The Real Fix Is a Faster, Better-Governed Last Mile

The shortest accurate answer to the central question is that a cross-border payment can slow down after reaching the beneficiary bank because receipt begins a local decision, not because the international network has necessarily failed. The subject is the last-mile path from beneficiary-bank receipt to end-customer credit. The final judgment is correspondingly specific: improve the information, interfaces, workflow, architecture, and governance that make a compliant credit possible, while preserving the rules that protect financial integrity and serve legitimate macroeconomic purposes.

That answer is stronger than either of the two simple stories that usually compete for attention. One story says the network is slow, so new technology should make the message travel faster. The other says regulation is the problem, so removing checks would solve delivery. The evidence supports neither as a complete explanation. A payment may travel through the international route and still wait for beneficiary information, payment purpose, supporting documents, account eligibility, amount limits, reporting, or an exception review. But the existence of a check does not prove that every hour of waiting is necessary. The relevant question is where the decision is pending and whether the surrounding system helps the bank reach it without avoidable repetition.

The post-arrival delay is a system outcome

A payment has at least three operational stages: initiation by the payer and originating bank, movement through the international message and settlement chain, and processing after the beneficiary bank receives it. The final stage is the one that is invisible when “sent” and “received” are treated as synonyms for “available.” A tracking record can show that the international instruction reached the receiving bank. The customer’s account is not complete from their perspective until the bank records the credit and makes the funds usable.

At that point, the beneficiary bank may need to establish who is receiving the money, what activity or purpose it represents, whether the account can receive the funds, and whether the supporting evidence is adequate. Inflow capital-control policies can require these checks, and some jurisdictions add transfer limits, account-type restrictions, or reporting duties. AML/CFT screening, sanctions checks, tax reporting, fraud controls, and ordinary internal procedures may overlap with them. The papers do not support a single universal sequence, but they do support a clear mechanism: local rules and local workflow can determine when a payment crosses from bank receipt into customer credit.

The difference between a short and a long wait is often a difference in dependencies. Complete, accurate information for a familiar customer can support processing in minutes, according to the IMF paper. An ambiguous purpose, missing document, account mismatch, or unresolved exception can keep the payment pending for much longer. Operating hours, batch processing, time zones, currency conversion, and intermediary handoffs can compound the wait. The customer sees one delay; the system may contain several interacting reasons for it.

This is why the last mile should not be reduced to a compliance queue. The queue is created by the interaction of a rule with the information available to apply it, the bank’s workflow, the systems on either side of the interface, and the institutions that make the cross-border connection possible. A necessary control can be slow because the evidence is genuinely incomplete. It can also be slow because the same evidence has been collected in a form the next institution cannot interpret. Those cases have different remedies even though they produce the same customer experience.

The IMF study gives this mechanism an important empirical anchor, but not a complete causal map. It examines 2020–2022 country-year data for 176 countries, combining anonymized Swift-derived Beneficiary Elapsed Time with an AREAER-based Financial Account Restriction Index. Average beneficiary elapsed time in the sample is 30.72 hours, with a median of 27.84 hours and an observed range from 2.50 to 120 hours. A one-standard-deviation increase in the inflow FARI is associated with nearly eight additional hours in the univariate specification and slightly more than four hours after the stated controls are added.

The unit of that result matters. The four-to-eight-hour range is an association for a one-standard-deviation difference in a country-level index, not a forecast for one customer or a guaranteed effect of changing one rule. FARI averages binary indicators of whether covered financial-account restrictions exist across 29 components. It does not measure the intensity of each restriction, how it is enforced, or how efficiently a bank implements it.