SWIFT's Shared Ledger: Ready for Use, Yet Without a Single Transaction

On 9 July 2026, the Society for Worldwide Interbank Financial Telecommunication (SWIFT) announced that its blockchain-based shared ledger is ready for use. Seventeen banks across six continents are preparing initial pilot transactions with tokenised deposits. The trade press reported the event as a launch, a rollout, a ribbon-cutting. The wording of the announcement is more restrained: the ledger is “ready for initial use", and the banks are “preparing to pilot live transactions". So far, no money has moved.

This distinction is not pedantry. It determines what a treasurer can plan for in the third quarter of 2026 and what they cannot. Between a ready infrastructure and a payment route a bank relies on for liquidity management lie pilot phases, supervisory clarifications and the question of whether enough counterparties join in.

What SWIFT announced, and what it did not

The shared ledger is an orchestration layer. Participating banks issue tokenised deposits on their own respective ledgers; the new layer coordinates how those deposits move between institutions. In the words of the announcement, it allows banks to move funds for customers “including overnight and on weekends – before completing final settlement through existing systems".

The finality of a payment, meaning the moment it becomes legally irrevocable, still arises in the relevant central bank's real-time gross settlement system or in the correspondent banking relationship. The blockchain layer creates no new settlement, but bridges the time before it: the customer can dispose of the amount before the payment is final.

Technically, SWIFT builds on Hyperledger Besu, an open-source architecture compatible with the Ethereum Virtual Machine (EVM), operated as a closed network of approved participants. Smart contracts enforce the transaction rules. Consensys had built the conceptual prototype that SWIFT presented at Sibos in Frankfurt in September 2025.

Tokenised deposit, stablecoin, central bank digital currency

The three terms are readily conflated in press coverage, although they describe three different liabilities. A tokenised deposit is the digital representation of a bank liability. It remains a liability on the issuing bank's balance sheet and is covered by existing deposit protection. A balance of 100 euros at one's own bank remains a claim against that bank, whether it is held in a core banking database or as a token on a ledger.

A central bank digital currency (CBDC), by contrast, is a liability of the central bank and carries no bank credit risk. A stablecoin, in turn, is predominantly issued by non-banks and may deviate from par. The Bank for International Settlements describes the underlying problem as a question of the singleness of money: a euro should be a euro, regardless of who holds it and where it sits.

SWIFT is therefore taking the most conservative of the three routes. The ledger moves money that already sits on bank balance sheets, under the regulatory regime that applies to that money anyway. For a bank, that is the path of least supervisory resistance.

What changes for treasury and payments operations

The benefit of moving tokenised deposits around the clock does not lie in the speed of individual payments. SWIFT itself states that 75 per cent of payments on the network reach the beneficiary bank in under ten minutes. That figure is self-reported: it measures not the credit to the end recipient, but delivery to the beneficiary bank. Anyone who believes the core problem of cross-border payments is transmission speed has not yet lived through bank holidays in the opposite direction.

The problem lies in the time zone and the calendar. A payment initiated on a Friday evening European time, arriving in a market whose real-time gross settlement system does not open until Monday, ties up liquidity across the weekend. Banks hold balances on nostro accounts for this purpose, meaning accounts with foreign correspondent banks. That money does not work. How much of it there is globally circulates in a surprising number of versions, none of which has a robust primary source; the order of magnitude is therefore better left unquantified.

If a bank can instead move tokenised deposits at the weekend, the pre-funding shifts. The customer receives availability before settlement is final. Someone bears the credit risk of that interim phase, and who that is depends on the contractual construction of the ledger. For payments operations, this means concretely: cut-off times in payment processes initially remain in place, because they are tied to the opening hours of the settlement systems. What changes is the customer experience beforehand.

Why bank consortia fail, when they fail

There is a graveyard of shared blockchain platforms in banking. we.trade and Batavia in trade finance, TradeLens in logistics, Marco Polo Network, B3i in reinsurance, Contour, the Australian exchange's CHESS replacement project. None of these ventures failed on the technology, but on governance. Banks do not hand their client relationships to a shared infrastructure, and competitors do not route their data through a rival's platform.

SWIFT sits at a different point in this sequence, because it already is the shared infrastructure. More than 11,500 institutions in over 200 countries and territories use the network; by its own account, the cooperative moves the equivalent of world gross domestic product every two to three days. The question of the neutral operator, on which the predecessor projects foundered, does not arise afresh here.

It is replaced by a different one. The same large banks piloting SWIFT's ledger are simultaneously investing in competing architectures: in Project Agorá of the Bank for International Settlements, in JPMorgan's Kinexys, in Partior. That is not a sign of convergence but of indecision. As long as the large banks hedge across several tracks, each individual route remains below critical mass, and network effects emerge on none of them.

The clause that says more about positioning than the announcement does

Thierry Chilosi, Chief Business Officer at SWIFT, places the ledger in a longer arc. He calls it a foundation for “future innovation in areas like programmable money and agentic commerce". Agentic commerce describes trade in which software agents initiate transactions independently, without a human approving each one.

An agent that pays on its own needs a means of payment that is available around the clock, carries programmable rules and sits within a regulated framework. Tokenised deposits on a permissioned ledger satisfy all three conditions. That the headline of this announcement is the seventeen banks rather than this clause says something about the attention economy of the trade press.

At the same time, this outlook sharpens the governance question. Once an agent may initiate payments, the interim phase between availability and finality becomes a window of attack. A bank opening tokenised deposits to agentic payment flows needs amount limits, approval thresholds and logging that attributes every initiation to an accountable system.

Recommendations for operational practice

For institutions outside the pilot group, the announcement is not a call to act, but an occasion to plan. Four fields of action stand out.