SWIFT gpi already credits roughly 60% of payments to end beneficiaries within 30 minutes, and nearly 100% within 24 hours, according to BIS Papers No. 167 (Claessens & Rice, March 2026), drawing on SWIFT's own gpi data. For wholesale, correspondent banking already works. That's the opposite of how most crypto-payments content opens this conversation.
Most of that content claims a blanket win over SWIFT but the real picture is more complex.
What SWIFT gpi actually delivers today
Per the BIS paper, wholesale cross-border payments, institution to institution, are "relatively low cost and fast" through correspondent banking, even if still less efficient than domestic transfers. Between major, well-connected markets, gpi performance can rival domestic payment systems. The G20's own 2025 progress report targets 75% of wholesale payments settling within one hour by 2027, and on the strongest corridors, that bar is already close to being met.
If your business moves money between two established financial centers, on a well-banked corridor, in a currency both sides trade heavily, SWIFT gpi is fast. But speed alone isn't enough.
Where correspondent banking actually breaks down
BIS data shows correspondent banking relationships have been retreating for years, concentrated in emerging markets and developing economies, driven largely by rising AML/CFT compliance costs for the banks maintaining those relationships. A bank doesn't necessarily exit a corridor because payments are slow through it, but because the compliance burden of maintaining that relationship stops being worth it for the volume involved.
We see this directly with clients. Established, fully legitimate multinational suppliers in certain sectors and geographies get shut out of correspondent banking entirely, not because anything about their payments is slow, but because banks have de-risked away from their sector or region. Not even traditional rails have solved for what is essentially an access problem.
What stablecoin settlement actually adds here
Stablecoin settlement offers something SWIFT can't: a payment path that doesn't route through the correspondent chain a bank has already exited. Value moves directly, on-chain, and settles in minutes rather than waiting on a banking relationship that may not exist.
Both ends of the payment still need an on-ramp or off-ramp into local currency, and BIS is blunt about this: AML/CFT enforcement for stablecoins is, in practice, only really feasible at those conversion points, not throughout the full transaction lifecycle. The advantage is access, not the elimination of compliance work, it just moves where that work happens.
Where crypto doesn't beat SWIFT
On the corridors SWIFT gpi already handles well, major markets, established banking relationships, high volume, stablecoin settlement doesn't have an obvious edge. You'd be trading a fast, well-understood correspondent banking path for one that still needs on/off-ramp conversion, plus the operational overhead of managing that conversion at both ends. For those corridors, speed was never the actual gap crypto is solving.
The limits
BIS flags three unresolved issues before stablecoins can reliably serve as a cross-border payment instrument at scale: a globally consistent legal framework (still uneven, with some jurisdictions banning private stablecoins outright), AML/CFT enforcement that currently only bites at conversion points rather than across the full chain, and limited interoperability with the traditional financial system stablecoins still need to connect to at both ends. The paper's own conclusion is measured: many observers think other approaches to improving cross-border payments are preferable given the risks, not that stablecoins are the clear answer.
Use-case decision table
| Corridor type | Typical size | Where the constraint sits | Better fit |
|---|---|---|---|
| Major-market wholesale (e.g., established EU-US flows) | Large, low frequency | Already fast via SWIFT gpi | Correspondent banking / SWIFT gpi |
| De-risked sector or geography (banking access limited or absent) | Mid-to-large | Access, not speed | Stablecoin settlement, with on/off-ramp handled by a provider built for it |
| High-frequency smaller B2B payouts | Small-to-mid, high volume | Cost per transaction and on/off-ramp friction | Depends on corridor; worth comparing both |
The real consideration isn't about crypto and whether it "beats SWIFT" but whether your corridor is constrained by speed or not. SWIFT gpi can solve this on well-banked routes. But if you are constrained by the lack of access, stablecoin is an apt solution.
Recommended reading
- On-ramps and off-ramps explained: how crypto becomes usable money
- The real cost of crypto payments: on-ramp, off-ramp and the hidden spread
- Gas fees explained: what they are and how networks compare
Frequently asked questions
Can cryptocurrency be used for cross-border B2B transactions?
Yes, and it's already happening, particularly in corridors where correspondent banking access, not payment speed, is the actual constraint businesses face.
What crypto is best for cross-border payments?
For B2B settlement, the relevant answer isn't a speculative coin, it's stablecoins, since they settle directly on-chain without needing an existing correspondent banking relationship in the corridor.
Can I transfer crypto from one country to another for business payments?
Yes. Stablecoin settlement moves value directly between wallets rather than through a correspondent banking chain, though both ends still need an on-ramp or off-ramp into local currency.
What is the "holy grail" of cross-border payments?
It's the industry shorthand for payments that are simultaneously fast, cheap, transparent and universally accessible, the same four criteria the G20's own roadmap targets, and no single rail, SWIFT gpi or stablecoin settlement, achieves all four in every corridor today.