CBDC Holding Limits May Backfire: Research Warns Caps Can Trigger, Not Prevent, Digital Bank Runs

Holding limits — caps on how much digital money a person can keep in a retail central bank digital currency (CBDC) wallet — are often presented as a simple safety valve. By restricting balances, central banks could curb sudden shifts of deposits out of commercial banks, reducing the risk of destabilising “digital bank runs” in a crisis. But a growing body of research suggests that these limits are not just a technical constraint. They are also a behavioural signal that can reshape how people choose to save, spend and manage their money — sometimes in ways that are not immediately intuitive.

Recent experimental evidence from euro-area consumers underscores the point. In an incentivised online study spanning four euro-area countries, researchers tested how people respond when the presence and level of a CBDC holding cap are varied exogenously. The design allowed them to disentangle two forces: the mechanical effect (people simply cannot hold more than the stated maximum) and the behavioural effect (people change their actions because a cap exists, even if they never intend to reach it).

That separation matters because policy debates frequently assume the cap’s impact is linear and predictable: lower limit equals lower CBDC holdings. The experiment indicates the reality is more nuanced. Even when participants were not immediately constrained by the cap, some adjusted their decisions in anticipation of hitting it later, or because the cap changed how they perceived the CBDC’s intended purpose.

Limits as a message, not just a rule

A holding limit can be interpreted by users as a cue about what a CBDC is “for.” A relatively low cap may implicitly frame the CBDC as a payments tool — digital cash for transactions — rather than a savings vehicle. That framing can lead people to keep balances deliberately below the threshold as a precaution against inconvenience, such as a payment being rejected if a top-up would breach the cap. In consumer finance, such “buffer” behaviour is common: people avoid operating right at the edge of constraints, whether credit limits, overdrafts, or mobile data caps.

The study’s findings suggest a similar dynamic for CBDC wallets. A cap can therefore suppress holdings by more than the amount mechanically forced by the limit, because users voluntarily steer clear of it. In policy terms, this implies that the behavioural response could help meet financial-stability goals: lower CBDC balances without the central bank having to set an extremely restrictive ceiling.

But the direction of behavioural effects is not guaranteed. Some participants may treat a cap as a focal point — a number that feels like a recommended maximum — and aim closer to it than they otherwise would have. Others may see a higher limit as an assurance that the CBDC is suitable for broader use, increasing trust and willingness to hold it. In that case, raising the cap could spur demand, potentially amplifying deposit outflows from banks in stressed conditions.

Designing for crises and for normal times

Holding limits are often justified by worst-case scenarios. In a crisis, a risk-free CBDC could look more attractive than a bank deposit, especially if consumers can convert deposits into CBDC instantly. A binding cap would mechanically stop large-scale shifts by individuals, but the behavioural layer complicates that picture.

If consumers interpret limits as a sign that access could be rationed in turbulent periods, they might pre-emptively increase CBDC holdings in calm times — “just in case” — up to whatever cap is available. That precautionary motive could be reinforced if people fear the cap might be lowered later, or that conversion channels could become congested during market stress. Such responses would change the steady-state demand for CBDC, not merely crisis dynamics.

The cross-country nature of the euro-area experiment is important because payment habits and trust in institutions vary across member states. While the study does not treat the euro area as behaviourally uniform, it provides evidence that design parameters have broadly comparable psychological effects even across different consumer contexts — a relevant point for any multi-country currency union considering a harmonised retail CBDC.

Implications for policymakers

The takeaway for central banks is that a holding limit is a behavioural instrument as much as a quantitative one. It can influence adoption, usage intensity, and perceived legitimacy. That creates both risk and opportunity. On one hand, poorly calibrated caps may inadvertently depress uptake, undermining public-policy objectives such as ensuring access to central bank money in a digital age. On the other, well-chosen limits could reduce financial-stability risks not only through hard constraints but also through voluntary self-limitation by users.

Researchers also caution that working-paper evidence should be read as part of an evolving discussion rather than a final blueprint for policy. Even so, the experimental approach offers a rare causal window into how real people might respond to CBDC design — an area where traditional economic modelling can struggle to capture psychological and operational frictions.

As retail CBDC proposals advance, the lesson from this research is straightforward: sometimes, “less” — a tighter design constraint — can produce “more” stability, not only by restricting balances but by reshaping behaviour in ways that multiply the cap’s effect.

Why it matters

CBDC holding limits are often treated as a technical parameter, but evidence suggests they can change consumer behaviour well beyond their mechanical role. That means central banks could inadvertently steer adoption and financial stability outcomes through the signals these caps send. Understanding those behavioural channels is essential before turning CBDC design choices into real-world policy.