Seigniorage and float income

Why it matters

A fiat-referenced token is a liability that pays its holder nothing, funded at zero and invested at the front end of the government curve. That single asymmetry is the business. Everything else about a large dollar-token issuer (the distribution deals, the licensing strategy, the choice of reserve manager) is downstream of the question of who ends up with the interest that the holder is not receiving.

The answer is not obvious from the outside, and it is not the issuer. The margin is real, but the bargaining power sits with whoever controls the wallet in which balances rest: one unit of a token is identical to every other, so the only thing an issuer can compete for is where those units are held.

How it works

The mechanics are three steps. A distributor wires fiat and receives newly minted tokens at par. The issuer invests the fiat in the permitted reserve assets and keeps the return. When tokens come back, they are burned and the fiat is returned at par, and the income stops. Float income is therefore the product of average balances outstanding and the yield on a short-dated portfolio, less whatever the issuer must pay to keep those balances in existence.

Holders receive nothing by law rather than by choice. The US statute of 18 July 2025 prohibits a permitted issuer from paying a holder any form of interest or yield solely in connection with holding, using or retaining the token; MiCA imposes the same prohibition on e-money token issuers and extends it to service providers. So the split of the carry is not negotiated with holders at all. It is negotiated with distributors.

Designs that sit outside the payment-stablecoin definition do the opposite, which is the useful control group. Sky reported a governance-set Sky Savings Rate of 3.52% on sUSDS in July 2026, paying the return away rather than retaining it. Ethena's sUSDe paid about 4% annualised in early 2026, sourced from perpetual funding and basis rather than from a reserve, which is why USDe falls outside the statutory definition and its yield prohibition.

Economic mechanism

Follow the money in four steps.

The source is the Treasury bill and repo market, which pays the coupon on the reserve. The first claimant is the issuer. In the quarter to 31 March 2026, Circle reported total revenue and reserve income of $694m, of which reserve income was $653m, up 17% on the year while average USDC in circulation grew 39%. The gap between those two growth rates is the policy rate, and it is the single most important sensitivity in the model: the volume of the float grew more than twice as fast as the income it produced.

The second claimant is distribution. Circle reported distribution, transaction and other costs of $407m for the same quarter, against that $653m of reserve income, a little under two-thirds of the gross carry. Net income from continuing operations was $55m and adjusted EBITDA was $151m. Circle's registration statement disclosed that Coinbase receives 50% of residual reserve income, and Coinbase held roughly 20% of USDC in circulation in 2024, against about 5% in 2022. The rent accrues to the balance-holder, not the reserve-holder.

That outcome follows from the product. Because the yield prohibition makes it illegal to compete for holders on price, and because the token is fungible across venues, the issuer's only lever is to pay intermediaries to keep balances resident. Payment for balance placement is the equilibrium, and its share rises with the concentration of wallets. A distributor holding a fifth of the float and half the residual income is a counterparty in a bilateral negotiation, not a channel.

The third claimant is the reserve manager. Most of the USDC reserve sits in a government money market fund managed by BlackRock, which earns a fee for running it. The fourth is the fisc, indirectly: the Treasury Borrowing Advisory Committee's presentation of 30 April 2025 noted that issuers already held more than $120bn of Treasury bills and put implied incremental bill demand at roughly $900bn if the sector reached about $2tn by 2028.

Who pays, finally, is the depositor who would otherwise have earned the rate, and the bank that would otherwise have intermediated the balance.

Participants

Issuers hold the reserve and the redemption obligation. Exchanges, wallets and brokers hold the balances and extract most of the margin. Asset managers run the portfolio for a fee. Banks lose a transactional deposit and the lending capacity attached to it. Legislators set the perimeter of the yield prohibition, and so decide how much carry can legally reach a holder indirectly.

Examples

Circle's first quarter of 2026, reported on 11 May 2026, is the only audited-quality window into the split, and it shows a large gross margin and a thin net one: $653m of reserve income, $407m of distribution and transaction costs, $55m of net income from continuing operations.

Tether shows the retained-earnings case. It reported quarterly profit of approximately $1.04bn for the first quarter of 2026 alongside $8.23bn of reserves in excess of liabilities, against consolidated token liabilities of $183.54bn. That is a materially higher retention of the carry, on a book with no equivalent distribution arrangement disclosed.

Sky is the deliberate counter-design: a 3.52% savings rate in July 2026 on a token that is not a payment stablecoin, and therefore an instrument that competes for balances on price rather than by paying intermediaries.

Risks and limitations

The perimeter of the yield prohibition is the live dispute and it is worth stating both sides accurately. The Bank Policy Institute argued on 12 August 2025 that the prohibition binds only issuers, that affiliates and exchanges can pay holders indirectly through arrangements jointly marketed with the issuer, and that deposit flight would reduce lending capacity and raise borrowing costs. Its brief cites a Treasury figure of as much as $6.6tn of deposit outflows. The underlying Treasury presentation of 30 April 2025 is more careful than the way it is usually quoted: $6.6tn is the size of the M1 transactional pool, while roughly $0.9tn of non-demand transactional deposits is identified as at risk of migration, with the impact likely modest under non-interest-bearing designs. The same document supports a large number and a small one depending on the assumption about yield, which is exactly why the perimeter is being fought over.

The OCC's proposal of 25 February 2026 would move that perimeter. It would establish a rebuttable presumption against affiliate arrangements designed to replicate the economics of yield, and reserve case-by-case authority to judge arrangements on economic substance rather than formal separation, reaching balance-based rewards, rebates and profit-sharing. If finalised in that form, the largest single cost line in an issuer's accounts becomes a regulatory question rather than a commercial one.

Rate risk is the structural limitation. The whole revenue line is a position in the front end, at a weighted average maturity supervisors want under three weeks. A 100 basis point cut passes through within a quarter, while distribution contracts struck as fixed payments rather than as a share of income do not fall with it.

The word seigniorage also overstates the analogy, and the BIS Annual Economic Report 2026 is the reference for why. Central bank seigniorage is the profit on issuing non-interest-bearing base money, and it is remitted to the fiscal authority. Private float income on a token that the BIS argues fails the test of singleness of money is retained by the issuer and its distributors. The mechanism rhymes; the beneficiary does not.

Key metrics

The ratio worth tracking is distribution and transaction costs to reserve income, 62% for Circle in the March 2026 quarter. Alongside it: reserve income per dollar of average balances, which isolates the rate from the volume; the share of float held by the single largest distributor; retained profit and excess reserves for issuers that do not report a distribution line; and the level of supply itself, since the whole revenue line scales with it.

Related research

Circle's quarterly disclosures are the only place the split between issuance and distribution is published in dollars. The Treasury Borrowing Advisory Committee's April 2025 presentation is the source most often cited on both sides of the deposit-migration argument and repays reading directly. Chapter III of the BIS Annual Economic Report 2026 sets out why float income on a private token is not the public revenue the term implies.

The data on this site

The series below are the ones that bear directly on this concept. Each carries its own source, cadence and coverage.