Why it matters
A token's price on an exchange is a market outcome. Par is a claim on an issuer, and the two are joined only by the channel that runs between them. Whether a dollar token behaves like money depends far less on what it trades for on a given afternoon than on who may present it to the issuer, at what cost, and how soon the issuer must pay.
That channel is deliberately narrow. Circle states that Circle Mint is for registered distributors (exchanges, institutional traders, wallet providers, banks and consumer-app companies) and is not available to individuals. Tether sets a minimum acquisition or redemption amount of $100,000. Almost every holder of a large dollar token therefore has no redemption right they can exercise themselves; they have a price at which someone who does hold that right will buy from them. The distinction is the whole of the subject.
How it works
Primary redemption is a bilateral transaction between the issuer and an onboarded counterparty: tokens are sent to the issuer and burned, and fiat is paid out through a bank. Secondary trading is everything else. A regulated regime works by putting terms on the first of those, on the theory that the second will follow.
The US statute enacted on 18 July 2025 requires a permitted issuer to maintain clear procedures for timely redemption, to disclose all fees in plain language, and to give at least seven days' notice before changing them. It does not itself define "timely". The Office of the Comptroller of the Currency's notice of proposed rulemaking, issued on 25 February 2026 under Bulletin 2026-3, supplies a number: redemption no later than two business days after a valid request. It also supplies the exception. Where an issuer faces redemption demands exceeding 10% of its outstanding issuance value within a single 24-hour period, the window extends to seven calendar days, with notice to the OCC inside 24 hours.
Price terms differ sharply between the two largest issuers. Circle charges nothing to mint or redeem and rations by admission; it reports more than 333bn USDC minted through the service. Tether charges the greater of $1,000 or 0.1% on redemption, 0.1% on acquisition, and a non-refundable verification fee of $150 in tokens. It rations by price and by ticket size.
Economic mechanism
The structure is the one used for exchange-traded funds, and it behaves the same way. A small set of counterparties with primary access arbitrages the gap between the secondary price and the redemption value, and the width of the band they cannot profitably close is set by their costs.
Tether's schedule makes the arithmetic explicit. At the $100,000 minimum the fee floor of $1,000 is 100 basis points. At $1m it is 0.1%, or 10 basis points, and it stays at 10 basis points above that. So a discount of 20 basis points is not worth arbitraging in small size and is worth arbitraging in large size, which is why persistent shallow deviations are a statement about ticket economics rather than about backing. Circle's zero fee narrows the band to the cost of funding and the operational risk of the transfer, and pushes the binding constraint onto who is allowed through the door.
The redemption window is the second cost. Two business days of exposure between paying the secondary price and receiving fiat has to be funded and hedged; the issuer keeps the reserve income over those two days and the redeemer wears the risk. Lengthen the window and the arbitrage band widens mechanically.
Immediacy is bought with foregone carry, and regulators have started to price it directly. The Bank of England's policy statement and draft Code of Practice of 22 June 2026 require a systemic sterling issuer to hold at least 30% of backing as deposits at the Bank, precisely so that redemptions can be met promptly, and cap the interest-bearing remainder at 70% of short-term UK government debt. Unremunerated central bank money earns the issuer nothing. The cost of a credible par promise appears on the issuer's income statement, and the benefit accrues to holders who will never call the issuer.
Participants
Issuers hold the obligation and the reserve. Registered distributors and authorised redeemers hold the right and, in practice, the arbitrage. Exchanges are the redemption venue for everyone else, which makes their balance sheets and their willingness to quote a two-way price a load-bearing part of the arrangement. Banks supply the fiat leg and set the hours during which par can actually be paid. Supervisors set the window, the fee disclosure and the conditions under which the promise may be suspended.
Examples
Circle Mint and Tether's fee schedule are the two live templates: free access for a vetted list, versus priced access with a six-figure floor. Neither is available to a retail holder, and both are consistent with a token that trades within a basis point of a dollar most of the time.
Ethena's USDe on 11 October 2025 is the cleanest demonstration that the two prices are separate objects. The token printed 65 cents on one venue during a liquidation cascade while, on the issuer's account, mint and redeem functionality remained operational with no downtime and the book remained over-collateralised. Nothing about par failed; the venue price did.
Risks and limitations
What holders own in an insolvency is genuinely contested, and the disagreement is not about the words of the statute but about how bankruptcy practice will meet them. Morgan Lewis, writing on 2 July 2025, read the insolvency provisions as treating holders more favourably than secured creditors, with reserves excluded from the estate and shortfall claims ranking ahead even of administrative expenses, and warned that protection that thick could obstruct a workable reorganisation. Adam Levitin, writing on 2 December 2025, argued the priority is largely illusory: secured repo and margin lenders, debtor-in-possession financing, professional fee carve-outs and setoff rights all reach the reserves first, leaving holders fifth in line, and the distribution clock only starts after a hearing that a DIP lender has every incentive to delay. Both readings are defensible on the text. Neither has been tested.
The stress exception is the second limitation, and it is not a loophole but a deliberate gate. A demand liability that converts to a seven-day term liability once redemptions pass 10% of outstanding in a day is a demand liability only in normal conditions. The cost falls on whoever is last in the queue, which is an incentive to be early, the structure the window was meant to suppress.
Access asymmetry is the third. A regime that governs the primary channel governs the price only through intermediaries who may withdraw. If market makers stop quoting, the redemption right is intact and unreachable, and the holder's realisable value is whatever the order book offers.
Key metrics
The informative series is net redemption over short windows rather than the level of supply. Tether's consolidated token liabilities fell about $3.0bn in the quarter to 31 March 2026, the first quarter-on-quarter decline in that series; USDC in circulation was $75.89bn at 29 May 2026 on the most recent examined figure available on 25 July 2026. Alongside those, track the redemption fee at the minimum ticket and at scale, the contractual and regulatory redemption window in business days, the share of backing held in unremunerated central bank deposits, and the number of counterparties with primary access.
Related research
The OCC's February 2026 proposal is the most specific published attempt to put a number on timeliness, and A&O Shearman's note on it sets out the two-business-day standard and the stress extension. The Morgan Lewis note and Adam Levitin's post together frame the insolvency question better than either does alone. The Bank of England's June 2026 policy statement is the clearest official statement of the trade between unremunerated backing and prompt payment.
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.
View as table
| Period | Stablecoin supply, USDC | Note |
|---|---|---|
| 2024-12-09 | $40,596,908,123 | USDC in circulation at the intra-month measurement date used by the examination report published 30 January 2025. |
| 2024-12-31 | $43,856,799,846 | |
| 2025-03-19 | $59,082,692,603 | |
| 2025-03-31 | $59,975,771,715 | |
| 2025-06-12 | $60,935,344,661 | |
| 2025-06-30 | $61,332,530,506 | |
| 2025-09-18 | $74,225,102,864 | |
| 2025-09-30 | $73,748,934,193 | |
| 2025-12-26 | $76,404,636,239 | |
| 2025-12-31 | $75,266,033,823 | |
| 2026-01-16 | $75,898,717,670 | |
| 2026-01-30 | $69,954,717,486 | A drop of about $5.9bn from the measurement date two weeks earlier, the largest fall in the series. |
| 2026-03-11 | $78,681,023,356 | |
| 2026-03-31 | $77,049,290,538 | |
| 2026-04-06 | $77,745,231,766 | |
| 2026-04-30 | $77,047,590,794 | |
| 2026-05-05 | $77,963,911,020 | |
| 2026-05-29 | $75,885,403,148 | Latest month-end figure in the most recent examination report available on 25 July 2026. |
View as table
| Period | Stablecoin supply, Tether | Note |
|---|---|---|
| 2024-12-31 | $136,613,782,874 | Liabilities relating to issued digital tokens, per the BDO attestation for the quarter. |
| 2025-03-31 | $143,678,070,758 | |
| 2025-06-30 | $157,100,255,857 | |
| 2025-09-30 | $174,356,634,812 | |
| 2025-12-31 | $186,450,610,920 | |
| 2026-03-31 | $183,438,487,810 | First quarter-on-quarter decline in the series, of about $3.0bn. |