In February 2026, stablecoins settled $7.2 trillion in transactions — surpassing the ACH network for the first time in history. The headline was striking, but the detail underneath it was more interesting: most of that volume wasn't crossing borders. It was moving locally.
For years, the clearest use case for stablecoins was obvious: cross-border payments. Send dollars to a supplier in Vietnam or a contractor in Colombia without the 3–5% FX spread, the 2–5 day settlement window, or the correspondent banking maze. The narrative was compelling, the pain was real, and the early traction seemed to confirm it.
Then a16z Crypto ran the numbers.
Their April 2026 analysis — nine charts covering adjusted stablecoin volume, transaction composition, velocity, and geographic distribution — surfaced something nobody in the industry was really prepared for: cross-border's share of stablecoin payment volume has fallen from roughly 50% in early 2024 to roughly 25% in early 2026. Intra-country transactions now account for approximately 75% of all stablecoin payment activity.
Stablecoins aren't primarily a cross-border story anymore. They're becoming local infrastructure.
How it happened
The clearest example is Brazil. The Brazilian real-backed stablecoin BRLA went from near-zero in early 2023 to approximately $400 million per month by early 2026, driven primarily by its integration with PIX, Brazil's domestic instant payment system. Brazilians aren't using BRLA to send money abroad. They're using it to move money inside Brazil, faster and with more programmability than their existing banking infrastructure allows.
This pattern — local currency stablecoin, local payment rail, domestic commerce — is the template for what stablecoins are actually becoming in markets with active adoption. MiCA in Europe accelerated the same dynamic from the other direction: USDT delistings created structural demand for euro-backed and other non-USD stablecoins, which have now settled at a persistent $15–25 billion per month in activity — higher than the pre-MiCA baseline.
The macro-level signal is velocity. Stablecoin velocity — transfer volume relative to circulating supply — doubled from 2.6x to 6x between early 2024 and early 2026. That's the signature of a real payments network, one where currency circulates rather than sits.
The shift that changes the product roadmap
If you were building infrastructure on the assumption that stablecoin payments are primarily a cross-border corridor play, this data requires a rethink.
The off-ramp problem is a good example. For years, the stablecoin off-ramp bottleneck was framed as an emerging-market issue: you receive USDC from an international buyer, and converting it back to local currency costs 6–8% and takes days. That's still a real problem. But if 75% of stablecoin payment volume is now intra-country, the off-ramp problem is also a domestic infrastructure problem. Local merchants, local payroll, local commerce — all of them need to convert stablecoin balances back to fiat, and none of them want to pay cross-border rates to do it.
The C2B data tells the same story from the demand side. Consumer-to-business transactions grew 128% year-over-year in 2025, to 284.6 million transactions. Stablecoin card programs — which let consumers spend USDC balances at any Visa-accepting merchant — have driven monthly collateral deposits from near zero to over $300 million per month in roughly a year. This is mostly local commerce, settled instantly.
The trust gap that comes with going local
Here's what the cross-border framing obscured: when stablecoin payments move from crypto-native corridors to everyday local commerce, the people on both sides of the transaction are no longer sophisticated users. They're ordinary merchants and ordinary consumers — and ordinary consumers expect chargebacks.
Stablecoin payments don't have them. A payment on-chain is final. There is no card network to call, no issuer to reverse the charge. That gap was manageable when stablecoin users were a self-selecting group who understood what they were signing up for. It becomes a structural barrier as stablecoins become general-purpose local payment infrastructure.
This is where programmable money's native advantages are most underexplored. Unlike a bank wire or card payment, a stablecoin transaction can be designed with conditions built into the payment itself — funds held in escrow until delivery is confirmed, released automatically when both parties have signed off, or routed to a dispute arbiter if something goes wrong. The conditionality that requires a separate legal contract and a bank's manual intervention on legacy rails can be encoded directly into the transaction. That's not a marginal improvement. It's a different category of financial primitive.
The infrastructure serving local stablecoin commerce will eventually need to offer protections equivalent to what cards offer today — but built natively for programmable money rather than grafted on from a different era of finance. The companies that figure out what that looks like will find a market that's already here, not one that's still arriving.
What this means for the next two years
Absolute cross-border volume is still growing. Nobody is arguing that the original use case is broken. But the proportion shift matters because it tells you where the new infrastructure spending needs to go.
The companies building for cross-border corridors are fighting over a share of the market that is shrinking relative to total activity. The gap that corresponds to where volume is actually growing is different: domestic off-ramps, local-currency stablecoin issuance, commerce-grade consumer protection, and the programmable payment logic that lets merchants and platforms do things with stablecoin rails they simply cannot do with a bank account.
The narrative hasn't caught up with the data yet. That's usually where the interesting infrastructure gets built.