top of page
Search

Inside Project Agorá: What the BIS Tokenisation Milestone Actually Means for Cross-Border Payments

  • May 30
  • 6 min read

Updated: Jun 4

In late May 2026, the Bank for International Settlements published the prototype report for Project Agorá, its largest-ever Innovation Hub experiment. The headline most outlets ran with was that tokenisation "works." The more useful read is narrower and, in some ways, more interesting: a coalition of central banks proved a specific architecture can settle cross-border payments atomically across currencies, and then deliberately chose to keep the most sensitive part of the system under national control. That design choice tells you more about where institutional finance is heading than the speed numbers do.

This piece breaks down what Agorá is, what the prototype actually demonstrated, and why it matters, including for India.


What Project Agorá is


Project Agorá is a public-private collaboration convened jointly by the BIS and the Institute of International Finance (IIF). That framing matters: the IIF, a global association of financial institutions, is a co-convenor, not a bystander. The project brings together seven central banks representing major reserve currencies and more than 40 private financial institutions, including JPMorgan, HSBC, Deutsche Bank, Mastercard, UBS, and Swift itself.


The seven central banks are the Federal Reserve Bank of New York, the Bank of England, the Bank of Japan, the Swiss National Bank, the Bank of Korea, the Bank of Mexico, and the Banque de France, the last participating on behalf of the Eurosystem (which is why there is no separate ECB seat). The Bank of Canada joined the week the report was published, with more institutions expected to follow.


The project builds directly on the BIS's "unified ledger" concept and earlier experiments like Project Helvetia and Project mBridge. It also slots into the G20's cross-border payments roadmap, the multi-year effort launched in 2020 to make international payments faster and cheaper.


The problem it's trying to solve


Cross-border wholesale payments still mostly run on correspondent banking and messaging rails like Swift. A payment hops through a chain of intermediary banks, each running its own compliance checks sequentially, each reconciling its own ledgers. The result is the familiar set of pain points: multi-day settlement, high cost, opacity about where a payment is, and settlement risk.


Agorá targets four of these directly: speed (settlement completing in seconds once funds are locked), efficiency (running compliance checks in parallel rather than one after another), transparency (all parties tracking payment status in real time), and risk reduction through atomic settlement.


How it actually works, and the design choice that matters


The core mechanism is atomic settlement. Tokenised central bank reserves and tokenised commercial bank deposits sit on a shared platform, and a transaction either completes in full across every leg or it doesn't happen at all. There is no state where one side has paid and the other hasn't. This is what kills Herstatt risk, the danger that a counterparty fails after you've already sent your side of the payment. It's named after a German bank whose 1974 collapse left counterparties exposed exactly this way.


Compliance is handled cleverly, rather than each bank in the chain running anti-money-laundering, sanctions screening, and fraud checks one after another, the prototype runs these in parallel, then locks the funds just before settlement. Existing AML and sanctions frameworks are preserved, not bypassed. That is the point: the system is designed to make compliance more efficient, not lighter.


The prototype is not a single shared ledger holding everyone's money. It is a layered architecture. Central bank money stays under each central bank's domestic control, while an interoperable shared layer sits on top to let the tokenised assets transact across borders. This was, by the BIS's account, the most consequential early design decision, and the requirement for central banks to retain local control over their own currency is what drove the entire two-layer structure. The whole architecture exists specifically to avoid pooling central bank money on one shared platform, which is its defining feature.


What the prototype actually showed, and what it didn't


The BIS report is a qualitative assessment of a prototype. It found that tokenisation could address inefficiencies "in a safe and secure manner," that atomic multi-currency settlement is achievable across jurisdictions, and that the layered architecture preserves central bank autonomy. Participants judged that the nine priority friction points they had identified were materially or partially addressed.


What the report does not provide are hard numbers on cost savings or settlement-time reductions from live transactions, because no real money has moved yet. The prototype demonstrated that the design is sound and the mechanics work as intended. Proving the economic benefits at scale is what the next phase is for.


What happens next


Agorá now moves from simulation and proof-of-concept to real-value testing, meaning actual money moving across the system with selected currencies and institutions. The choice to add the Bank of Canada, a major liquid-currency central bank, suggests the testing will prioritise large, liquid corridors rather than exotic pairs.


Two things to keep front of mind. First, the BIS stresses the project remains experimental, and there is no timeline for production deployment. Second, this is not a crypto play. The BIS deliberately excluded private tokens and stablecoins from the architecture. The institutional message is pointed: the path to modernising international payments runs through tokenised central bank money, not stablecoins or crypto-native rails. The BIS reinforced this the same week by separately warning that stablecoins and exchange "earn" products can expose users to unsecured repayment risk.


The analytical takeaways


The compliance story is the real unlock. The instinctive worry about blockchain in finance is that it means less oversight. Agorá argues the reverse: parallel, automated, embedded compliance can make AML and sanctions screening more efficient than the sequential mess of correspondent banking. In a period of heightened geopolitical tension over sanctions enforcement, that is the feature that earns government trust, and it is probably more decisive for adoption than raw speed.


Sovereignty was the binding constraint, and it shaped everything. The layered architecture is not an engineering preference; it is a political requirement made technical. Central banks were never going to surrender control of their currency to a shared pool. Any future cross-border tokenisation standard will likely inherit this two-layer pattern for the same reason.


Concentration risk is the quiet danger. With the participant list skewing toward the largest global banks, there is a real question of whether smaller banks and emerging-market central banks get meaningful access or get designed out. The interoperability promise only matters if the on-ramps are genuinely open. This deserves watching as the standard solidifies.


The unresolved hard problems are legal, not technical. Cross-border legal recognition of tokenised assets, data privacy across jurisdictions, and interoperability between divergent national systems are not things a prototype solves. They are the work of years and treaties, not code.


The India angle


For an Indian readership, Agorá lands at a notably resonant moment. Days before the BIS report, on 26 May 2026, SEBI Chairman Tuhin Kanta Pandey announced at the CareEdge Debt Market Summit in Mumbai that SEBI has approved a pilot to tokenise corporate bonds using distributed ledger technology. The pilot targets four benefits that rhyme directly with Agorá's goals: shorter settlement cycles, better traceability, automated debt servicing, and improved transparency. Pandey indicated a limited-scale rollout first, with implementation expected to take roughly six to nine months.


For context, India's corporate bond market has grown from around ₹17.5 lakh crore at the end of FY15 to more than ₹59 lakh crore now, and FY26 debt issuance of ₹9.1 lakh crore was nearly double the equity markets' haul over the same period. SEBI's pilot is not meant to replace the existing bond framework but to test whether DLT can add an efficiency layer on top, which is, structurally, the same "augment, don't replace" philosophy Agorá embodies.


The distinction worth drawing, Agorá is wholesale cross-border payments between central and commercial banks, while SEBI's pilot is domestic capital-market infrastructure for corporate debt. Different layers of the financial stack. But both reflect the same global pattern, regulators using controlled tokenisation pilots to modernise plumbing while keeping safeguards intact. India is not a participant in Agorá, but it is clearly reading from the same playbook.


Conclusion


Project Agorá did not prove tokenisation is the future of money. It proved something more specific and more credible: that a carefully bounded design, atomic settlement, parallel compliance, and a sovereignty-preserving layered architecture, can work as intended for cross-border wholesale payments. The next phase, moving real value, is where the economic case gets tested. For now, the most important signal is not the technology. It is that seven major central banks, more than 40 of the world's largest banks, and a 40-plus-institution private coalition agreed on an architecture and chose to keep control where it has always been. That alignment, more than any speed metric, is what makes this worth watching.


Sources:

Comments

Stay connected for updates and insights.

Get in Touch

Write us at
contact@projectsutra.org

  • LinkedIn
  • X
  • Instagram

Privacy Policy

bottom of page