The G20 set end-2027 targets to make international payments faster, cheaper, more transparent and more accessible. With the deadline approaching, the rails have accelerated - but the user experience still depends on the slowest bank, rulebook, FX market and compliance process in the chain.
Global Banking & Finance Review | Research article | Updated 11 August 2026
International payments have spent the better part of a decade being described as ripe for disruption. Domestic transfers can now move in seconds in many markets, smartphones have compressed the customer experience to a few taps, and new payment networks promise near-continuous settlement. Yet sending money across a border can still produce an old-fashioned result: uncertain fees, opaque foreign-exchange spreads, manual compliance reviews and a beneficiary who waits far longer than the underlying payment message took to travel.
That gap matters more now because the policy clock is running. The Financial Stability Board (FSB) targets for enhancing cross-border payments set most of the G20 outcome targets for the end of 2027. For retail cross-border payments, the global average cost is meant to be no more than 1%, with no corridor above 3%; 75% should make funds available to the recipient within one hour, with the rest within one business day. Remittances have a similar speed ambition, while their cost target follows the UN Sustainable Development Goal of a global average below 3% by 2030 and no corridor above 5%.
The uncomfortable reality is that the deadline is approaching faster than the outcomes. The FSB said in its 2025 consolidated progress report that global indicators had improved only slightly since the first KPI calculations in 2023 and that satisfactory global improvement was unlikely to arrive on the original 2027 timetable. In March 2026, the FSB moved the programme into a more implementation-heavy phase, asking authorities for jurisdictional and regional action plans and placing greater emphasis on public-private delivery rather than additional high-level policy design.
The deadline was always about outcomes, not faster messaging
It is tempting to judge cross-border payments by the fastest segment in the chain. That produces impressive numbers. Swift says that around 75% of payments on its network reach beneficiary banks within ten minutes, many in seconds. But the G20 target is end-to-end: the customer must actually have the money. A message arriving at the beneficiary bank is not the same thing as funds being credited to the beneficiary account.
This distinction explains much of the apparent contradiction between rapid infrastructure progress and stubborn customer frustration. A payment can traverse the international messaging leg quickly and then stall at the receiving institution because of local operating hours, sanctions screening, missing beneficiary data, exchange-control rules, liquidity management, manual exception handling or legacy core-banking processes. Swift itself highlights this last-mile problem, noting that local processes can extend a payment that moved quickly between banks into an experience measured in hours or days.
The evidence therefore does not show that nothing has changed. It shows that the bottleneck has moved. International messaging and correspondent-bank processing are materially faster than they were. The remaining delays are increasingly concentrated in the parts of the chain that are hardest to standardise globally: the point where a transaction meets domestic regulation, local data requirements, local banking systems and the recipient account.
The global scorecard: faster in places, still expensive in aggregate
The cost picture is more sobering. The European Central Bank, drawing on FSB data, reported in 2026 that for nearly one-third of the cross-border business payments in its dataset, costs exceeded 3% of the transaction value, while only about 40% of international business-to-business payments were settled within one working day. The ECB also noted that the global provision of correspondent banking services had declined by roughly 20% compared with the mid-2000s, a trend that can reduce choice in lightly served corridors. ECB Economic Bulletin analysis also finds that interlinking fast-payment systems is associated with higher bilateral trade, illustrating why payment efficiency is an economic infrastructure issue rather than merely a bank-service issue.
Retail transfers show the same unevenness. The Eurosystem comprehensive payments strategy says that nearly one-fifth of global retail payment corridors in 2025 still exceeded the G20 maximum-cost ambition of 3%, against a target global average of 1%. For migrants sending smaller amounts home, the World Bank Remittance Prices Worldwide programme reported a global average cost of 6.36% for sending remittances in its third-quarter 2025 data - more than twice the 3% long-term global objective.
Those averages hide enormous corridor-level variation. Competitive digital corridors can be cheap and fast, while low-volume routes with weak local competition, cash-heavy distribution or currency controls can remain expensive. This is one reason global targets are difficult: a handful of world-class corridors cannot compensate for structurally weak ones when the policy ambition is broad accessibility and consistently low cost.
Why domestic instant payments did not automatically become global instant payments
Domestic instant-payment systems solved a comparatively bounded problem. Participants operate under one legal framework, one currency regime, one settlement asset and a relatively coherent set of data and consumer-protection rules. Cross-border payments add foreign exchange, multiple legal systems, sanctions regimes, anti-money-laundering obligations, data-localisation rules, different operating hours and sometimes several intermediaries.
The result is a system of fast domestic islands with incomplete bridges. The BIS Committee on Payments and Market Infrastructures (CPMI) said in its May 2026 monitoring survey covering 82 jurisdictions that expanded payment-system access, longer operating hours and interoperability by design are foundational to better cross-border services. It also highlighted ISO 20022 alignment and standardised API frameworks as ways to reduce inefficiency. The important word is alignment: introducing a modern format in one market does little if every market implements it differently.
This is why the cross-border payments problem has proved resistant to simple technological narratives. A faster rail can reduce transmission time, but it does not by itself harmonise legal liability, sanctions obligations, customer-identification rules or FX liquidity. The infrastructure can be instant while the institutional process around it remains sequential.
ISO 20022 is becoming the unglamorous centre of the reform
Few payment reforms sound less dramatic than message standardisation, but richer and more consistent data may do more for the 2027 agenda than another consumer-facing app. The CPMI updated its harmonised ISO 20022 data requirements in February 2026, with adoption expected by the end of 2027. The goal is not simply to make messages more detailed. It is to ensure that banks and payment systems carry consistent structured information across the chain so that screening, reconciliation, repair and exception management can be automated rather than repeatedly reconstructed by each intermediary.
Poor payment data creates real cost. A truncated beneficiary name, inconsistent address fields or unstructured remittance information can trigger manual investigation, false-positive compliance alerts and requests for clarification. Those interventions are individually rational for banks managing legal risk, yet collectively they make the system slower and more expensive. Standardisation is therefore an operating-model reform disguised as a data project.
There is also a limit to what data harmonisation can achieve. Banks will still apply different risk appetites, national authorities will still interpret some requirements differently, and privacy or localisation rules can restrict data movement. ISO 20022 can reduce avoidable friction; it cannot erase legitimate differences in law and supervision.
Compliance is not a removable friction
The industry often speaks about removing friction, but some friction exists because the financial system is expected to stop illicit money, sanctions evasion, fraud and misdirected payments. The relevant question is not whether to remove controls, but whether the same control can be executed with better data, clearer responsibility and more automation.
That is the direction of travel in the Financial Action Task Force revision of Recommendation 16. Agreed in June 2025, the changes standardise information requirements for certain peer-to-peer cross-border payments above USD/EUR 1,000, clarify responsibilities across the payment chain and require technologies that help prevent fraud and error. The changes are due to take effect by the end of 2030, well after the G20 2027 deadline. That timing itself illustrates the coordination challenge: important global rule changes operate on a slower calendar than infrastructure innovation.
For banks, this means the future of faster payments is inseparable from the future of financial-crime operations. Real-time settlement leaves less time for manual intervention. Screening, name matching, beneficiary verification and transaction monitoring need to become more precise before institutions can safely remove queues and cut-offs. Faster payments without faster risk decisions can simply move fraud faster.
Interlinking fast-payment systems may be the most practical bridge
One of the strongest alternatives to rebuilding global payments from scratch is to connect domestic systems that already work. The logic is simple: if two countries already have low-cost instant-payment rails, linking them can shorten the correspondent chain and allow customers to benefit from infrastructure that has already been paid for domestically.
The BIS Innovation Hub developed Project Nexus as a standardised model for connecting instant-payment systems, with the aim of enabling most cross-border payments to reach recipients in under 60 seconds. In Europe, the ECB is using TIPS to develop cross-currency and external fast-payment links. Its cross-currency service for euro, Swedish krona and Danish krone was implemented in 2025, while work is progressing on links with systems including India's UPI and other international arrangements.
The case for interlinking is strengthened by the ECB research finding that countries with connected fast-payment systems trade about 4% more with one another on average, after controlling for other factors. That is evidence of an association derived from econometric analysis, not proof that every new link will create the same gain. The inference for policymakers is nevertheless important: payment interoperability can have effects beyond consumer convenience, especially where existing cross-border payment costs are high.
The counterargument is scale and governance. Bilateral links can become a web of bespoke connections; multilateral hubs can reduce technical duplication but introduce common rulebooks, shared governance and concentration questions. Interlinking also works best where participating systems have compatible access rules, operating models and compliance standards. The technology is only one layer of the agreement.
Correspondent banking is changing, not disappearing
New rails are often framed as replacements for correspondent banking. That is plausible in selected retail corridors, but too sweeping as a global forecast. Correspondent banks still provide deep currency liquidity, credit, treasury services, sanctions expertise and reach into markets that are unlikely to build direct system links with every trading partner. Large-value corporate payments also require features that retail instant-payment systems may not provide, including complex FX execution, liquidity facilities and richer transaction services.
The more likely architecture is hybrid. Correspondent chains become shorter and more data-rich; fast-payment links handle a growing share of retail and SME flows; specialist fintechs compete aggressively on customer experience and FX; and tokenised or blockchain-based settlement is introduced where it solves specific coordination problems. That is an inference from the direction of current projects, not a settled end-state.
Swift is itself evidence of that hybrid path. Rather than waiting to be displaced, the network has pushed ISO 20022 adoption, payment tracking, pre-validation and new retail service standards. In March 2026, Swift said more than 25 banks would participate in a retail payments framework designed to provide cost certainty, full-value delivery and end-to-end traceability across selected corridors. At the same time, Swift is exploring blockchain-based infrastructure for 24/7 cross-border payments. The incumbent network is therefore trying to improve the present architecture while creating options for a different future one.
Why cost is harder to fix than speed
Speed can often be improved by engineering: longer operating hours, faster clearing, better routing and automated exception handling. Cost is more political and commercial. A cross-border price contains not only technology costs but also FX margins, liquidity costs, regulatory overhead, fraud losses, customer-acquisition expense, agent commissions, capital use and the economics of serving a particular corridor.
That is why some providers can make a payment technically instant without making it cheap. A customer may receive funds in seconds but still pay through a wide exchange-rate spread. Conversely, a bank may offer low explicit fees while recovering economics through FX. The G20 transparency target matters because competition cannot work well if customers cannot see the full price.
Competition is therefore as important as infrastructure. Expanded access for non-banks, interoperable systems and common APIs can lower barriers to entry, but regulators must balance those benefits against operational resilience and financial-crime risk. A market with more participants is not automatically a market with lower all-in costs if access to FX liquidity or last-mile distribution remains concentrated.
Stablecoins and tokenisation: catalyst, alternative or distraction?
The slow progress of conventional reform has created space for a more radical argument: if existing cross-border chains are structurally inefficient, why not route around them using tokenised deposits, stablecoins or shared ledgers? The attraction is obvious. Digital settlement assets can move around the clock, support programmable workflows and reduce the need for sequential reconciliation across separate systems.
But new settlement technology does not eliminate the core questions that slowed the old system. Users still need trusted conversion between currencies, legal finality, identity and sanctions controls, safeguarding of funds, dispute processes and regulated access points into the real economy. Tokenisation may compress the chain; it does not make governance optional. The Eurosystem's 2026 strategy explicitly takes a two-track view, improving existing infrastructure while also exploring tokenised settlement assets rather than treating the two as mutually exclusive.
The practical question is therefore not which rail wins in the abstract. It is which architecture can deliver lower all-in cost and faster end-to-end availability while preserving compliance, resilience and monetary trust. In some corridors that may be a linked instant-payment system. In others it may remain a modernised correspondent model. In wholesale markets, tokenised approaches may become more important. The global system is likely to converge on interoperability among several rails, not a single universal network.
What the approaching deadline means for banks
For banks, the 2027 deadline should be treated less as a compliance date than as a competitive benchmark. Customers will increasingly compare international payments with the instant domestic experience they already know. Institutions that can expose total cost upfront, commit to delivery times, automate beneficiary validation and reduce manual exceptions will have a defensible advantage even if the global targets are missed.
The investment priority is consequently not a single new rail. It is orchestration across rails. Banks need payment engines that can choose between correspondent routes, instant-payment links and emerging digital settlement options based on currency, destination, value, cost, urgency and compliance requirements. They also need cleaner reference data and stronger observability, because a bank cannot promise end-to-end performance if it cannot see where a payment is failing.
What it means for fintechs, regulators and investors
For fintechs, the remaining friction is an opportunity, but the easy part of the market is becoming crowded. Customer interfaces and low-cost routing are increasingly replicable. Durable advantage is more likely to come from regulatory licences, local payout reach, FX liquidity, fraud controls and direct connectivity to payment systems. The companies that can turn fragmented infrastructure into a predictable end-to-end service may capture more value than those that simply advertise faster rails.
For regulators, the next phase is about domestic execution of global principles. The FSB implementation phase launched in March 2026 reflects that shift. Authorities have to decide who can access domestic systems, how operating hours can be extended, where data rules can be harmonised, how non-bank participation should be supervised and how to align consumer protection with instant irrevocable settlement. Global targets ultimately depend on national rulebooks.
For investors, the approaching deadline is a reminder to distinguish infrastructure claims from customer economics. A provider may demonstrate impressive transaction speed but still depend on expensive FX, third-party payout networks or regulatory arbitrage. The most valuable payment businesses are likely to be those that own or secure scarce advantages at the difficult points in the chain: regulated access, liquidity, identity, compliance, distribution and interoperability.
The likely 2027 outcome: a missed target, but not a failed reform
The FSB has already prepared the market for the possibility that the global objectives will not be fully achieved on schedule. That should not be confused with zero progress. Payment messages are moving faster, domestic instant-payment systems are proliferating, ISO 20022 is becoming a common data foundation, and new cross-border links are moving from prototypes toward implementation. The reform has created infrastructure and standards that can compound over time.
The more critical test is whether those improvements become visible to ordinary users. A policy programme cannot ultimately be judged by the number of standards published, pilots launched or systems connected. It must be judged by what a small exporter pays, how long a migrant family waits, whether a corporate treasurer can predict settlement, and whether a payment reaches the right account without a manual investigation.
That is the central lesson of the approaching deadline. Cross-border payments are not slow because the world lacks fast technology. They are slow and expensive because technology has to cross institutions, currencies, laws and risk frameworks that were never designed as one system. The next phase of reform is therefore less about inventing speed than making speed survive the entire journey.
Conclusion
The world is unlikely to wake up on 1 January 2028 to a universally instant, one-percent cross-border payment system. The G20 deadline is more likely to mark a transition from global standard-setting to corridor-by-corridor execution. That may sound less dramatic than a payments revolution, but it is probably how the revolution will actually happen.
Banks will shorten chains, interlink domestic systems, improve data and automate compliance. Fintechs will exploit gaps in price and service. Regulators will have to make access and rulebooks more interoperable. New digital-money rails will compete with, and increasingly connect to, existing infrastructure. The winners will not necessarily be those with the fastest technology. They will be the institutions that can make the entire cross-border transaction - from quote to final credit - behave like one coherent service.
References
1. G20 Targets for Enhancing Cross-border Payments. Financial Stability Board. Targets page, accessed August 2026.
2. G20 Roadmap for Cross-border Payments: Consolidated progress report for 2025. Financial Stability Board. 9 October 2025.
3. FSB kicks off new implementation phase to enhance cross-border payments through public-private partnership. Financial Stability Board. 12 March 2026.
4. Enhancing cross-border payments step by step: insights from the 2025 monitoring survey. BIS Committee on Payments and Market Infrastructures. 27 May 2026.
5. Harmonised ISO 20022 data requirements for enhancing cross-border payments - updated report. BIS Committee on Payments and Market Infrastructures. 26 February 2026.
6. Remittance Prices Worldwide. World Bank. Latest available public cost data accessed August 2026.
7. Remittance Prices Worldwide - Data Catalog. World Bank. Metadata updated May 2026.
8. Unlocking trade potential: the benefits of improving cross-border payments. European Central Bank. Economic Bulletin Issue 2/2026.
9. The Eurosystem's comprehensive payments strategy. European Central Bank. 31 March 2026.
10. Cross-border payments: improving cross-border payments with TIPS. European Central Bank. Accessed August 2026.
11. Project Nexus: enabling instant cross-border payments. BIS Innovation Hub. Updated 27 August 2025.
12. FATF updates Standards on Recommendation 16 on Payment Transparency. Financial Action Task Force. 18 June 2025; update 28 October 2025.
13. G20 goals for enhancing cross-border payments. Swift. Accessed August 2026.
14. Swift accelerates transformation of consumer payments as banks roll out new framework for retail transactions. Swift. 5 March 2026.