The Interoperability Imperative

I have spent the better part of two decades inside payment schemes. As CPO at Interac and in roles at Klarna and Nexi, I watched the same conversation happen in different companies with different accents: "Our domestic success proves our model works. Cross-border is the next growth chapter. We will figure out the economics when we get there."

Every time I heard that sentence, I knew the scheme was about to stall.

The economics of cross-border payment processing are not a mystery. They are governed by participant incentives: if every institution in the chain has a profitable reason to process the transaction, volume scales; if one institution does not, volume stalls regardless of customer experience.

The A2A payment schemes that have scaled internationally PayNow-PromptPay, UPI's bilateral corridors[4] did so by internalizing this lesson. The ones that have stalled most prominently, Project P27 in the Nordic region[3] stalled precisely because they treated cross-border expansion as an engineering problem when it is fundamentally an economics problem.

This paper is directed at scheme CEOs and strategy leads who are evaluating cross-border expansion in the next six to eighteen months. The question is not whether to pursue cross-border. The question is which architecture to choose, and how to structure the participant economics so that volume actually flows once the pipe is built. That decision needs to happen now.

---

Section 1: The Localization Fallacy

Swish won Sweden by being Swedish. BLIK won Poland by being Polish. iDEAL won the Netherlands by being Dutch. Every major domestic A2A scheme built its moat through localization: UX that fits how people in a specific country think about money, bank integrations that leverage existing domestic infrastructure, and features purpose-built for local regulatory and cultural contexts.

That localization advantage is real. I am not arguing otherwise. BLIK's mobile-first QR experience is genuinely superior to pulling out a Visa card for a domestic Polish transaction. The speed, the familiarity, the trust: these are hard to replicate, and they are the reason A2A schemes have captured meaningful market share in their home markets against card incumbents with vastly more capital.

The localization fallacy is the belief that what worked domestically can be extended cross-border by building more of the same.

It cannot and the reason is structural.

A domestic payment transaction involves a small number of known participants: a sending bank, a receiving bank, and the scheme in the middle. The regulatory environment is uniform. The AML framework is shared. The FX rate is 1.0. The economic model is simple: the scheme charges a small clearing fee, the banks earn on account relationships, and every participant has a clear and predictable revenue model.

A cross-border payment transaction involves an entirely different set of problems. The sending bank and receiving bank are in different regulatory jurisdictions with different AML obligations. The transaction crosses at least one currency boundary, introducing FX risk, FX liquidity requirements, and FX economics that must be allocated across multiple parties. The AML liability question: who is responsible for screening, and what happens when a transaction fails post-clearing does not have a domestic analogue.

These are not engineering problems that better APIs will solve. They are economic design problems. Who holds the FX risk? Who earns the FX margin? What does the sending bank receive in exchange for exposing itself to cross-border compliance obligations? What does the receiving bank charge for accepting a foreign-originated credit? What is the scheme's take?

The card schemes answered these questions forty years ago. Their answer is what funded the Visa and Mastercard networks: a multi-layer fee structure in which the card scheme, the issuing bank, the acquiring bank, and the FX provider all earn a defined margin on every cross-border transaction. The economics are not favorable to consumers. Cross-border card fees are among the highest-margin products in retail financial services but they are extremely favorable to participants. Every bank in the Visa network has a compelling economic reason to accept a Visa card from another country. The economics are designed to make that true.

A2A schemes that launch cross-border without having answered the equivalent questions will build infrastructure that sits idle. No bank will route volume through a cross-border A2A corridor if the economics of doing so are worse than routing through Visa Direct. The pipe gets built. The volume does not follow.

---

Section 2: The Economics That Determine Whether Volume Flows

Cross-border A2A does not fail on messaging formats. It fails when one participant in the chain has no economic reason to route the transaction. In practice, three variables decide that outcome in real time: the FX quote, the fee split, and the liability allocation.

FX is the central lever.

If the spread is not allocated so each participant can cover compliance, operations, and balance-sheet cost, volume will not scale. This is why technically functional corridors can remain commercially thin: the rail is live, but incentives are mispriced.

A2A does have a structural cost-stack advantage, but the claim has to be precise. A transfer priced at 0.5% does not produce more gross revenue than one priced at 3.0%. The opportunity is that lower operating complexity can distribute a smaller pool more efficiently while still giving customers a better price.

That only works when economics are designed before launch: who earns which part of FX, what sending and receiving banks are paid for compliance load, and how sanctions exceptions and failed payments are handled.

The recurring failure mode and P27 remains the warning is to build infrastructure first and economics second. Programs that do this announce integration milestones and then stall when banks conclude per-transaction returns are weaker than incumbent routes.

This is where most schemes still under-resource the work. Corridor economics design is not a side workstream. It is the workstream.

---

Section 3: Five Corridors, Five Lessons

The empirical record on A2A cross-border is now sufficient to draw strategic conclusions. Here are the five cases that every scheme operator should have internalized.

P27: The Centralized Approach That Stalled

Project P27 was the pan-Nordic multi-currency clearing initiative backed by six of the largest Nordic banks.[3] The ambition was correct: Nordic consumers and businesses make enormous volumes of cross-border payments within the region, and the existing infrastructure primarily correspondent banking and SWIFT was expensive, slow, and opaque.

The approach was to build a single, centralized multi-currency clearing scheme that would internalize all of the cross-border and cross-currency complexity. One clearing house, four currencies (SEK, DKK, NOK, EUR), governed jointly by the participating banks.

The initiative was wound down in 2023.[3] The reasons are instructive.

The centralized architecture required participating banks to make substantial investments in legacy system overhauls to connect to the new clearing infrastructure. Those investments were not justified by the per-transaction economics of the scheme, because the per-transaction fee model did not generate enough revenue to make the capital expenditure worthwhile for individual banks. The governance complexity of aligning six major banks across four jurisdictions on technical standards, regulatory interpretation, and commercial terms added years to the timeline. And the competitive dynamics between participating banks each of whom was also a competitor made commercial agreement on participant economics structurally difficult.

P27 did not fail because the concept was wrong. Pan-Nordic cross-border clearing is a real market opportunity. P27 failed because the centralized approach created an economic structure in which the cost of participation was front-loaded system overhaul, governance investment, compliance build and the revenue was back-loaded: transaction fees that would only be meaningful at scale. No individual bank had a compelling reason to absorb that imbalance. The collective action problem was never solved.[3]

Lesson for scheme operators: Centralized, heavy-infrastructure approaches to cross-border clearing require participant economics that justify the front-loaded investment cost. If you cannot show a participating bank a credible model for when their transaction volume through the new scheme covers their integration and compliance costs, the scheme will not scale.

PayNow-PromptPay: The Bilateral Bridge That Scaled

The Singapore-Thailand linkage launched in April 2021 as the world's first bilateral real-time payment linkage between two sovereign A2A schemes.[5][6] It is the clearest proof point available that A2A cross-border works when the economics are designed correctly.

The architecture is deliberately simple. Designated FX liquidity providers specific banks in each market are responsible for quoting pre-transaction FX rates and funding settlement. Transactions are capped at SGD 1,000 per day, which limits intraday liquidity exposure for the FX providers and reduces AML screening complexity.[5][6] The consumer sees the FX rate before initiating the transaction, which means the economic model is transparent and auditable.

The fee structure was designed to undercut traditional remittance corridors the Thailand-Singapore corridor is one of the highest-volume remittance routes in Southeast Asia, traditionally dominated by Western Union and money transfer operators charging 2–5% while leaving enough margin for participating banks to cover their API integration and ongoing compliance costs.

The corridor has remained an important reference point because it operationalized real-time bilateral linkage with transparent pre-transaction FX quotation and capped exposure from day one.[5][6] It is frequently cited in regional cross-border discussions as a practical proof that A2A interoperability can scale when participant incentives are explicit.[7]

What the PayNow-PromptPay team got right was sequence: they solved the economics first, built the simplest possible technical architecture second, and launched with a narrow but economically defensible use case P2P remittance before expanding to merchant payments.

Lesson for scheme operators: The bilateral bridge model works when (1) the corridor has high existing remittance volume to capture from incumbents, (2) the FX liquidity provider role is clearly allocated to specific banks with transparent pre-transaction rates, and (3) transaction caps manage the intraday liquidity and AML risk that would otherwise prevent bank participation.

UPI International: The Asymmetric Acquirer Overlay

India's NPCI has taken a deliberately different approach to cross-border expansion.[4] Rather than building bilateral scheme-to-scheme linkages for all use cases, NPCI has pursued a two-track strategy: bilateral linkages for specific high-volume corridors UPI to PayNow, UPI to UAE's Aani and direct merchant acquirer partnerships for tourist and diaspora use cases.

The acquirer overlay strategy signing partnerships with local payment processors such as Lyra in France to enable UPI QR code acceptance at European merchants is an asymmetric move.[4] It exports the consumer experience for Indian travelers abroad without requiring a bilateral agreement with a European A2A scheme or solving the full participant economics stack for a new corridor. The economics rely on traditional correspondent banking and acquiring fees for the backend, which is not optimal, but it gets UPI acceptance points established quickly and at low scheme-level cost.

UPI's international strategy is explicitly about geographic footprint and consumer experience capture first, with economics optimization as a second-order problem. This works because UPI has a specific structural advantage: a massive, homogeneous user base of Indian travelers and diaspora with strong demand for paying abroad in a familiar interface. Not every scheme has that demand anchor.

Lesson for scheme operators: The acquirer overlay is a viable fast-start strategy for schemes whose domestic users travel heavily or send remittances to specific corridors. It establishes consumer UX without solving full bilateral scheme economics. However, it does not create a defensible competitive position the backend economics remain subject to correspondent banking pricing, and Visa Direct can offer the same acquirer economics without the scheme integration overhead.

EPC OCT Inst: Standardizing Gateway Economics

The European Payments Council's One-Leg Out Instant Credit Transfer scheme (OCT Inst) addresses a specific but important gap in European cross-border A2A infrastructure: the opaque and slow economics of cross-currency instant payments entering or leaving the Eurozone.

The scheme standardizes message requirements (ISO 20022-based) and codifies timing and participant obligations for the Euro leg of one-leg-out instant transfers, including explicit handling of amount, exchange-rate, and charge-bearer data elements.[8] This standardization matters because it lowers integration friction for gateway banks and creates a shared operational baseline across participants.

OCT Inst does not itself set corridor FX pricing that remains commercial but it creates a common execution and data framework that makes FX terms more comparable and contestable than fully bespoke bilateral setups.[8]

Lesson for scheme operators: Standardization of message formats and SLAs creates the conditions for competitive FX economics without requiring scheme-level FX infrastructure. If your scheme operates in or adjacent to the EEA, OCT Inst compliance should be on your technical roadmap not because it solves everything, but because it creates a common interface through which competitive FX liquidity providers can operate.

BIS Project Nexus: The Multilateral Model for the Next Decade

Project Nexus is the Bank for International Settlements' answer to the bilateral linkage problem. The mathematics are straightforward: connecting 20 national A2A schemes through bilateral agreements requires 190 separate linkages. Each linkage requires its own technical integration, FX liquidity agreement, and regulatory framework. This does not scale.

Nexus solves this with a standardized multilateral gateway. Participating schemes connect to the Nexus hub once, through a standardized API and rulebook based on ISO 20022. The hub handles routing to any other participating scheme. Critically, Nexus incorporates a built-in FX marketplace in which financial institutions bid to provide the best rate for each transaction across specific corridors. The scheme does not take on FX risk it routes transactions through whichever liquidity provider is offering the best pre-cleared rate for that corridor at that moment.

BIS materials describe Nexus as having completed its phase-three blueprint and moving toward live implementation, with central-bank and IPS workstreams involving Malaysia, the Philippines, Singapore, and Thailand, joined by India in the next phase; Indonesia is listed as a special observer.[7] This should be distinguished from separate bilateral linkage programs in the region.

Nexus is the most architecturally elegant solution to the cross-border A2A problem I am aware of. It solves participant economics through marketplace competition rather than bilateral negotiation. It solves the scaling problem through standardization rather than infrastructure replication. And it solves the FX risk problem by allocating it to specialist liquidity providers who price it into their bids rather than forcing it onto scheme operators or participating banks.

Lesson for scheme operators: Nexus compatibility ISO 20022 compliance and alignment with the Nexus rulebook should be on every scheme's technical roadmap, regardless of which primary cross-border strategy you choose. A scheme that is already ISO 20022-compliant and Nexus-ready can participate in any bilateral corridor that connects to the Nexus hub, without re-integrating for each new corridor.

---

Section 4: The Card Scheme Threat: Visa and Mastercard Are Not Standing Still

I want to address directly what every scheme CEO is already thinking when they read the case studies above.

As of Visa's 2025 investor disclosures, Visa Direct reports 11B+ endpoints across cards, accounts, and wallets in 195+ countries and territories, and support for 150+ currencies.[1]

As of Mastercard's published Move metrics (valid as of October 2025), Mastercard Move reaches 200+ countries and territories, 150+ currencies, approximately 17 billion endpoints, and more than 95% of the world's banked population.[2]

These are not roadmap items. These are live products, with active sales teams, being sold right now to the same banks that participate in A2A schemes.

The historical parallel that I find instructive is the European mobile operator consortium that launched "joyn" an attempt to build an interoperable messaging platform to compete with WhatsApp. The strategic logic was sound. The mobile operators collectively had 100% of the European mobile user base. An interoperable messaging standard that crossed operator boundaries should have been a winning proposition. Every argument for cooperation between competitors was logically valid. They failed because bilateral negotiation between competing mobile operators was structurally too slow too many governance meetings, too many commercial disagreements about data sharing and revenue allocation and WhatsApp moved faster than the consortium could resolve its internal conflicts. By the time joyn launched, network effects were already irreversible.

A2A schemes face the same structural risk with Visa Direct. The card schemes already have global bank relationships, established FX infrastructure, and broad regulatory operating coverage. They are layering new cross-border use cases onto existing rails, which can lower their marginal cost of corridor expansion relative to schemes starting from zero.[1][2]

The strongest counter-argument to this threat assessment is: "Visa Direct and Mastercard Move carry card scheme economics. The fees are high. A2A schemes can undercut them on price." This is correct, and it is the core of the A2A strategic opportunity. Cross-border card pricing is often embedded in FX spreads and transfer fees that remain expensive for end users in many corridors. An A2A scheme that solves participant economics can offer lower end-customer pricing while still funding a viable participant margin stack. The cost advantage is real.

But that cost advantage only materializes if the A2A scheme has actually built the cross-border infrastructure and designed the participant economics. A scheme that waits three more years to begin that work will find that the banks it needs as FX liquidity providers are already committed to Visa Direct agreements with favorable commercial terms. The window for building competitive cross-border A2A infrastructure is not permanently open. Every month that passes, the card schemes deepen their bank relationships and corridor coverage, raising the entry cost for an A2A scheme that wants to compete in those corridors.

Scheme operators who do not have a defined cross-border strategy in 2026 will be routing international volume through card scheme rails by 2028 and paying card scheme economics to do so. That outcome would undermine the cost advantage that justified the A2A scheme's domestic existence in the first place.

---

Section 5: Three Strategic Paths for Scheme Operators

Based on the evidence from the five corridors above and my operational experience in scheme economics, I see three viable strategic paths for A2A scheme operators evaluating cross-border expansion. These are not mutually exclusive UPI pursues both the bilateral bridge and the acquirer overlay but they have different cost structures, timelines, and competitive implications.

Path 1: The Bilateral Bridge (PayNow-PromptPay Model)

Best suited for: Schemes with one or two high-volume corridors with identifiable remittance or tourist demand.

The bilateral bridge requires: (1) a designated FX liquidity provider bank in each market, willing to quote pre-transaction rates and absorb intraday liquidity exposure; (2) an agreed transaction cap to manage AML risk; (3) a transparent fee structure that undercuts incumbent remittance operators while leaving margin for participating banks; and (4) bilateral technical integration between the two scheme clearing systems.

Timeline (operator estimate): 18–36 months from negotiation start to live transactions at scale, based on the PayNow-PromptPay experience and typical bank integration cycles.

The key variable is whether designated liquidity-provider banks already have correspondent and settlement relationships in place across both markets. Timelines trend toward the lower end when those relationships and AML/KYC operating lanes already exist, and toward the upper end when they must be built from scratch.

Cost structure: Moderate upfront investment technical integration, legal framework, compliance with ongoing FX liquidity costs borne by the designated liquidity provider banks.

Competitive implication: Creates a defensible position in the specific corridor. Does not create a platform for rapid multi-corridor expansion without additional bilateral negotiations.

The bilateral bridge is the right first move for most schemes, particularly those that have identified one corridor with enough volume to build a credible business case for participating banks. The PayNow-PromptPay corridor was specifically chosen because the Singapore-Thailand remittance corridor was already large, established, and expensive characteristics that made the business case for bank participation straightforward.

Path 2: The Acquirer Overlay (UPI Model)

Best suited for: Schemes whose domestic users travel heavily or send remittances to dispersed global destinations that do not lend themselves to bilateral linkages.

The acquirer overlay does not require bilateral scheme-to-scheme agreement. The scheme partners with local payment processors in target markets to enable acceptance of the scheme's QR codes or mobile credentials at merchant point-of-sale. The backend economics rely on acquiring and correspondent banking infrastructure.

Timeline (operator estimate): 6–18 months per market, significantly faster than the bilateral bridge.

The key variable is acquirer readiness: existing processor integrations, merchant onboarding rails, and local regulatory approvals determine whether rollout is a rapid enablement program or a longer market-entry project.

Cost structure: Lower upfront investment; ongoing economics subject to acquiring and correspondent banking fees that the scheme does not control.

Competitive implication: Establishes consumer UX and acceptance network quickly, but does not create a structurally cheaper cross-border product than card schemes. The backend economics will converge toward card scheme pricing unless the scheme subsequently builds bilateral or multilateral clearing infrastructure.

The acquirer overlay is a strong fast-start play for schemes whose domestic users are actively traveling and encountering demand for the product in real time. Its limitation is that it is a consumer experience play, not an economics play and without the economics, a card scheme competitor can replicate or undercut it.

Path 3: The Multilateral Gateway (Nexus/OCT Inst Model)

Best suited for: Schemes seeking broad regional relevance across multiple corridors simultaneously.

The multilateral gateway requires: (1) ISO 20022 message format compliance; (2) alignment with the gateway's technical rulebook; and (3) connection to the central hub infrastructure. In the Nexus model, this connection gives the scheme access to every other participating scheme in the network, with FX economics handled through the marketplace.

Timeline (operator estimate): ISO 20022 compliance is the long pole typically 24–36 months for a scheme not already on ISO 20022 messaging. Once compliant, connection to a Nexus hub is an incremental project.

The key variable is legacy migration scope: schemes with modernized ISO 20022-adjacent messaging can compress the program, while schemes that must replace core message orchestration, exception handling, and certification tooling will land at the longer end.

Cost structure: Front-loaded investment in ISO 20022 compliance; low marginal cost per additional corridor once connected.

Competitive implication: Creates the most scalable architecture for multi-corridor expansion. The strongest long-term competitive position for schemes that want broad regional relevance rather than ownership of specific corridors.

Evaluating All Three Against the Card Scheme Alternative

Before committing to any of these paths, scheme operators should run one additional calculation: what is the cost of doing nothing, and routing cross-border volume through Visa Direct or Mastercard Move?

I am not being provocative. For some schemes, in some corridors, the card scheme option may be economically rational in the short term particularly if the scheme's cross-border volume is too small to justify the investment in a bilateral bridge. But that calculation needs to be made explicitly, with clear visibility into what Visa Direct or Mastercard Move charges for cross-border routing, and what the long-term competitive implications are of becoming dependent on card scheme infrastructure for international volume.

A scheme that routes the majority of its cross-border volume through Visa Direct for five years will find it structurally difficult to build competing infrastructure afterward. The banks will be in long-term Visa Direct agreements. The consumer habit will be formed. The commercial case for independent cross-border infrastructure will be substantially weaker. This is not a theoretical risk. In my judgment, based on operator conversations and observed go-to-market patterns, several second-tier A2A schemes are already on this trajectory without having made the choice consciously.

---

Conclusion: The Window Is Closing, and the Incumbents Are Already Inside

The interoperability imperative for A2A payment schemes is not a future challenge. It is a present one, compressing in real time.

BIS Project Nexus is moving from pilot to implementation.[7] The ASEAN corridors are being locked in. OCT Inst is standardizing the Euro-zone gateway.[8] UPI continues to sign bilateral and acceptance agreements across Asia, the Middle East, and Europe.[4] Every corridor that gets locked into a bilateral or multilateral architecture in the next 24 months becomes substantially more expensive for a late entrant to displace.

And the card schemes are already inside the building: Visa Direct reports 11B+ endpoints globally, while Mastercard Move reports reach to over 95% of the world's banked population and roughly 17B endpoints.[1][2] Both continue to expand cross-border money-movement capabilities at global scale.

The question I put to scheme CEOs is this: in three years, do you want to be running cross-border transactions on your own economics, or paying Visa for access to their network? That decision is being made right now not by the choice to start or not start a cross-border program, but by the choice to prioritize or deprioritize it relative to domestic roadmap work.

I have worked inside payment schemes. The economics of cross-border are not a mystery. They are well understood by the card networks and, in my experience, consistently underestimated by A2A scheme operators. The PayNow-PromptPay team proved that the problem is solvable. BIS Nexus is building the infrastructure that will make it scalable. The operational expertise required to design participant economics that make volume flow is available.

What is required is the decision to prioritize this work before the window closes.

Onwards.

---

Ted Bowman is the Founder of Outside Context Labs, an advisory firm specializing in payment scheme strategy, cross-border payments economics, and fintech growth. He has previously served as Chief Product Officer at Interac and in product and strategy roles at Klarna, Nexi/Nets and other fintechs.

---

Sources

1. Visa, 2025 Investor Day Consolidated (slide showing Visa Direct footprint: 11B+ endpoints, 195+ countries/territories, 150+ currencies), February 2025.
2. Mastercard, Mastercard Move product page (metrics valid as of October 2025: 200+ countries/territories, 150+ currencies, ~17B endpoints, >95% of world's banked population).
3. P27 Nordic Payments Platform, public governance updates and 2023 wind-down communications from participating banks and program entities.
4. NPCI International and partner announcements covering UPI-PayNow linkage, UPI-Aani linkage intent, and international merchant acceptance partnerships (including France acquirer integrations).
5. Monetary Authority of Singapore and Bank of Thailand joint launch communications on PayNow-PromptPay linkage (April 2021).
6. Bank of Thailand and Monetary Authority of Singapore updates on PayNow-PromptPay transaction cap and operating model.
7. BIS press release, Project Nexus completes comprehensive blueprint and prepares for work towards live implementation, July 1, 2024.
8. European Payments Council, 2025 OCT Inst Scheme Rulebook (EPC158-22 v1.0, issued November 28, 2024).