0x3ac2348f3ac2…3ac23492

ConfirmedInfrastructure663 vB51 sat/vB3 min decode

BoE-Led FX Trial Proved Atomic Settlement Across Three Ledgers

Project Meridian FX APAC settled FX atomically across three central bank ledgers — but Singapore's system forced the operator into an escrow wallet, a custody problem hiding in the detail.

Three Central Banks Settled FX Across Separate Ledgers, and One Had to Hold the Money
WitnessThree Central Banks Settled FX Across Separate Ledgers, and One Had to Hold the MoneyTumisu / pixabay

Outputs

  1. The BoE published Project Meridian FX APAC results on 13 November, run with MAS and the Bank of Thailand.

  2. Atomic PvP worked across three systems; Singapore's SGD Testnet required the operator to manage a dedicated escrow account.

  3. Only $5 trillion of $14 trillion in daily FX settlement (36%) clears through PvP, per the BIS 2025 Triennial Survey.

  4. Netting cuts 90 gross payment legs among ten participants to 10 movements, an 89% reduction.

  5. The BoE is committed to launching a live synchronisation service in 2028.

The Bank of England, the Monetary Authority of Singapore and the Bank of Thailand settled foreign exchange atomically across three separately designed central bank money systems, according to the Project Meridian FX APAC summary published on 13 November. The headline finding is that payment-versus-payment worked across an RTGS simulation, a tokenised wholesale ledger and a wholesale CBDC testnet. The detail that matters is that one of the three only worked because the operator held the cash.

The experiment ran in simulated environments and did not reproduce a live FX trade lifecycle. Even so, it answers a question that has stalled cross-border settlement design for years: banks do not need to converge on one shared platform to settle atomically. The cost of not converging lands somewhere else — on the synchronisation operator.

What did the operator actually do?

Partior built the operator prototype, and it stayed deliberately thin. Every transaction passes four stages, per the technical report: a participant submits a settlement instruction with currency, amount, counterparty, value date and a priority flag; the operator matches the two legs; each settlement system earmarks funds so they can't be reused unless the operator cancels; and only when every leg is secured does settlement execute, binding the all-or-nothing condition.

The reversible window is the design's useful part. Until the first settlement instruction executes, the operator can unwind everything. After the first leg executes, it can't. The report calls that fork the point of commitment and separates it from legal settlement finality, which the policy report handles.

Four exception scenarios tested the fork — insufficient funds, an invalid account identifier, both faults in one batch, and an infrastructure outage. Nothing was left half-settled. The report credits idempotency, queryable transaction status and reliable event delivery.

Why the escrow wallet is the real finding

The three environments were genuinely different. The BoE's Synchronisation Lab paired a simulated RTGS system with prototype earmarking. Thailand's Project SAN ran a token-based wholesale ledger with an interoperability layer. Singapore's SGD Testnet used a shared ledger where commercial bank wallets hold wholesale CBDC, with no central intermediary.

Singapore's environment could not earmark funds for a third party. So the operator replicated reservation with a wallet it controlled. "In MAS' test environment, the SO managed a dedicated escrow account to replicate funds reservation," the summary notes, adding the arrangement "is not an inherent feature of synchronisation." That sentence carries weight: an operator parking participant funds in its own wallet acquires custody exposure and a legal relationship to the money that the non-custodial baseline avoids. One of three test environments — a third of the test bed — needed the workaround.

The summary also states that "synchronisation worked across systems with materially different technologies and operating models."

Netting added pressure in the same direction. Multilateral net FX PvP worked across all three systems, but because the operator calculated offsetting positions, the report warns it "may begin to assume functions traditionally associated with a clearing house, rather than acting solely as a settlement co-ordinator." Ten participants trading bilaterally produce 90 gross payment legs; netting collapses that to 10 movements, an 89% reduction. One floated remedy: central bank systems perform the netting themselves.

What happens before 2028?

The stakes are large. Citing the BIS 2025 Triennial Survey, of roughly $14 trillion settled daily in April 2025, only $5 trillion (36%) cleared through PvP. Another $7.6 trillion (54%) used risk-reducing methods like pre-settlement netting, and more than $1.4 trillion (10%) settled gross and bilateral, fully exposed.

The BoE says the findings inform "its commitment to launch synchronisation in 2028." Unresolved gaps include non-overlapping operating hours, unmodelled earmark expiry and out-of-scope commercial bank money settlement. MAS and the Bank of Thailand made no equivalent implementation commitment, while the Bangko Sentral ng Pilipinas and the Reserve Bank of Australia observed the policy workstream — the likely source of the next round of jurisdictions.

via cryptonews.net (Source)

More from Daniel Okafor

Daniel Okafor

Show full bio

Correspondent covering people, launches and day-to-day practice in crypto and web3 industry at Mempool Brief.

1,221 articles