Swift payments reach banks in seconds but beneficiaries still wait
by Ben Poole
Forty seconds is now enough for half of payments sent over Swift to reach the receiving institution. Three-quarters arrive within ten minutes. Yet the slowest 10% take more than 18 hours to reach the beneficiary’s account or wallet.
Swift’s Spotlight on Speed 2026 report examines the gap between rapid cross-border delivery and the domestic processing that follows. Once a payment reaches the recipient institution, it may still need to pass through local systems, checks and operating windows before the funds become available.
Across the report’s beneficiary-crediting data, 25% of payments were credited in under a minute and half in around 16 minutes. But the slowest quarter took roughly six hours or longer.
Corporate finance teams live with that tail. Until the credit appears, the cash forecast is still missing a receipt. Supplier queries can arrive just as treasury is making intraday liquidity decisions.
The analysis covers 40 receiving markets in the first quarter of 2026 and plots the full spread of crediting times. That shows where payments start to bunch up and where a small number fall badly behind.
The wait begins after the receiving bank
Reaching the beneficiary institution is an important milestone. It still isn’t the end of a payment. Funds need to pass through local processing and be credited to the beneficiary’s account or wallet. Swift calls this domestic stage the last mile.
Previous editions of the report found that the last mile accounts for more than 80% of an average cross-border payment’s total processing time. The international leg has accelerated to the point where domestic processes can now dominate the customer’s experience.
The G20 has set a target for 75% of international payments to be credited to the end user within one hour by the end of 2027. Swift’s 40-second and ten-minute figures measure only the journey to the receiving institution, so they can’t be compared directly with that end-to-end target.
An 18-hour wait for some recipients can therefore coexist with rapid bank-to-bank delivery. With the international leg now moving quickly, domestic hold-ups are harder to ignore.
Averages blur the outliers
Across the full data set, the median beneficiary crediting time was around 16 minutes. On its own, that figure sounds reassuring. It says little about the slowest payments, which account for a disproportionate share of total waiting time and the uncertainty felt by customers.
The percentile breakdown shows how quickly the tail lengthens. The fastest quarter was credited in less than a minute. At the other end, a quarter took around six hours or more and one in ten exceeded 18 hours.
Seconds shaved from an already-fast payment won’t help a company whose transfer is stuck for most of a working day. Spotting the hold-up sooner and fixing it faster would.
Its analysis covers international customer credit transfers between and within banking groups, excluding cover payments and transactions carrying a future value date. Swift presents the 40 receiving markets anonymously, arranging them by the distribution of their last-mile crediting times.
Where markets lose time
Roughly one-third of the markets fall into Swift’s “consistently fast” group. Their median last-mile time is at or below one hour, while average times generally range from 30 to 90 minutes. In the best-performing markets, 90% of payments are credited within about an hour. Near the edge of the group, that threshold extends towards seven hours.
Markets in the middle show a different problem. Many still have median times below one hour, proving they can credit payments quickly, yet their averages sit between two and five hours. Their 90th-percentile times range from around seven to 15 hours, and the wider spread among the middle half of payments points to uneven execution.
At the slower end, the issue is broader than a thin tail of exceptions. Median crediting times are around seven hours across much of the group, with averages of roughly seven to eight hours. In these markets, 90% of payments are credited within about 15 to 18 hours, and one market performs substantially worse than its peers.
Where the delay sits determines the response. If most payments are fast and a few get stuck, exception handling is the place to look. If more than 75% take over an hour, the drag is more likely to lie in market rules, operating conditions or infrastructure.
Why payments stall after they arrive
Some delays begin with something as mundane as a missing account field. Missing or inconsistent account details may trigger a manual review. Unclear remittance or purpose-of-payment information can do the same, particularly where local reporting rules demand fields that weren’t completed at initiation.
Sometimes the payment is ready and the market isn’t. The next system, liquidity window or processing team may be unavailable. Extending opening hours helps only if settlement arrangements, foreign exchange availability and exception handling keep pace.
Regulatory controls can hold up otherwise straightforward transactions. Swift points to purpose-of-payment reporting, capital controls and foreign exchange checks as examples. Precautionary reviews may also last longer when institutions lack clear guidance or need to request the same information more than once.
Visibility sometimes breaks as a payment moves between the cross-border network and domestic systems. Without a shared transaction identifier or timely status updates, participants may know that a transfer has stopped progressing without knowing where it sits, why it stopped or who must act.
What an 18-hour tail means for treasury
Once funds are at the receiving bank but remain uncredited, the recipient can’t necessarily use them. That affects the timing of incoming liquidity and can leave a receivables team chasing a customer whose treasury function believes the payment has already arrived.
Uncertainty matters even when the money turns up later the same day. Cash positioning depends on when funds become available; a payment message reaching a bank doesn’t create usable liquidity. Unpredictable credit times may encourage companies to keep larger buffers or delay outgoing payments until an expected receipt can be confirmed.
For treasury, dependable arrival times beat the occasional record-breaking payment. Swift links more predictable payments to better working capital management and smoother trade flows. Reliability also gives companies a firmer basis for scheduling supplier payments and funding accounts across markets.
Delays create work as well as uncertainty. Payment-status queries pull treasury, accounts receivable, banks and counterparties into work that should shrink when tracking continues through to confirmed beneficiary credit.
Domestic rails hold the next gains
Swift’s main recommendation is wider use of “one-leg-out” arrangements. Under this model, a domestic instant payment system handles the receiving-country leg of an international transfer, while tracking continues until the beneficiary has actually been credited.
That hand-off only works if the domestic scheme can accept the payment data and preserve its identity. Domestic systems need common data requirements and operating conditions that accommodate an international payment. The cross-border and domestic legs also need to retain the same transaction identity, otherwise speed improves while visibility falls away at the handover.
Selected live transactions show how fast that hand-off can be. Swift says selected payments using its retail framework have travelled end to end from Turkey to Spain in 15 seconds and from Australia to India in 37 seconds. A Brazil-to-US transaction completed in one minute. These individual corridor examples don’t describe market-wide performance, but they show that final credit in seconds is technically possible.
Thierry Chilosi, Swift’s chief business officer, said: “Cross-border processing speeds are consistently fast, but if we’re to deliver the experience that customers expect then the challenge now is to tackle the inconsistency that we see in the last mile in markets around the world.”
Broader adoption will require payment-system operators, banks and regulators to agree how international transactions enter domestic instant schemes. Tracking must cover that final leg as well. A faster domestic rail offers less value if the sender still can’t see whether the beneficiary received the money.
The metrics need to follow the money
Better data at the start can prevent some payments entering the long tail. Swift calls for stronger checks of beneficiary and account details before an instruction is sent. Purpose-of-payment and remittance information also need to meet the requirements of the destination market.
ISO 20022 provides the structured fields needed to carry richer information through the chain. Pre-validation and account-verification services can catch errors earlier, while common code sets reduce the ambiguity that sends otherwise routine payments into manual handling.
When a payment stalls, someone needs to know quickly and know what to do next. The Unique End-to-end Transaction Reference, mandatory for international payment flows over Swift since 2018, allows institutions to trace the same transfer across systems. Status updates and reason codes can then show whether it is waiting for information, liquidity, a regulatory review or an operating window.
Average speed remains useful, although it can flatter a service in which most payments are quick and a small number are extremely slow. Swift argues for closer attention to the median and selected percentiles, supported by data on how many payments enter exception handling and why.
Two clocks matter once a payment stalls: how long it takes someone to notice and how long they take to fix it. Digital case management can route requests to the right owner, while automated repair handles predictable data problems. More complex cases can be triaged for human review without holding up lower-risk transfers.
The next test is consistency
Most of the tools are already on the shelf. The work now is connecting them well enough that fewer payments drop into a manual queue.
Lower-volume markets won’t always be able to copy the model used by a major payment centre. Their systems and economics differ. The standard should still feel familiar to the customer: a clear arrival time, fewer unexplained interruptions and quicker answers when something goes wrong.
A 40-second journey to the bank sets a high bar. Progress will now show up at the other end of the distribution: fewer 18-hour waits, spotted and fixed sooner.
Like this item? Get our Weekly Update newsletter. Subscribe today
