The Bank of England, Monetary Authority of Singapore and Bank of Thailand have completed a trial coordinating foreign exchange settlement across traditional payment systems and distributed ledgers. Findings published on November 13, 2026, describe Project Meridian FX APAC, with Partior building the prototype that connected the central banks' test environments.
The Bank of England plans to launch a live synchronisation interface in 2028. That date applies to its own programme: participation in the experiment carries no equivalent deployment commitment from Singapore or Thailand. The trial tested settlement mechanics in controlled environments.
The practical result is a demonstration that different types of financial infrastructure can coordinate currency payments while retaining their own settlement designs.
Different ledgers, one coordinated exchange
The project brought together systems representing three approaches to wholesale money.
The UK's test environment simulated an account-based Real-Time Gross Settlement system. Thailand's Project SAN used a tokenised settlement model with an interoperability layer. The Singapore environment used a non-production configuration of SGD Testnet, supporting wholesale central bank digital currency.
Partior developed the Synchronisation Operator prototype that coordinated their transactions. Its job was to match the currency legs of an FX trade, arrange the reservation of funds and initiate settlement once the required money was available.
The central protection is payment-versus-payment, or PvP. Delivery of one currency depends on delivery of the other. This addresses the risk that a participant pays out the currency it owes but fails to receive the currency it bought.
The scale of that exposure is substantial. The central banks' announcement cites the 2025 BIS Triennial Central Bank Survey: more than $1.4 trillion, or 10% of average daily FX settlement in April 2025, remained fully exposed to settlement risk.
For institutions assessing tokenised infrastructure, the experiment supports a specific design choice: connecting different settlement systems through a coordination layer can preserve domestic infrastructure while making linked transactions possible.
Reserved funds carry the transaction through failures
The technical report describes a process with a commitment boundary.
Before commitment, the operator arranges reservations across the relevant systems. If funds cannot be secured, it can cancel reservations already obtained. After commitment, it proceeds with settlement. If a participating system becomes unavailable during that phase, the operator retries the outstanding instruction when the system returns.
That sequence explains how the project's atomic, all-or-nothing outcome works. It depends on funds being reserved and on the operator managing the transaction through completion.
Researchers tested insufficient funds, invalid account identifiers and infrastructure outages. Across the selected scenarios, transactions either completed or were fully unwound without leaving individual payment legs to proceed independently.
Integration required different arrangements for each system. In Singapore's test environment, a dedicated bank-owned wallet managed under bank authorisation replicated funds reservation through an escrow arrangement.
This raises a practical design question. A coordinator that temporarily controls funds takes on responsibilities beyond passing instructions between ledgers. The reports examine how those functions affect operational risk and oversight.
The engineering lesson is that interoperability depends on the functions a system exposes, including reservation and transaction-status information. A shared description of the technology cannot substitute for those capabilities.
Netting changes the liquidity calculation
A second use case tested multilateral net FX settlement across the three environments.
Instead of settling every eligible payment separately, the prototype grouped non-urgent transactions into settlement cycles and calculated participants' net positions. Offsetting obligations can reduce the money needed to complete a batch.
The difficult part extended beyond arithmetic. The operator had to coordinate a common group of transactions that every participating system could accept while keeping the necessary liquidity reserved.
The reports identify a trade-off in assigning that work. If the operator calculates offsetting positions and constructs settlement instructions, its responsibilities can begin to resemble clearing functions. More support within the central bank systems could reduce that burden.
The policy workstream therefore considered access, oversight, settlement finality, governance and information sharing. Technical completion and legal finality require separate treatment, particularly when several jurisdictions are involved.
For the Bank of England, these findings feed into development of its planned 2028 service. They provide tested workflows and concrete integration problems for that programme to address.
Overview
Project Meridian FX APAC demonstrated coordinated FX settlement across central bank test systems using different technologies. It tested both individual currency exchanges and multilateral netting, with failure scenarios examining whether linked payments stayed intact.
The trial shows that synchronisation can connect account-based and tokenised settlement environments, provided they supply the necessary capabilities or adaptations. The Bank of England's implementation target is 2028.
