BASEL, 22 AUG 2026 — Project Nexus, the Bank for International Settlements initiative to connect domestic instant payment systems through a single shared hub, has moved from blueprint into implementation. Governance now sits with Nexus Global Payments, a not-for-profit incorporated in Singapore in March 2025, and a tender for a network operator is under way.
The participants are the central banks and payment system operators of India, Malaysia, the Philippines, Singapore and Thailand, with Bank Indonesia as a special observer. The go-live year is reported as both 2026 and 2027.
Where the project stands
The current phase covers a formal rulebook, technical implementation guides and ISO 20022 specifications, overseen by the new managing entity. Bank Indonesia participates as an observer rather than a full member.
What Nexus does differently from the bilateral links
Singapore and Thailand already connect PayNow and PromptPay directly, and several other bilateral corridors exist across the region. Nexus is not just another set of bilateral links; it is a multilateral hub, and the difference is what makes it ambitious.
Bilateral linking scales badly. Connecting every country to every other country requires a number of arrangements that grows with the square of the participants, each with its own contract, technical integration, settlement arrangement and dispute process. Five countries linked bilaterally is ten arrangements; ten countries is forty-five.
The Nexus model uses a single scheme — one rulebook, one message format, and one technical connection — to give each participant access to all the others. The work stops growing quadratically and starts growing linearly.
It is a better design, but it is also why progress is slow. A bilateral link needs two parties to agree. A multilateral scheme needs everyone to agree, and then needs an operator, a governance body and a rulebook that all of them accept.
Two published go-live dates is a fact about the project
The launch is reported as being in both 2026 and 2027, and it’s better to acknowledge the conflict than to pick one.
A project with a new governance entity, a rulebook in development, and an open operator tender is not weeks from launch. The sequence of selecting an operator, contracting, building, testing, and securing regulatory approval in five countries takes years, not months. This makes 2027 the more plausible date, and even that assumes the tender concludes promptly.
The discrepancy is ordinary rather than sinister. A central bank describing an intended date and a scheme entity describing a delivery date are answering different questions, and a project this size accumulates several official-sounding timelines. But it does mean that any single date quoted for Nexus should be treated as an aspiration attached to a named speaker rather than a schedule.
The checkable milestone is the operator award. Until that is announced, the delivery clock has not started.
Why the regional case is stronger than the global one
Nexus is framed as a global project, and its value is concentrated in exactly this region for reasons that are structural.
Southeast Asia and India combine very high domestic instant payment adoption with heavy intra-regional migration and remittance flows. UPI processes transaction volumes that dwarf most national systems. Millions of workers move money between these five countries routinely, and they currently pay correspondent-banking costs and wait correspondent-banking times to do it.
This combination of mature domestic systems and heavy cross-border remittance traffic justifies the high coordination cost of interlinking. It does not exist to the same degree between, say, two large economies with low instant-payment adoption and modest bilateral remittance volumes.
Even if Nexus succeeds here, that will not prove the model works elsewhere. The regional case is the strong case.
The unglamorous problems are the ones that decide it
Message formats and connectivity are the easy part. Three harder problems get less attention.
First is foreign exchange. An instant cross-border payment needs a rate quoted and held at the moment of transfer, so someone has to carry the position and charge for it. How transparently they do so will determine if users find Nexus cheap.
Second is sanctions and financial-crime screening. A payment crossing five jurisdictions inherits five separate screening regimes. Reconciling them without reintroducing the latency Nexus was built to remove is a difficult design problem.
The third and least-discussed problem is liability. When an instant payment goes wrong across borders — a bad recipient, fraud, a failed leg — the rulebook has to specify who reimburses whom, and under which country's law. This is why rulebooks take years to write.
It is racing private rails that do not need anyone's permission
The competitive context is rarely mentioned in official material and is the most important thing about the timeline.
While five central banks negotiate a rulebook, stablecoin settlement and private remittance operators are moving small-value money across these same corridors today. They do it without multilateral agreement because they do not need one, and they are improving on a product cycle rather than a governance cycle.
Nexus has advantages that private rails cannot easily match, including settlement in central bank money, regulatory certainty for banks, and direct access to the domestic systems where most people keep their accounts. Those matter, and they are why banks would prefer it.
The disadvantage is arithmetic. A project whose delivery date is disputed between 2026 and 2027, and whose operator has not been appointed, is competing against alternatives that ship continuously. Every quarter of coordination is a quarter in which the behaviour Nexus intends to serve settles somewhere else, and payment habits are sticky once formed.
This is the honest case for urgency, and it is a stronger one than any of the project's own efficiency claims.
What remains unconfirmed
No go-live date is confirmed and the two reported years cannot be reconciled from the available material. The operator tender has not been reported as awarded, and no shortlist or timetable is described.
Which specific payment systems will connect in the first phase is not established, nor whether all five participants will launch simultaneously or in stages. Pricing, foreign-exchange arrangements, transaction limits and the liability framework are not described. Whether Bank Indonesia intends to move from observer to participant is not stated.
What to watch for
The operator award is the milestone that converts intent into a delivery programme, and it is the next thing that can actually be observed.
The second signal is the rulebook, specifically whether it is published or held privately among participants. A public rulebook lets banks and payment firms prepare and lets outside parties assess the liability model; a private one means the terms become visible only to those inside the scheme.
The third is Indonesia. An observer becoming a participant would be the clearest evidence that the arrangement is attractive to those watching it, and Indonesia is the largest economy in the region currently outside it.
