteed launch date.