Stablecoins are a new layer of global payments, not a new payments system

Every few years the payments infrastructure industry discovers a technology that is going to make banks, card networks and correspondent banking obsolete by roughly next Tuesday. Stablecoins are the current holder of that title. The claim is wrong, but it is wrong in an interesting way, and the interesting part is worth your time.

What a stablecoin actually is, and what it is not

A stablecoin is a token issued on a blockchain that is designed to hold a steady value, usually one unit of a fiat currency. The credible ones — and the distinction matters — are fully backed by reserves: cash and short-dated government paper held one-for-one against the tokens in circulation, with regular attestation of those reserves. That is the model behind the large regulated issuers. The algorithmic variety, which tried to hold its peg through clever incentives rather than actual money in a bank, had its reckoning in 2022 and we need not dwell on it.

So a payment stablecoin is, in essence, a bearer claim on a dollar or a euro that happens to travel on a public ledger. This is genuinely useful. It is also considerably less revolutionary than the marketing suggests. A tokenised dollar is still a dollar; the novelty is in how it moves, not in what it is.

But are stablecoins not already moving trillions

You will have seen the number. Stablecoins settled somewhere around 35 trillion dollars on-chain in 2025, a figure that comfortably exceeds the annual throughput of some card networks and gets quoted as proof that the revolution has already happened.

It has not, and the number is the reason to be sceptical rather than impressed. When analysts strip out trading, arbitrage and the automated bot-to-bot transfers that make up the overwhelming majority of that traffic, the volume that represents an actual payment — someone paying someone else for something in the real economy — was on the order of 390 billion dollars for the year, by McKinsey and Artemis estimates. That is barely 1% of the headline. It is still real, still growing fast, and still worth having. But 35 trillion dollars of mostly-machinery is not a payments system displacing the incumbents. It is a very active trading venue with a useful payments business attached.

The Bank for International Settlements makes the point more soberly: stablecoin market capitalisation stood at roughly 320 billion dollars in mid-2026, which it describes, accurately, as dwarfed by the trillions sitting in bank deposits. Keep both facts in view. The technology is growing quickly and the base is small. Both things are true, and only quoting the first is how you end up with a strategy built on a press release.

Will stablecoins replace Swift and the banks

No. And the reason is instructive, because it is the same reason every “Swift killer” of the past twenty years has failed to kill Swift.

Moving value from A to B is the easy part. It was never the hard part. The hard part is everything wrapped around the movement: knowing that A and B are who they claim to be, that neither appears on a sanctions list, that the money is not the proceeds of something unpleasant, that the transaction can be reported to the right regulator in the right format, and — the part everyone forgets — that the recipient can actually spend the result. A market trader in Dar es Salaam cannot pay for vegetables in USDC. She needs shillings, in a wallet or an account she already uses.

A blockchain does the value-movement part beautifully. It does none of the rest. Compliance, identity, foreign exchange, and the last-mile conversion back into money people can use — none of that is native to the token. It has to be supplied by regulated institutions operating under real licences. Which is to say: by the incumbents, or by companies like ours that connect to them.

Stablecoins add a settlement option. Every layer above and below still has to be there.

Where stablecoins genuinely add value today

Set the hype aside and four uses stand up to scrutiny.

Settlement between institutions

This is the strongest case. Two payment firms, or a firm and its liquidity provider, can settle obligations in minutes on a weekend instead of waiting for correspondent banking cut-offs and the next business day. The token here never touches a consumer. It is plumbing between professionals, and as plumbing it works well.

Treasury and liquidity

A business moving money across a dozen markets has to pre-fund accounts in each of them — capital sitting idle against payouts that may or may not come. Stablecoins let treasury teams move that liquidity between corridors far faster than a wire, which means less trapped capital and tighter funding. This is not glamorous. It is exactly the kind of efficiency that compounds.

Business-to-business cross-border payments

This is where the real payments volume actually sits. Business-to-business transactions make up roughly 60% of genuine stablecoin payment activity, and grew several times over in the past year. For a supplier invoice between two companies in markets with a thin or expensive correspondent banking relationship, a stablecoin leg in the middle can cut both cost and time. The first and last miles remain fiat; the middle is tokenised. The customer sees an invoice paid faster and does not care what happened in between, which is precisely as it should be. It is telling that the institutions moving fastest here are the incumbents: Visa has been settling in stablecoins across several blockchains, and Mastercard added settlement support for USDC, PYUSD, USDG and RLUSD in mid-2026. The card networks did not conclude that stablecoins would replace them. They concluded that stablecoins were a useful new rail, and picked them up.

Programmable payments

Because a stablecoin lives inside a smart contract, you can attach conditions to it: release on delivery, split automatically between parties, escrow until a milestone. This is early and mostly promise rather than production, but it is a real capability that older rails do not have.

What the regulators have decided

The interesting shift of the past two years is that the “wild west” framing is now out of date. The rules have arrived.

In Europe, the Markets in Crypto-Assets regulation, MiCA, brought payment stablecoins — what it calls e-money tokens and asset-referenced tokens — under a formal regime in force since 2024, with reserve, authorisation and disclosure requirements, and a transitional window for service providers running into mid-2026. In the United States, the GENIUS Act was signed into law in July 2025 and is now in its rulemaking phase ahead of taking full effect. It requires payment stablecoins to be backed one-for-one by cash and short-dated government paper, and — a detail worth noticing — bars issuers from paying interest on them. Singapore’s framework, the UAE’s payment-token rules and the United Kingdom’s final crypto regime, published in 2026, all point the same way. The specifics differ, and the exact dates are worth confirming against the current text before they go in a board paper, but the direction is unmistakable: Payment infrastructure stablecoins are becoming regulated financial instruments, issued by supervised entities, backed by real reserves.

This is the opposite of what the early enthusiasts wanted, and it is the reason stablecoins now matter to serious institutions. A regulated instrument that settles in minutes is a tool a bank can actually pick up. An unregulated one was a headline.

A practical note for anyone building on this: regulation is what turns a stablecoin from an interesting experiment into something you can put in front of a compliance committee. The compliance obligations do not disappear because the money is on-chain. Travel rule, sanctions screening and know-your-customer apply to a stablecoin payment exactly as they apply to a wire. If a provider tells you otherwise, that is not a feature. It is a liability.

So what should a bank or a payments firm do

The BIS puts the deeper reason for this well. Sound money needs three properties: singleness, so that a dollar is always worth a dollar wherever it is held; elasticity, so that obligations settle without the system seizing up; and integrity, the safeguards against financial crime. A stablecoin, on its own, delivers none of these fully. It borrows them from the regulated system it sits inside. That is not a criticism. It is the job description of a layer.

So treat stablecoins as what they are: another rail in the toolkit, to be used where they beat the alternative and ignored where they do not. The winning position is not “for” or “against” stablecoins. It is the ability to move value across whichever rail is fastest and cheapest for a given corridor at a given moment — a card network here, an instant payment scheme there, correspondent banking where it still makes sense, and a stablecoin leg where it genuinely helps — while presenting the customer with a single, simple experience and handling the compliance underneath.

That is the argument my colleague Arup makes in the endpoint economy, and it is the same argument in a different dress. The future is not one rail defeating the others. It is orchestration across all of them. Stablecoins are a welcome and increasingly serious addition to that set. They are a new layer of global payments. They are not a new payments system, and the companies that understand the difference will get considerably more out of them than the companies still waiting for next Tuesday.