The stablecoin margin war: why Circle's real competitor was never Tether, 2026 edition

Every explanation of Circle's business starts with the same two facts: it issues USD Coin, and USD Coin is backed one for one by cash and short-term Treasuries. Both are true and both miss the point. Circle is not really in the business of issuing a token. It is in the business of managing a very large, very liquid bond portfolio, and then splitting the interest on that portfolio with whichever partner brought in the deposits. The token is the packaging. The interest is the product. Understanding that single fact explains why Circle's stock has moved more on distribution-partner headlines this year than on anything resembling stablecoin technology, and why a coalition that includes Visa, Mastercard and Stripe just built an entirely different company to fight over exactly the same interest income.

What Circle actually sells

USDC in circulation reached $77.0 billion at the end of Circle's first quarter of fiscal 2026, up 28 percent year over year. Circle's total revenue and reserve income for that quarter was $694.1 million, of which $652.5 million, 94 percent, came from reserve income: interest earned on the cash and Treasury bills backing that $77 billion, at a reserve return rate of 3.5 percent. Subscription, services and transaction revenue, the part of the business that actually resembles a technology company, contributed just $41.6 million.

The reserves themselves sit overwhelmingly in short-dated US government paper managed through a BlackRock-run fund and custodied at BNY, with monthly attestations published by an independent accounting firm confirming the cage holds at least a dollar of qualifying reserves for every token in circulation. That structure was not always in place. When Silicon Valley Bank collapsed in March 2023, Circle disclosed that $3.3 billion of USDC reserves, about 8 percent of the total, sat in accounts at the failed bank. USDC broke its peg over that weekend, trading as low as roughly 87 cents, until federal regulators guaranteed SVB's deposits and arbitrage traders restored the dollar peg within days. The industry's lasting lesson from that episode, and from the collapse of the algorithmic stablecoin TerraUSD the year before, was that a stablecoin's entire value proposition rests on the boring, verifiable quality of what backs it, not on any clever mechanism.

The other half of the ledger: what Coinbase actually takes

Here is the part of Circle's economics that gets less attention than it deserves. Circle does not keep its reserve income. Under a revenue-sharing agreement dating to Circle and Coinbase's 2023 restructuring of the original Centre Consortium, Coinbase receives 100 percent of the interest earned on USDC held directly on Coinbase's own platform, and 50 percent of the interest earned on USDC held everywhere else, on other exchanges, in DeFi protocols, in third-party wallets. Because Coinbase's own platform holds only about 20 to 22 percent of total USDC supply, the arithmetic of that second clause does most of the work: in 2024, Coinbase received $908 million from Circle, roughly 54 percent of Circle's total revenue that year, capturing more than half of Circle's economics from a direct holding of about a fifth of the float. Circle cannot unilaterally exit this agreement, which renews on a rolling basis with its next major review window in August 2026.

Run the same math forward and the picture for any given quarter is stark. Circle's Q1 2026 distribution, transaction and other costs totaled $407 million, up 17 percent year over year, the great majority of which reflects the Coinbase split. Adjusted EBITDA for the quarter was $151 million, and net income from continuing operations was $55 million, down 15 percent year over year even as USDC circulation grew 28 percent, because the reserve return rate fell 66 basis points as short-term rates declined. Full-year 2025 tells an even sharper version of the same story: $2.75 billion in total revenue and reserve income, and a net loss of $70 million, driven substantially by $845 million in stock-based compensation triggered by Circle's June 2025 IPO. A company earning billions in interest income can still report negative net income once distribution costs and IPO-related compensation are subtracted, which is exactly what happened.

Two variables, one structural and one political

Circle's future income depends on two things that move independently of each other, and separating them is the key to understanding every headline the stock generates. The first is the Federal Reserve. Reserve income is simply USDC circulation multiplied by the prevailing short-term yield, so every rate decision flows straight through to the top line with no lag and no room for Circle to offset it. Circle's own disclosures have quantified the sensitivity: a full percentage point cut in rates removes roughly hundreds of millions of dollars in annual reserve income, a mechanism with no analog to, say, a payments network whose revenue depends on transaction volume rather than a policy rate set by twelve people in Washington eight times a year.

The second variable is the distribution split itself, and this one is not set by the Fed, it is set by negotiation. The share of reserve income Circle keeps after paying Coinbase, its Retention Ratio, barely moves when rates fall. What it does move on is the terms of the underlying partnership agreement, which is precisely why the August 2026 renewal window matters more to Circle's long-term economics than any single Federal Open Market Committee meeting. A rate cut shrinks the whole pie. A worse distribution deal shrinks Circle's slice of whatever pie remains, and those are two entirely different risks that happen to show up in the same revenue line.

The consortium built to end the moat entirely

If the Coinbase agreement is a negotiated split of one issuer's economics, what happened on June 30, 2026 is a wholesale rejection of the issuer model itself. More than 140 companies, including Visa, Mastercard, Stripe, BlackRock, BNY, Standard Chartered, Google and, notably, Coinbase, announced Open USD, a stablecoin governed by an independent entity called Open Standard and led by Zach Abrams, the former co-founder of Bridge, the stablecoin infrastructure firm Stripe acquired in 2024. Open USD charges no minting or redemption fees, imposes no volume caps, and is structured to return nearly all reserve income to its member partners after a management fee, rather than retaining it for a single issuing company. Circle's stock fell as much as 17.5 percent on the day of the announcement, closing down roughly 15 to 17 percent depending on the session measured, and extended a monthly decline of close to 39 percent.

The structural logic is the inverse of Circle's business model. Where Circle keeps the float and negotiates away a share of it to distributors like Coinbase, Open USD is designed so that the distributors themselves are the shareholders of the money, collectively capturing the interest that would otherwise accrue to a single issuer. It is, in effect, an attempt by the largest payment and technology companies in the world to disintermediate the entire category of company Circle represents.

Sixteen days later, on July 16, 2026, Visa turned that announcement into shippable infrastructure. The Visa Stablecoin Platform launched in beta, giving banks, fintechs and crypto-native firms a single Visa-managed environment to mint, redeem, hold and transfer stablecoins, plugged directly into Visa's existing settlement, treasury and fraud systems reaching roughly 15,000 financial institutions and 200 million merchants worldwide. Tellingly, the platform launches with Open USD first, but also supports Circle's USDC and Paxos's USDG. That coin list is the real signal: Visa does not need Open USD to beat USDC. If USDC keeps winning, Visa still earns platform economics from every transaction that flows through its rails. If Open USD wins, Visa earns a partner's share of the float directly. Visa built a structure where it profits regardless of which stablecoin ultimately dominates, a position no single-issuer stablecoin company can replicate.

Every bank wants a piece of the same interest income

Open USD is the most coordinated challenge to Circle's model, but it is not the only one. JPMorgan's blockchain unit, Kinexys, has already launched JPM Coin under the ticker JPMD, a dollar-denominated deposit token for institutional clients on Base, the Coinbase-incubated Ethereum layer two network. Because JPMD represents a bank deposit rather than a stablecoin reserve, JPMorgan can legally pay interest directly to holders in a way Circle cannot offer on USDC itself, an advantage that goes to the heart of why traditional banks view tokenized deposits, not stablecoins, as their preferred entry point into the same market. PayPal's PYUSD, Ripple's RLUSD, issued through Standard Custody and Trust under a New York trust charter, and Paxos's USDG, backed by a consortium that includes Mastercard, Robinhood and Kraken, round out a field in which nearly every major payments and banking incumbent now has a competing dollar token.

Tether remains the largest single stablecoin by a wide margin, with USDT circulation of roughly $184 billion as of late July 2026, close to 60 percent of the total stablecoin market, against USDC's approximately $73 to 77 billion, and Tether posted a $1.04 billion profit in the first quarter of 2026 alone while keeping the entirety of its reserve interest with no distribution-partner split at all. But Tether operates offshore, outside the US national trust bank framework that Circle, Ripple, Paxos, BitGo and Fidelity have all pursued, and the Genius Act's provisions targeting foreign issuers create a genuine long-term question over Tether's access to the compliant, bank-integrated corridors that Visa's platform and Open USD are both racing to own.

The regulatory deadline that arrived with no rules attached

All of this is unfolding against a specific and unusual regulatory backdrop. The Genius Act, signed into law on July 18, 2025, gave five federal agencies, the OCC, the Federal Reserve, the FDIC, Treasury and FinCEN, exactly one year to finalize the implementing rules covering reserves, capital, custody, redemption and state certification for payment stablecoins. That deadline passed on July 18, 2026 with zero final rules published across any of the five agencies, despite eleven proposed rulemakings and dozens of open comment periods. Crucially, the statute carries no penalty for a missed deadline and no court is positioned to compel faster action, which means the law's true effective date defaults to its statutory backstop: 120 days after final rules are issued, or January 18, 2027, whichever comes first. At the current pace, the January backstop looks like the more likely trigger.

That regulatory vacuum has not slowed anyone down. In the same narrow window, the OCC granted Circle a national trust bank charter, authorizing Circle National Trust to open on or after July 10, 2026, while also issuing charters or conditional approvals to Ripple, Paxos, BitGo, Fidelity Digital Assets and Crypto.com. Visa launched its stablecoin platform. The Open USD consortium went live. None of these moves waited for a finished rulebook, because waiting was never going to be rewarded: the companies that moved first are shaping what the eventual rules will need to accommodate, rather than the other way around.

What actually decides who wins this fight

Strip away the announcements and three separate contests are running simultaneously. The first is between issuers: Circle, Tether, Ripple and Paxos competing on trust, compliance infrastructure and bank integrations. The second is between issuers and distributors: whether Circle can hold its Retention Ratio steady when its Coinbase agreement comes up for renewal in August 2026, at the same moment Coinbase has joined the Open USD consortium that exists specifically to make single-issuer arrangements like Circle's obsolete. The third is between the entire stablecoin category and tokenized bank deposits like JPMD, which sidestep the reserve-income question altogether by letting banks pay yield directly.

Circle's own hedge against all three is real but still small: its Arc blockchain, valued around $3 billion after a $222 million token pre-sale, its Circle Payments Network processing roughly $8 to 10 billion in annualized volume across more than 130 financial institutions, its new national trust charter, and a partnership with Nomura to build USDC-based settlement for the roughly $440 billion a day Japanese corporate market. Together these initiatives generate around 6 percent of Circle's current revenue. The escape plan from being purely a rate-sensitive bond fund with a partner-revenue-sharing problem is underfunded relative to the core USDC business, but it is the only part of Circle's story that does not depend on either the Fed or Coinbase.

For DN's readers, the practical takeaway is that a stablecoin issuer's headline market share tells you almost nothing about how much of its economics it actually keeps. The Retention Ratio, not the circulation figure, is the number worth tracking, and it applies equally to any DePIN or on-chain protocol claiming a revenue share from a partner integration. Model your own scenario below.

For readers looking to position around stablecoin and payments infrastructure exposure directly, spot and derivatives markets are available through most major exchanges, including Bybit, OKX and MEXC. As always, this is not financial advice. Distribution economics are measurable. Which structure wins the next phase of this market is a judgment call that deserves independent due diligence.

Frequently asked questions