Higher bank deposit rates appear to have a greater effect on USDC adoption than on USDT, according to a Bank for International Settlements study published on 13 October.
The research found that a one-basis-point rise in deposit rates was associated with about 4.8% less cumulative USDC growth over 10 weeks. The equivalent reduction for USDT was 1.9%.
Rashad Ahmed and Inaki Aldasoro examined weekly data from January 2019 to June 2025. The October publication followed an earlier version released on 24 March, which had already identified the stronger response from USDC.
The study measured deposit rates using an average of six national savings and money-market deposit-account series. That measure is separate from the Federal Reserve’s policy rate and from the rate offered by any individual high-paying account.
Its baseline ordinary least squares (OLS) estimates indicated that deposit-rate increases were linked to slower growth in both stablecoins. For the market as a whole, the estimated reduction was about 2.8% in cumulative growth over 10 weeks for each basis point of repricing.
The authors said USDC’s larger response was consistent with its closer links to US institutional and decentralised finance users. Those holders may be more likely to compare bank deposits with stablecoin balances. They associated USDT’s smaller response with its offshore and developing-economy use.
The data cannot show which holders moved their funds or identify the bank accounts involved, meaning the explanation for the difference between the issuers remains an interpretation of the results.
A second model, using an instrumental-variable approach intended to isolate deposit repricing, produced a larger estimate of about 9.1% less growth. It used the fact that banks passed on policy-rate changes more strongly when the federal-funds target midpoint four weeks earlier had been at least 3.125%.
The model also controlled for policy and Treasury rates, Bitcoin prices, the dollar index and market volatility. However, the authors noted that the approach had a first-stage F-statistic of 9.78, below the conventional threshold of 10. Since the sample contained only one high-rate episode, they said the larger estimate had limited support when applied to another rate cycle.
In the instrumental-variable charts, USDC’s 10-week 95% confidence interval stayed below zero, while USDT’s included zero. Comparisons between different deposit tiers also supported the direction of the relationship. Tests covering the Silicon Valley Bank episode, crypto events, trading volume and DeFi yields continued to show negative relationships, although results for balances supplied to lending protocols were imprecise.
The study said exchange rewards introduced another source of competition for stablecoin balances. Its historical analysis used Coinbase and Binance as examples of different payment models, while acknowledging that the distinction was not absolute.
Under reserve-based remuneration, an exchange passes some income from an issuer’s reserve assets to holders. Under activity-based remuneration, returns are linked to lending, trading or other market activity. Coinbase’s historical USDC yields followed policy and reserve yields more closely, while Binance’s were more exposed to crypto activity.
Coinbase’s current terms describe USDC Rewards as a loyalty programme funded by the exchange, without using or lending a customer’s USDC without instruction. Customers must be in an eligible region, be eligible to trade and hold at least $1 of USDC. Coinbase One membership is also required in several regions, including the US and UK.
Rates vary by country, account type and ongoing experimentation, and may be changed or withdrawn. Coinbase directs customers to account-specific rates rather than offering one universal annual percentage yield. It also says USDC balances are not deposit accounts and are not covered by FDIC or SIPC insurance.
The authors cautioned that the findings do not directly measure lost bank lending. That would require evidence on reserve locations, replacement funding and changes in credit provision. They also said the coefficients cannot be applied directly to a Federal Reserve rate move, because bank repricing and the return available on each platform are separate factors.
A useful forecast, the study concluded, would distinguish between issuers, holder preferences, bank repricing and access to exchange rewards.
