When Congress passed the GENIUS Act in July 2025, the debate was framed as consumer protection and payments policy. The bigger effect may be on the plumbing of two markets that rarely share a sentence: the roughly $6 trillion Treasury bill market and the deposit base of U.S. community banks. As of October 5, 2026, neither effect has arrived at scale. The forecasts on both sides rest on assumptions that are worth checking one at a time.

What the statute makes issuers hold, and what it doesn't

According to the Council of Economic Advisers' April 2026 analysis of the Act, permitted issuers must back stablecoins at least one-to-one. The reserve menu is cash, demand deposits at insured institutions, Treasury bills with 93 days or less to maturity, Treasury-backed repurchase agreements, government money market funds, and deposits at the Federal Reserve. Issuers may not pay holders interest or yield. The Act is a menu, not a Treasury bill mandate. Where demand ends up depends on issuer economics and on how regulators write the details.

Current practice leans heavily toward bills. The CEA, citing Circle's December 2025 reserve report, notes that about 12% of USDC reserves sat in bank deposits and 88% in bills and repo. It also reports that Tether held only $34 million in bank deposits against $147.2 billion in reserves at the end of 2025. A Federal Reserve Board staff note by Jessie Jiaxu Wang (December 17, 2025) puts Circle's deposit share at about 13% and Tether's near zero. That split matters later, because a bill-heavy reserve mix and a deposit-heavy one affect banks very differently.

Where the rulemaking stands

The statutory deadline for implementing rules was July 18, 2026. By my reading of agency documents and secondary legal summaries, the major rules are still proposals. The OCC's proposal appeared in the Federal Register on March 2, 2026. It would require full segregation of reserves and would require issuers to maintain and demonstrate the operational capacity to monetize every type of reserve they hold. Its Question 77 asks whether to limit Treasury notes and bonds that qualify under the 93-day rule, because they may be more thinly traded and could sell at a discount in a stress.

The FDIC approved its own proposal on April 7, 2026. It generally requires redemption within two business days, and it says deposits held as reserves are insured to the issuer, not on a pass-through basis to token holders. Treasury proposed a rule on certifying state regimes as substantially similar on April 3, and one account of the Federal Reserve's announcement says the Board issued two proposals on September 24, 2026. I found no final rule from the OCC, FDIC or Treasury. Under the statute's own timetable, the Act takes effect on the earlier of January 18, 2027 or 120 days after final regulations, so the bank-side details may still be unsettled when the market is already live.

Sizing bill demand: three scenarios, one honest caveat

Today the footprint is modest. A Bank for International Settlements paper by Rashad Ahmed and Iñaki Aldasoro, revised in June 2026, reports that stablecoin assets under management exceeded $270 billion in December 2025. Issuers bought nearly $35 billion of T-bills in 2025, about what the largest U.S. government money market funds did and more than most foreign buyers. A Kansas City Fed Economic Bulletin by economist Stefan Jacewitz makes the proportions concrete: Circle held about $20 billion in bills, roughly 43% of its assets, and if every issuer did the same the industry would hold around $125 billion, under 2% of outstanding bills.

The big forecasts are about 2028. The Treasury Borrowing Advisory Committee's April 30, 2025 “Digital Money” presentation described stablecoins as a source of “materially heightened demand” for bills. As reported, it estimated about $120 billion of bills backing stablecoins then and more than $1 trillion of additional bill demand if the market reached $2 trillion by 2028. Standard Chartered's analysts reached a similar $0.8 to $1.0 trillion in February 2026, while the Kansas City Fed bulletin notes J.P. Morgan's much lower $500 billion projection for 2028.

To keep the numbers anchored in cited inputs, I applied two reserve ratios to the growth implied by each projection, starting from a base of roughly $300 billion (the CEA's February 2026 figure). The low ratio is the Kansas City Fed's roughly $0.50 of Treasuries per $1 of stablecoins. The high ratio is Circle's 88% in bills and repo. The net column applies the Kansas City Fed's offset for funds drawn from bank deposits. The bulletin states $0.30 per $1, but its own inputs ($0.50 minus $0.08) give $0.42, so I show both. This is my arithmetic on their assumptions, not a forecast from any of these institutions.

The $2 trillion path looks strained against recent behavior. Standard Chartered, as reported by CoinDesk, noted supply had stalled just above $300 billion after rising from $238 billion in April 2025, about $60 billion in ten months. Reaching $2 trillion by the end of 2028 from about $304 billion in early September 2026 (the DeFiLlama-based tally some trackers publish) means adding around $63 billion a month for 27 months. That is roughly ten times the recent pace. So the lower two rows are the more defensible planning range, and the top row should be read as a ceiling.

Why gross demand is not net demand

The Kansas City Fed bulletin makes the central point in one sentence: “Funds flowing into stablecoins have to flow out of another source.” If the money comes from households selling Treasuries, or from money market funds that already own bills, the net addition is small or zero. The bulletin says that if the sources of funds sell Treasuries as fast as issuers buy them, a larger stablecoin market has no net effect on Treasury demand. Wang's note, citing Aldasoro and co-authors (Economics Letters, 2025), adds that U.S. monetary-policy shocks push money into prime money funds and out of stablecoins. That points to a substitution margin between the two products.

Foreign demand could push the other way. The CEA, citing Brookings, says over 80% of stablecoin transactions occur outside the United States, and it cites the IMF's July 2025 External Sector Report finding that issuers already hold more bills than Saudi Arabia. Dollars that arrive from abroad do not displace a U.S. depositor. That makes them the cleanest source of net new bill demand, though no cited source quantifies it.

The price effect: small today, uneven later

Ahmed and Aldasoro find that a $3.5 billion inflow lowers three-month bill yields by 0.71 basis points on impact and about 4 basis points within 10 days, with no spillover to longer maturities. The effect is stronger when Treasury-market intermediaries are stressed and has strengthened as the sector grew. Earlier drafts of the same paper reported a smaller range of roughly 2 to 3.5 basis points, a reminder that these point estimates move between versions. BIS Bulletin No. 108 (July 2025) adds a warning about asymmetry. It reports that outflows raise yields two to three times as much as inflows lower them, consistent with issuers having to liquidate holdings quickly.

Put the proposed rules next to that finding. A two-business-day redemption standard and an OCC requirement to monetize every reserve type quickly both push issuers toward the shortest, most liquid assets. That is good for run resistance, but it concentrates any forced selling in the same narrow segment of the bill market. This is my inference, not a finding in the proposals, but it explains why the OCC asked about longer-dated notes at all.

Community banks: two readings of one balance sheet

The pessimistic reading comes from Wang's note, which says stablecoins can “reduce, recycle, or restructure bank deposits.” Using deposit-to-loan pass-through multipliers of 0.6 to 1.26, she estimates loan contractions of $65 to $141 billion if stablecoins grow by $200 billion with half of reserves recycled into banks, $190 to $408 billion at $500 billion, and $600 billion to $1.26 trillion at $1 trillion if issuers earn interest at the Fed. The Kansas City Fed's simpler version has a $650 billion shift out of bank deposits cutting loans by about $325 billion, or 2.5%. She flags relationship lenders in digitally young markets as the most exposed. She also notes that wholesale deposits can be two to three times as volatile as retail deposits, citing an OFR brief on liquidity coverage ratios.

The CEA's April 2026 analysis argues the effects are far smaller, and it deserves a fair hearing. It treats this as a general-equilibrium problem, in which the bank that gains a reshuffled deposit can lend it. Only reserves held as bank deposits at 100% backing are locked out of lending, and for Circle that is about 12%. Its model asks what happens to lending if yield is prohibited, which bounds the loss from adoption in the opposite direction. The baseline answer is $2.1 billion of added lending, 0.02% of loans, and $500 million at community banks. Even stacking worst-case assumptions, including that the Fed abandons ample reserves, it gets $531 billion overall and $129 billion at community banks. It also cites research by Charles River Associates and economist Tsyrennikov finding no statistically significant link between USDC market cap and community bank deposit changes, and observes that issuer reserves are custodied at large institutions such as BNY Mellon.

These camps are not fully contradictory. A Federal Reserve Bank of New York staff report by Michael Junho Lee and Donny Tou (No. 1185) finds that banks partnering with stablecoin issuers hold more reserves and lend relatively less. The CEA calls that mechanism isomorphic to its own. The cost therefore falls on large custody banks that process stablecoin flows, not on Main Street lenders, unless reserves become scarce or deposit flows reach small banks directly. The CEA also acknowledges the yield loophole: the Act bars issuers from paying yield but not intermediaries, and Coinbase's USDC rewards, funded through revenue sharing with Circle, offered returns similar to high-yield savings accounts as of February 2026. If that loophole is the real channel for deposit competition, the exposure sits wherever yield-chasing depositors are, which a reserve-composition rule cannot address.

The FDIC's no-pass-through proposal pulls both ways. Issuer reserve deposits would sit at banks as uninsured corporate balances, which Wang describes as flighty funding. But that outcome depends on which banks hold them, and today that is mostly big ones.

Signals worth tracking through 2027

Four things will settle this better than any projection. First, whether final rules keep bills and repo as the practical reserve core and how the OCC resolves Question 77. Second, whether yield-sharing arrangements are closed by legislation or left alone. Third, the Fed's reserve regime, since the CEA's lending estimates change by orders of magnitude if reserves become scarce. Fourth, the monthly reserve disclosures the proposals would require, which will show how much of each new stablecoin dollar actually lands in bills, bank deposits or money funds. Until those arrive, the defensible statement is narrow: stablecoins are already a visible bill buyer, their net contribution is probably a fraction of the headline numbers, and the community bank threat is possible but unproven.