Swift Spent Fifty Years Sending Messages About Money. Now It Is Reaching for the Money Itself.

Swift Spent Fifty Years Sending Messages About Money. Now It Is Reaching for the Money Itself.

Last week, I used Swift to send money to Australia. You’ve probably also used Swift for your salary, your rent, and your last overseas invoice. But we never think much about it. Like all good plumbing, we notice only when it backs up.

Swift itself moves nothing, yet it carries the instructions behind roughly $150 trillion in cross-border payments a year, about 1.4 times global GDP.  It sets a standard format and carries the instructions that tell each bank what to pay, while the cash itself settles through correspondent accounts and national payment systems. For an organization that touches so much money, it keeps none; SWIFT is run as a non-profit cooperative owned by its 3,500 members, who hold shares in proportion to how much they send and elect an unpaid board.

On July 9, 2026, Swift unveiled larger ambitions. It declared its blockchain-based shared ledger ready for use, with 17 banks preparing to pilot tokenized cross-border payments (Swift). Citi, HSBC, UBS, Standard Chartered, DBS, MUFG, Wells Fargo, First Abu Dhabi Bank and nine more are on the list. The ledger, on an Ethereum-compatible network with permissioned access, records how much each bank owes the others and lets tokenized deposits move around the clock, including nights and weekends.

The change behind the headline is small but material. Previously, Swift could carry instructions on a Saturday, but no money would move until the banks reopened on Monday. On the new ledger, a bank can transfer a tokenized deposit—a digital claim on commercial-bank money—and the recipient can hold it, see it, and pay it onward before the weekend is out.

What still lags is final settlement. Swift’s network runs 24/7/365, but the banks and their core settlement systems, Fedwire and T2, do not. Call it a faster way to promise, and a partly faster way to pay, not yet a faster way to settle with finality.

This is a small step against an enormous problem: cross-border payments are neither fast nor cheap. The average cost of sending a $200 remittance was 6.49% in early 2025, more than double the G20’s 3% target for 2030.

Much of the cost is idle cash. To pay on demand in currencies where they hold no local account, banks pre-fund nostro accounts (“our money at your bank”) with other banks around the world. By one widely cited estimate, around $27 trillion sits in these accounts to support cross-border settlement, and that money earns little. A bank holding $1 billion in a foreign account gives up about $50 million a year at a 5% rate, and that drag repeats across every currency.

With a 24/7/365 ledger, a claim on money can move and be re-used at any hour, allowing banks to pre-fund less and redeploy the difference into more productive activities. Companies receive usable funds sooner, which shortens the gap between shipping goods and getting paid. And a dollar that settles faster can work harder, clearing two transactions where previously a weekend in limbo allowed one.

The hardest part of all this was never the technology; it’s getting humans, with all their egos and agendas, to agree. In the early 1970s, banks sent payment instructions by Telex, a telegraph-style machine that printed messages onto paper tape. Telex was slow, with no shared format, and left room for error and fraud on every line. In 1973, 239 banks from 15 countries founded Swift to fix this, agreeing on a single message format that every member would use. On May 9, 1977, Prince Albert of Belgium sent the first message on Swift, and within a year Swift carried 10 million messages. Swift won because it got hundreds of bankers to agree to speak one language.

Swift is not alone in wanting to tackle this problem. JPMorgan settles billions a day on its Kinexys ledger and has put a deposit token to work. Partior, built by JPMorgan, DBS and Temasek, runs a shared ledger for interbank settlement. Fnality settles wholesale payments in central-bank money and won a UK settlement-finality designation in 2024. The Bank for International Settlements is tokenizing reserves and deposits through Project Agorá, with Swift itself taking part.

We’re moving toward a world that never closes, where I can send money to my cousin in Australia on a Sunday and it will settle the same day. The open question is whose ledger sits underneath: a member-owned cooperative, a single large bank, or a central-bank consortium. Swift has the reach, and the historic ability to get banks to agree. Whether that will be enough is worth watching.