The G20 cross-border payment targets are slipping. Here is the scorecard

Five years ago the G20 made cross-border payments a priority and endorsed hard numbers to hold the system to: most with a deadline of end-2027. In October 2025 the Financial Stability Board, which monitors those targets, delivered its verdict on the halfway mark: “It is unlikely that satisfactory improvements at the global level will be achieved in line with the 2027 Roadmap timetable.” This page is the scorecard: what was promised, what the 2025 numbers show, and what the gap means for institutions that move money.

The targets

The G20 endorsed 11 global targets in 2021 across three market segments. The headline ones:

Segment	Target	Deadline
Wholesale	75% of payments credited within one hour of initiation	End-2027
Retail	Global average cost no more than 1%; no corridor above 3%	End-2027
Retail	75% of payments credited within one hour	End-2027
Remittances	Global average cost no more than 3%; no corridor above 5%	2030

The 2025 scorecard

The FSB’s own summary is blunt: “The KPIs for 2025 show only a slight improvement at the global level since the KPIs were first calculated in 2023.”

Wholesale speed is the bright spot. Across the March 2023 to March 2025 monitoring period, the share of wholesale payments credited within one business day held above 90%, and the share credited within one hour, now above 50%, increased. The improvement came largely from faster processing on the beneficiary leg, traditionally the slowest part of the journey, which Swift attributes to greater automation in banks’ internal processes and to the ISO 20022 migration. Above 50% is still a long way from 75%, but the direction is right.

Costs are stuck. The FSB describes average global costs as “sticky”, improving mainly in the most expensive regions. FXC Intelligence, which supplies the retail data behind the FSB’s indicators, reported the average cost of person-to-business cross-border payments easing from 2% to 1.9% in 2025, still roughly double the 1% target. For remittances the World Bank’s global average remains 6.36%, more than twice the 2030 target.

Transparency cannot even be fully measured. The FSB reports slight improvement in some areas, notes the available data have limitations, and currently lacks a data source for wholesale transparency altogether.

The regional spread is wide. North America is the fastest region for wholesale payments and the Eurozone recorded the largest improvement; sub-Saharan Africa remains the slowest region for receiving wholesale payments while being the fastest for receiving remittances, though costs there remain high. Europe and Central Asia have the cheapest retail cross-border payments.

The policy work is done. The outcomes are not

What makes the 2025 report striking is the mismatch it documents. The international policy program is largely complete: the FATF finalized its revision of Recommendation 16 on payment transparency in June 2025, the FSB finished its recommendations for regulating bank and non-bank payment providers and for aligning data frameworks, and the CPMI’s harmonized ISO 20022 data requirements are set for industry implementation by end-2027. As the report puts it, “these efforts have not yet translated into tangible improvements for end-users at the global level.”

Why not? The FSB lists the frictions: misaligned AML/CFT controls, inefficient capital controls, limited transparency, interoperability challenges, insufficient competition in some segments, and the long lead times of infrastructure change. There is also a telling detail in the monitoring survey: the share of fast payment systems that completed or planned at least two priority improvements rose from 91% to 98% between 2023 and 2024, while the same share for RTGS systems fell from 71% to 58%. The newer rails are modernizing faster than the core wholesale infrastructure the targets depend on.

What the misses mean for institutions

For a bank, payment provider, exchange, platform, or corporate treasury, the scorecard carries a practical message: the system’s official improvement track will not, on its own current trajectory, deliver hour-level, percent-level cross-border payments by 2027. The correspondent chain’s frictions, pre-funding, cut-offs, per-hop fees, are being sanded, not removed. Institutions that need target-level performance in specific corridors are increasingly assembling it themselves, from the alternatives that already exist: upgraded incumbent rails, direct-connection networks, and regulated digital-cash rails.

What to watch between now and 2027

Three markers will tell whether the trajectory changes. First, the wholesale one-hour share: it is above 50% and rising on the strength of the beneficiary leg, and Swift’s evidence suggests speed improves further as both ends of a corridor complete the ISO 20022 migration, so the end-2027 industry deadline for the harmonized data requirements doubles as the targets’ best hope. Second, the FSB’s next annual report, due October 2026, which will show whether the 2025 uptick in RTGS operating-hours extensions and non-bank access translates into measured speed. Third, remittance costs in the most expensive corridors, where the FSB’s own corridor analysis found the binding constraints are cash dependence, weak competition, and non-bank providers locked out of payment systems, problems that rule changes address slowly and infrastructure addresses faster. None of these markers requires optimism to track; they are all published numbers, and this page will be updated as they move.

Where Frame fits

Frame is the settlement layer for global finance: one integration to orchestrate payments at scale across fiat rails, stablecoins, and tokenized deposits, with compliance enforced on every transaction and settlement in seconds instead of days.

The G20 targets are corridor-by-corridor outcomes, and no single rail hits them everywhere. That is the case for a settlement layer that is rail-neutral by design: Frame routes each payment across whichever rail meets the corridor, the counterparty, and the governing policy, an upgraded fiat rail where that is fastest, a regulated stablecoin or tokenized deposit where it is not. Compliance runs inside settlement through Frame’s Rules Engine, so the checks that the FSB lists among the roadmap’s persistent frictions are enforced once, as a condition of settlement, rather than repeated at every hop. The targets set the bar at hours and single-digit percentages. Settlement in seconds clears it with room to spare.

See how a rail-neutral settlement layer approaches this in the Frame Blueprint.