BIS paper links higher deposit rates to slower stablecoin growth

A Bank for International Settlements working paper published October 13 found evidence that reserve-backed stablecoins and U.S. bank deposits act as partial substitutes: when banks raised deposit rates, stablecoin market capitalization subsequently grew more slowly.

The result matters for monetary policy and stablecoin regulation because it suggests that competition for liquid dollar balances extends beyond banks and money-market funds. It does not prove that stablecoins drain the banking system, cause lending to contract or respond only to interest rates.

The study follows weekly data through June 2025

Researchers Rashad Ahmed and Iñaki Aldasoro analyzed weekly U.S. observations from January 2019 through June 2025. They related changes in bank deposit rates to stablecoin market-capitalization growth over subsequent weeks while controlling for policy rates, Treasury yields, exchange rates, bitcoin prices, financial-market volatility and broader cryptocurrency conditions.

The paper uses local projections, a statistical method for estimating how an outcome evolves after a change in another variable. To address the possibility that deposit rates and stablecoin demand influence each other, the researchers also exploited a nonlinear feature of the data: banks repriced deposits more aggressively once the federal funds rate exceeded approximately 3%.

Against average stablecoin growth of 15% over a 10-week horizon in the study sample, the ordinary least-squares specification associated one basis point of deposit repricing with roughly three percentage points less growth. A one-standard-deviation repricing move was associated with about 3.75 percentage points less growth. These are model estimates for the January 2019–June 2025 sample, not forecasts for current supply or a claim that every bank-rate increase will produce the same response.

The estimated relationship was stronger for USDC than for USDT. The authors interpret that difference as consistent with USDC’s closer connection to U.S. depositors. They also found similar results across different deposit products and stablecoins supplied to decentralized-finance protocols, while their placebo test did not find the same pattern for non-stablecoin cryptocurrencies.

Partial substitution is not a deposit-drain estimate

The paper measures how stablecoin capitalization changed after deposit-rate repricing. It does not observe each user moving money from a bank account into a token, and it does not calculate the net change in aggregate bank deposits.

That distinction is central. A Federal Reserve staff analysis says stablecoin purchases can reduce, recycle or restructure deposits depending on who buys the tokens and how issuers invest their reserves. If an issuer holds reserves at commercial banks, money leaving retail accounts may return as more concentrated wholesale deposits. If reserves instead move into Treasury securities or central-bank balances, the banking-system effect can differ.

The BIS study also spans one recent transition into a federal-funds-rate environment above the approximately 3% threshold used for identification. That provides useful variation, but it limits confidence that the estimated magnitude would remain stable across future rate cycles, different regulatory regimes or much broader payment adoption.

The policy implication is narrower than the headline risk

The evidence supports treating deposit pricing as one driver of stablecoin demand. It also suggests that monetary-policy transmission may extend into stablecoin markets because higher returns on bank balances raise the opportunity cost of holding tokens that generally pay little or no direct interest.

It does not establish the reverse proposition that stablecoin expansion has already weakened bank funding or credit supply. Federal Reserve researchers have emphasized that the outcome depends on reserve structure, domestic versus foreign demand, banks’ competitive responses and whether funds ultimately recycle through the banking system.

The study was presented at a Boston and New York Federal Reserve conference on May 22 and formally published in the BIS working-paper series on October 13. Coinburn is reporting the publication on October 14. The next useful evidence would be account-level or institution-level data that can identify who moved funds, where stablecoin reserves were placed and whether bank lending or funding costs changed afterward.