The euro stablecoin gap

Executive summary

MiCA has been in force since 30 December 2024 and it worked as a licensing statute. ESMA's interim register of notified e-money token white papers carried 46 rows covering 20 distinct issuers at its 24 July 2026 update, and euro is the most common reference currency on it. The companion register of asset-referenced token issuers carried a header row and no data rows on the same date. Every fiat-referenced token authorised in the European Union has come through the e-money token route.

The market did not follow the licences. Piero Cipollone of the ECB Executive Board told an audience on 12 February 2026 that US dollar-denominated stablecoins "currently account for 99% of the global stablecoin market". The cleanest measure of the gap is one issuer against itself: Circle reported €381.8m of EURC in circulation as of 22 June 2026, against USDC of $72.6bn on RWA.xyz's 25 July 2026 reading. Same company, same five networks, one authorisation each.

The reasons are economic rather than procedural, and three of them are decisive. Float income on a euro reserve is lower than on a dollar reserve, both because euro short rates are lower and because the eligible euro reserve is a worse portfolio. MiCA prohibits paying interest on an e-money token, which removes the one lever an issuer has to buy adoption. And demand for a fiat-referenced token is derived demand for the unit that onchain markets quote in, which is the dollar. Where euro demand onchain does exist, it has gone to the wrapper that is allowed to pay a return: Spiko's euro T-bill money market fund held €886.24m on 25 July 2026, more than twice EURC's circulation.

Key findings

1. The euro segment is roughly half a per cent of its issuer's dollar book. EURC circulation of €381.8m on 22 June 2026 compares with USDC of $72.6bn on 25 July 2026, and with an aggregate stablecoin supply of $312.14bn on DefiLlama and $297.88bn on RWA.xyz, both read on 25 July 2026. The ECB's own framing on 12 February 2026 was that dollar tokens are 99% of the global market.

1. Zero asset-referenced token issuers is an equilibrium, not a backlog. ESMA's ART register held no data rows on 24 July 2026. A token referencing a single official currency is an e-money token by definition, which is the lighter regime and the one that matches what users want; the asset-referenced category exists for baskets and commodities, and nobody wants to be paid in a basket.

1. The euro reserve earns less and looks different. The ECB deposit facility rate was 2.25% with effect from 17 June 2026; the federal funds target range has been 3.50-3.75% since 11 December 2025. Circle's transparency disclosure, accessed 25 July 2026, shows USDC reserves in a government money market fund, overnight reverse repo and Treasuries maturing inside three months, while EURC reserves are disclosed as deposits at systemically important institutions and other bank deposits.

1. The euro instrument that grew is a fund, not a token. Spiko's own published figures on 25 July 2026 show EU T-Bills at €886.24m across 3,112 users at a 2.16% thirty-day yield, against US T-Bills at $159.59m across 815 users at 3.41%. The euro vehicle is more than five times the dollar vehicle at the same issuer, and it pays a return that a euro e-money token is forbidden to pay.

1. Euro onchain government exposure is concentrated in three instruments. On RWA.xyz's 25 July 2026 reading of non-US government debt, total $1.41bn across 24 assets and 10,008 holders, the euro-denominated products were Spiko's EU T-Bills at $1,023.3m, the NRW1 bond issued through Cashlink at $114.1m, and Backed's GOVIES 0-6 Months Euro at $9.8m.

Analysis

Start with the issuer's income statement, because a fiat-reserved token is a zero-coupon liability funded at nothing and invested in short government paper. Gross revenue is supply multiplied by reserve yield; the cost base is an authorisation, safeguarding arrangements, audit, distribution and redemption operations, and it is largely fixed. With the deposit facility rate at 2.25% and the federal funds range at 3.50-3.75%, a euro token earns roughly a third less per unit of float than a dollar token before any allowance for portfolio composition.

Composition makes it worse. MiCA requires an e-money token issuer to hold at least 30% of funds received as deposits with credit institutions, with the remainder invested in the categories that apply to asset-referenced tokens; the CEPS analysis of September 2025 sets out the mechanics of Article 54. Bank deposits pay below the deposit facility rate, because the bank keeps a spread, and they carry bank credit risk that a bill does not. Circle's own reserve disclosure shows the consequence plainly: the dollar book holds a government money market fund and reverse repo, the euro book holds bank deposits. There is also no single euro-area bill market comparable to the US Treasury bill market; euro sovereign paper is issued by twenty governments with different credit and different liquidity, so the euro reserve manager is choosing among national curves rather than buying one instrument.

Put the two together and the break-even supply for a euro token is several times the break-even supply for a dollar token at the same cost base. That is the whole of the supply-side story, and it explains why the euro register is full of small licensed issuers rather than one large one: the licence is obtainable, the business is not, below a scale that nobody has reached.

The demand side is where MiCA's design is binding. Article 50 prohibits issuers of e-money tokens from granting interest. In dollar markets the standard route to distribution is to share float income with the venues and wallets that hold balances, directly or through rebates; a euro issuer cannot do that with the token, and in any case has less to share. So the euro token must win on utility alone, against euro bank deposits that already move through SEPA at low cost within the currency area. A dollar stablecoin offers a non-US holder something a local deposit cannot: dollar exposure and access to dollar payments without a US bank. A euro stablecoin offers a euro-area holder nothing the holder does not already have, and offers a non-euro-area holder a currency they were not seeking.

This is why the ECB's framing matters more as a sovereignty argument than as a market forecast. Cipollone's warning on 12 February 2026 was that dollar tokens "could also start to gain a foothold in Europe, starting with retail cross-border payments" and in "fringe use cases such as gaming, micro-payments and machine-to-machine payments", and that European banks "could lose fees, data and deposits to stablecoins". The policy response is not a bigger euro stablecoin. It is the digital euro, and on 19 June 2026 Cipollone set out the timetable: if the Regulation is adopted by the end of 2026, a pilot in 2027 and the first digital euro "in the course of 2029". A central bank liability that pays no interest is not competing with a private token that cannot pay interest either; it is competing on legal tender status and reach.

Note the asymmetry that the data exposes. MiCA already restrains the substitute: it sets quantitative limits on the use of non-euro-denominated tokens as a means of exchange, which the CEPS paper describes as designed to protect the euro's domestic payment role. Dollar tokens are therefore not free to displace euro tokens in EU payments, and euro tokens still did not grow. That rules out competitive displacement as the explanation and leaves the two internal ones: the float economics do not work at current rates, and the product is not permitted to compensate holders for using it.

The Spiko comparison is the strongest single piece of evidence for the second point. The same issuer, the same networks, two UCITS money market funds, and the euro fund is more than five times the dollar fund despite yielding 125 basis points less. Euro demand for an onchain cash instrument exists. It went to the wrapper that pays.

Methodology and data

Licensing counts are taken from ESMA's two interim MiCA registers, downloaded on 25 July 2026. The e-money token file's stated last update was 24 July 2026, with 46 data rows corresponding to 20 distinct issuers; because several issuers have notified more than one white paper, the row count runs ahead of the issuer count, and one third-party mirror of the same file counts 21 issuers rather than 20. The asset-referenced token file contained its fifteen column headers and no data rows.

Supply figures are on a stated-as-of basis and are not contemporaneous with each other. EURC circulation of €381.8m is Circle's own figure as of 22 June 2026 from its EURC page. USDC of $72.6bn is RWA.xyz's 25 July 2026 reading; Circle's monthly examination reports, which are the audited-adjacent series, put USDC at $75.89bn at the 29 May 2026 measurement date. Aggregate supply is quoted from two trackers on the same day because they disagree by about $14bn on inclusion rules.

Policy rates are from the ECB's key interest rates table, top row 17 June 2026, and from the Federal Reserve's record of open market operations, most recent change 11 December 2025. Fund figures are the issuer's published data, accessed 25 July 2026, and are cross-checked against RWA.xyz's non-US government debt page for the same day. Reserve composition is as disclosed on Circle's transparency page, whose reserve data was marked current as of 23 July 2026.

Limitations

There is no official series for euro stablecoin supply. Neither the ECB nor ESMA publishes outstanding amounts by reference currency, so this piece is built from one issuer's self-reported circulation plus third-party trackers, and the EURC figure is the only euro number in it with the issuer's name on it. Secondary reporting has put the whole euro segment around $900m in mid-2026 with EURC at roughly half of it; that estimate is consistent with the figures used here, but it is not a measurement and this piece does not rely on it.

The 99% figure is a central banker's characterisation in a speech, not a statistic with a published methodology. It is directionally supported by the trackers, and the precise share depends on whether yield-bearing and synthetic dollar tokens are counted, which the trackers treat differently.

Two of the causal claims are not directly testable with the available data. The break-even argument requires issuer cost structures, and no MiCA e-money token issuer publishes revenue or cost of a euro programme, so the claim that the euro business does not clear its fixed costs is an inference from reserve yields and disclosed reserve composition rather than from accounts. The interest-prohibition argument rests on a comparison between two different legal wrappers at one issuer; Spiko's funds and Circle's token differ in eligibility, redemption terms and investor onboarding as well as in whether they may pay a return, and those differences are not separable in the data.

Finally, dates do not line up. The EURC figure is 22 June 2026, the USDC and tracker figures are 25 July 2026, and the ESMA registers are 24 July 2026. Ratios computed across them are approximate, and none of the conclusions here would change if the euro figure were twice what is reported.