Nobody Launches a Stablecoin Anymore. Everybody Launches the Rails Underneath It

Two companies can hold the same license, serve the same corridor, and settle in the exact same token, and one of them earns a few basis points on reserves while the other owns the app a customer opens every morning. The difference isn't the token. It's which layer of the stack they decided to build on.

For most of the last five years, the assumed answer to "how do we win in stablecoins" was to issue one. In 2026 that assumption quietly stopped being true, and the market told you so in the least ambiguous way available: with acquisition prices.

What "Distribution Beat Issuance" Actually Means

Issuance is the business of minting and redeeming a token and managing the reserves behind it. It sounds like the whole industry, and for a while it was. The float income is real, the balance sheet is enormous, and the brand sits on every screen.

Distribution is everything between that token and a person or a business actually using it: the wallet it lives in, the custody and compliance that make it legal to hold, the chain it settles on, the API that routes it, and the card that spends it. None of that mints anything. All of it sits closer to the customer.

The 2026 shift is that competitive advantage migrated up the stack, which is exactly what happened in card payments, in cloud, and in mobile. The unit of value became access rather than the asset. Financial institutions increasingly want stablecoin functionality without the regulatory burden of issuing, and that preference alone reprices every layer above the issuer.

Why This Flipped in 2026 and Not Earlier

Three things had to be true at the same time, and 2026 is the first year all three were:

• Issuance became a regulated commodity. The GENIUS Act gave the US its first federal rulebook for payment stablecoins in July 2025. Once the rules for minting a compliant dollar token are written down, minting one stops being a moat and starts being a checklist. Tether's response tells you where the pressure went: it had KPMG US complete the first full independent audit of its 2025 financials, competing on credibility rather than novelty.
• Supply stopped compounding. Total stablecoin supply is up 14.3% year over year, which sounds healthy until you look at where that growth landed. It happened in late 2025. Through 2026 the number has been essentially flat, peaking at $322.4B on 17 May and sitting at $308.0B on 13 August. If the pie isn't growing, the fight moves to who serves the slices.
• The consortium model attacked issuer economics directly. On 30 June 2026, a group of more than 140 companies operating as Open Standard announced Open USD, a dollar token where nearly all reserve income flows to participating institutions after a management fee, rather than to a single issuer. Visa, Mastercard, Stripe, Coinbase, BlackRock, Google Cloud, BNY and Shopify are all in it, and it's led by Zach Abrams, who co-founded Bridge before Stripe acquired it. Circle's stock fell 15–17% the day it was announced. The market read it correctly: the point of Open USD isn't to be a better token, it's to make the token layer a shared, low-margin utility that the distributors control.

That third one deserves a note on timing. Open USD was announced in June and is expected to go live later in 2026 on Solana, Stellar, Base and Polygon. As of this writing it's a signed consortium and a stated design, not settled volume. What it has already changed is the negotiating position of every distributor in the market.

Where the Margin Actually Sits

Walk the stack from the bottom and the pattern is hard to miss.

Issuance still holds the reserves, and at current rates that is a genuinely large business. It's also the layer with the most regulatory exposure, the most political attention, and now an organized consortium of its own largest distributors working to compress its take.

Settlement rails turned into a serious arena in 2026, and the notable part is who is building them. Tempo, from Stripe and Paradigm, went to mainnet in March with roughly 0.6-second finality, no native volatile token, and fees paid in stablecoins through a built-in fee AMM. It added Stripe, Visa and Zodia Custody as its first external validators in April. Circle raised $222M in an ARC presale on 11 May at a $3B fully diluted valuation for a chain that uses USDC as gas. The largest issuer in the market spent its year building a rail.

Custody and compliance is the layer that decides whether a regulated institution can touch any of this, which is why it prices like infrastructure rather than software.

Orchestration is where the clearest number landed. Mastercard agreed on 17 March 2026 to acquire BVNK for up to $1.8B, including $300M in contingent payments, and closed the deal on 3 August. BVNK doesn't issue anything. It moves roughly $30B a year across more than 130 countries for customers like Worldpay, Deel and Flywire. Coinbase had reportedly been in talks for it at around $2B before those discussions collapsed. Two of the largest companies in payments and crypto both concluded that the routing layer was worth roughly two billion dollars, and neither of them was buying a token.

Distribution is the layer everything above is really competing for, and it's the one showing the cleanest growth curve. Crypto card spending reached $1.04B in July 2026, roughly triple the $382M recorded in August 2025, across more than 10 million tracked transactions. Dollar-backed stablecoins funded about 70% of them, with USDC at 50.8% of volume and USDT at 20.3%. The average payment rose to about $86 from $59 a year earlier, which is the detail that matters most: that's grocery-and-ride-hailing money, not treasury movement.

The Numbers, and the Caveat Behind Them

Every stablecoin deck you'll see this year opens with a transfer-volume number, and most of them are technically accurate and practically misleading.

The 2025 headline is roughly $33T in annual transaction volume across USDC and USDT alone, up about 72% year over year, and monthly settlement figures that passed the ACH network in February 2026. Those are real measurements. They're also gross on-chain transfers, which include bot activity, arbitrage loops, exchange rebalancing, and money moving between wallets that belong to the same entity.

Analysis of 2025 flows puts genuine economic activity at roughly 7% of transfer volume, about $4.2T out of $62T. Payments for actual goods and services were smaller again, in the range of $350–550B on BCG's analysis. That's not a small business, but it is roughly one percent of the number that usually leads the slide.

The honest version of the adoption story sits in the survey data rather than the on-chain data. EY-Parthenon's 2025 research found around 13% of organizations using stablecoins, split as 23% of financial institutions and 8% of corporates, with 54% of non-users expecting to start within six to twelve months and 41% of users reporting cost savings of 10% or more. That's a market in the early-majority handoff, not a solved one, and it's why the distribution layer is being bought now rather than later. Whoever owns the integration when those 54% arrive owns the relationship.

What This Changes If You're Building

If you're a startup, a fintech, or a bank looking at stablecoins as a roadmap item, the practical consequences of the shift are fairly concrete:

• Launching your own token is now a distribution question, not a treasury one. A branded stablecoin with no wallet, no card program, and no corridor behind it is a liability with a logo. The teams shipping branded tokens successfully in 2026 attach them to an existing spend or settlement flow on day one.
• Multi-token support is the default assumption. Harness, the middleware StablecoinX shipped on 2 July, is built around exactly this: accept any major stablecoin, hold it in the form you want, deliver it in the form the receiver wants. Products that hard-code one token are building in a migration project.
• The orchestration layer is a build-or-buy decision, and it's mostly buy. Bridge, BVNK, Zerohash and their peers exist because routing, reconciliation, and off-ramp coverage across 130 countries is a multi-year problem, not a sprint. The teams that build it themselves usually do so because their corridor is genuinely unusual.
• Compliance is a product surface, not a legal appendix. Screening, attestations, and jurisdictional routing show up in your UX as delays, holds, and rejected transfers. Designing for those states is product work.
• Distribution-first also means consumer-grade onboarding. The last mile ends at a person tapping a card or a business approving a payment run. That's a wallet-and-interface problem, and it's the reason embedded wallet infrastructure and stablecoin infrastructure keep converging on the same roadmaps.

How to Pick Your Layer

A short set of questions before committing engineering time:

• Do you already have distribution — users, merchants, payroll, a corridor — or are you hoping the token will create it? If the second, you're solving the harder problem first.
• Is your differentiator the movement of money or the reason someone is moving it? Movement is a crowded, well-capitalized layer. The reason is usually your actual product.
• Can you name the specific regulatory permission your product needs, and do you hold it, rent it, or plan to acquire it? Every credible answer at the custody and issuance layers starts here.
• Would a customer notice if you swapped the underlying token tomorrow? If not, the token isn't your moat and shouldn't be your roadmap.
• Where does your revenue come from — float, spread, per-transaction fees, or software? Float income is the most exposed line item in the market right now, and consortium economics are aimed directly at it.

The Bottom Line

The stablecoin market didn't stop growing in 2026 so much as it stopped growing in the place everyone was watching. Supply went flat, transfer volume stayed inflated by activity that isn't commerce, and the real money went into buying the layers between a token and a customer: a $1.8B routing company, a payments chain with no gas token, a 140-company consortium designed to make issuance a shared utility, and a card business that tripled in a year on ordinary purchases.

If you're planning a stablecoin product, the useful question isn't which token to use. It's which layer you intend to own, and whether you can defend it once the companies that already own the customer decide they want it too.

Frequently Asked Questions