Cross-border payments are still slow, costly and opaque. They run through chains of correspondent banks, each adding fees, cut-off times and compliance checks. The tokenisation debate is often framed as banks against crypto. We think that framing misses what is happening. Banks are borrowing specific features from blockchain: programmability, atomic settlement and shared ledgers. They are applying them inside regulated, permissioned environments. For ASEAN banks, the question has moved from "whether" to "where first, and under what controls".
The problem the G20 asked us to fix
The Financial Stability Board coordinates the G20 cross-border payments programme. It set quantitative targets on cost, speed, access and transparency, to be met by 2027. Its progress reports have consistently found that progress is uneven and that the targets are at risk. That gap is the commercial case for new rails. Faster settlement does little if the surrounding compliance and messaging processes stay slow.
Tokenised deposits, not open stablecoins
Many banks prefer tokenised deposits to stablecoins. A tokenised deposit is a bank liability that stays on the bank's balance sheet. It keeps deposit insurance treatment, interest and supervisory oversight, and it can move on a shared ledger. Stablecoins issued by non-banks carry reserve, redemption and run risks that boards would have to underwrite.
The BIS's Project Agorá tests this model. It brings central banks and commercial banks together to explore tokenised commercial bank deposits alongside tokenised central bank reserves on a single platform. In a press release of 27 May 2026, the BIS said Agorá shows how tokenisation can improve wholesale cross-border payments, and that the work will advance to real-value testing. Real-value testing means live money, so operational and legal questions can no longer be deferred.
ASEAN's own building blocks
The region has been building a layered set of connections:
- Project Nexus (BIS Innovation Hub). It links domestic instant payment systems to one another through a standard gateway. Malaysia is among the founding participants, together with India, the Philippines, Singapore and Thailand. It does not need a ledger, which is a reminder that not every improvement requires tokenisation.
- ASEAN instant payment and QR connectivity. Bilateral links between Malaysia and its neighbours have already made retail and small-value cross-border payments routine. They show what regulators and banks can deliver when messaging standards and settlement arrangements are agreed.
- Project Mandala. It explored embedding compliance rules, including jurisdictional requirements, directly into cross-border transaction logic. Its message is that compliance can be automated rather than handled after the fact.
- Project Dunbar. The BIS, Bank Negara Malaysia, the Monetary Authority of Singapore and other central banks tested a shared multi-CBDC platform for settlement between banks. It proved the concept of a shared wholesale platform.
Malaysia and Singapore: sandbox first, then scale
Bank Negara Malaysia has taken a structured approach to digital assets and tokenisation. It uses a Digital Asset Innovation Hub and sandbox-style testing to examine real use cases, including tokenised deposits and other tokenised financial assets, under supervisory oversight. Our reading is that the direction is clear: controlled experimentation first, with use cases tied to real economic purposes.
The Monetary Authority of Singapore's Project Guardian works the same way. It brings banks and asset managers together to test tokenised assets and DeFi-style protocols in permissioned settings. Its main lesson is about governance. Participants can use public-chain technology if the participants, assets and protocols are controlled and the rules are enforceable.
Four questions boards must settle before scaling
- Compliance. Can the bank apply AML, sanctions and travel-rule obligations at the same level on a token rail as on SWIFT? Who is responsible when a counterparty's wallet or node fails a check?
- Liquidity. Tokenised settlement can free up trapped nostro balances. It can also create new demands for intraday funding and for 24/7 liquidity management. Treasury should model this before volumes arrive.
- Interoperability. Isolated networks repeat the fragmentation they were built to remove. The board should ask which standards, networks and central bank settlement assets the bank is committing to, and what the exit costs are.
- Governance and accountability. Smart contracts, validators and platform operators raise legal questions about settlement finality, liability and data. In our experience, directors need a clear view of who owns operational resilience and model risk for the new infrastructure. This should be settled before launch, not afterwards.
What an ASEAN bank should do in the next 12 months
- Pick two or three corridors with real flows. Choose those with high cost or delay, and set baselines for cost, speed and failure rates. Without baselines you cannot show that a pilot works.
- Join, don't build alone. Take part in regulator-led sandboxes and industry consortia. Prioritise those aligned with Nexus, Agorá-style work and the regional connectivity programmes.
- Map compliance onto the new rail. Involve compliance, legal and technology from day one. Test how screening and reporting will work on a ledger.
- Write a position on tokenised deposits versus stablecoins. Approve it at board level, including limits on counterparties and on any exposure to non-bank stablecoins.
- Prepare treasury and operations. Plan for 24/7 operation, intraday liquidity and new reconciliation processes.
- Set the success measures now. Link each pilot to the FSB's four dimensions of cost, speed, access and transparency, so results can be reported to the board in the same terms.
Our view
The shift from pilot to practice is real, but it is gradual and selective. Banks that treat tokenisation as a payments-efficiency tool, set within clear governance, will gain. Those that treat it as a technology showcase will not.