Higher CBDC holding limits can increase balances rather than cap demand
High holding limits increased digital currency balances in an experiment reported in European Central Bank (ECB) Working Paper 3300, published October 15, 2026. Automatic funding increased adoption while reducing holdings.
A ceiling becomes an anchor
Participants allocated 10 tokens under no limit, a 3- or 8-token limit, then a limit with reverse waterfall funding.
Overall holdings fell from 2.68 to 2.35 tokens after limits appeared.
Among initially unconstrained participants, however, holdings rose from 2.04 to 2.23 tokens (p=0.005): 21.0 percent increased balances and 16 percent reduced them.
Low-limit holdings fell from 2.80 to 1.93 tokens; high-limit holdings rose from 2.56 to 2.76 (p=0.07).
Attraction and repulsion offset under the lower limit.
Reverse waterfall funding automatically converted bank funds for payments, raising adoption from 85.2 to 92.1 percent and lowering overall holdings to 2.05 tokens.
Exactly one-token holdings rose from 19.8 to 44.8 percent.
Real incentives, stylized payment choices
Thomas Garcia, Claudia Lambert, Cosimo Pancaro, Claire Rimbaud and Alice Soldà conducted the preregistered experiment on Prolific in November 2025.
The 761 participants came from France, Germany, Italy and Spain.
The bank-card instrument paid 7 percent interest and had 90 percent seller acceptance; the CBDC proxy paid no interest, had universal acceptance and cost 0.5 tokens to adopt.
Conditions followed a fixed order, with no feedback until completion.
The authors distinguish mechanical balance restrictions from behavioral attraction and repulsion.
Experimental evidence, not policy calibration
The results support combining holding caps with automatic funding to separate payment use from stored balances.
The authors caution against generalization: their approximate €1,000 and €5,000 equivalents are exploratory, not calibrated policy thresholds.
These are authors' findings, not the ECB's views.