The central number in the White House case for restricting stablecoin yield is not large: $2.1 billion in additional U.S. bank lending. That is 0.02% of total bank loans under its baseline model. For community banks, the projected gain is about $500 million, or 0.026%.
Those estimates do not establish that stablecoin yield is irrelevant to banks. They do, however, make it difficult to present a ban as a meaningful near-term credit-expansion measure on the administration's own assumptions. The same analysis estimates an annual net welfare loss of roughly $800 million, with costs exceeding the modeled lending benefit by 6.6 times.
The dispute therefore turns on a narrower but consequential question: whether a model calibrated to the present stablecoin market and the Federal Reserve's current operating framework is a useful policy baseline, or one that will be overtaken by market growth and a changed reserve environment.
A $54.4 billion deposit shift yields only $2.1 billion in new loans
Roughly $54.4 billion would move from stablecoins into bank deposits under the White House model. The downstream figures are far smaller: the Council of Economic Advisers estimates $6.5 billion in marginal lending capacity and about $2.1 billion in additional loans after liquidity and reserve buffers.
That spread is the substance of the model, not a discrepancy. The CEA presentation distinguishes funds leaving stablecoins, deposits entering banks and new bank credit. Removing yield first induces a reallocation toward deposits; the model then estimates how much of that reallocation affects marginal capacity and actual lending. Most of the initial balance-sheet change is absorbed by buffers, so a deposit inflow does not automatically become a loan.
For an argument about deposit location, the $54.4 billion shift is the pertinent result. For an argument about expanding bank intermediation, it is the $2.1 billion loan figure. The White House’s September FAQ places that increase at 0.02% of all bank loans and 0.026% for community-bank loans. The FAQ figures show how little of the modeled deposit movement reaches lending.
The modeled costs exceed the lending benefit by 6.6 times
The administration's welfare calculation makes the case for a lending-driven prohibition still harder to sustain. It estimates that stablecoin holders would incur approximately $940 million in annual welfare losses, while the benefit from greater lending is put at $140 million. The resulting net effect is a negative $800 million a year.
In other words, the model does not treat the additional loans as a free gain. It counts the lost value to stablecoin holders alongside the benefit associated with extra credit, and finds that the former is substantially larger. The policy's principal modeled burden falls on the users whose returns or choices are constrained, not on banks.
This is not a forecast of every effect a yield rule could have, nor is it a comprehensive judgment on stablecoin regulation. It is a specific comparison within the White House framework. But it poses an awkward test for a policy defended primarily as a way to direct funds back into banks: the estimated welfare cost is 6.6 times the estimated lending benefit.
That comparison also changes how the $2.1 billion should be read. A modest positive loan number can be politically useful in isolation. Once it is set against the model's $940 million holder loss and $140 million lending benefit, it becomes evidence of a trade-off that is unfavorable under the baseline assumptions.
Ample reserves are the constraint on the lending case
The White House’s baseline estimates that banning stablecoin yield would increase lending by $2.1 billion, a result that depends heavily on the Federal Reserve’s current ample-reserves framework. Holding the other baseline assumptions constant, switching to scarce reserves raises the estimated lending effect to approximately $15 billion, or about 0.13% of loans, according to the White House FAQ.
That sensitivity identifies the operative constraint: the result reflects not only assumptions about stablecoin demand or household appetite for yield, but also how an added deposit dollar changes lending capacity under the prevailing monetary framework. The scarce-reserves estimate is more favorable to the lending case, though it remains small relative to the overall loan book.
The White House’s extreme stress case is much larger, at $531 billion in additional lending, or 4.4% of loans. It assumes stablecoins reach roughly 10% of deposits, reserves are held entirely in cash rather than Treasury bills, households are unusually sensitive to yield, and the system returns to scarce reserves. Those stacked assumptions make the result a scenario for what would have to change before the lending effect became macroeconomically large, not the White House’s central estimate under current conditions.
The fight is over whether today’s stablecoin market is the right denominator
The stablecoin market reached approximately $317 billion by April 6, 2026, more than 50% above its level in early 2025, according to a Federal Reserve analysis.
The White House calibration, however, is tied to a roughly $300 billion market and a 1.7% stablecoin share of bank deposits. Its baseline estimates a roughly $2.1 billion lending effect from a yield prohibition. Market growth makes that reference point a moving target, but does not by itself invalidate it.
The Independent Community Bankers of America disputes whether the reference point is appropriate for policy. Its June comment letter says the analysis rests on today’s still-immature market and cites a separate estimate that failing to extend the yield prohibition could eventually reduce community-bank lending by $850 billion through a $1.3 trillion reduction in deposits.
Those are different scenarios, not interchangeable forecasts. The ICBA estimate assumes a much larger eventual shift in deposit funding, while the White House extreme case itself requires stablecoins to reach roughly 10% of deposits and also assumes changes in reserve holdings, household yield sensitivity and the reserve framework. The unresolved issue is whether that longer-run market structure should replace the current-market denominator.
