Cut-off time

A cut-off time is the hidden clock inside every payment. Banks and settlement systems run on business days, and a business day has to end somewhere: there is a moment after which today’s queue closes and new instructions belong to tomorrow. That moment is the cut-off, and for a payment crossing several institutions, it is the single biggest determinant of whether the journey takes hours or days.

Domestically, cut-offs are a minor nuisance. A missed deadline costs one day, and instant payment systems have removed even that for many retail flows. Cross-border is where the clock bites. A payment through a correspondent chain faces a separate cut-off at every hop, each set in the local time zone of a different institution. A transfer leaving Singapore in the local afternoon reaches a New York intermediary before dawn Eastern time, which is fine; one leaving New York in the afternoon reaches Asia after hours, and waits. The famous worst case is the Friday payment: miss one cut-off heading east and the transfer sits through the weekend, clears one leg Monday, misses the next deadline, and lands Tuesday. Four calendar days for a few seconds of actual processing.

Cut-offs are why the value date of a cross-border payment is so hard to promise, and they interact badly with everything else in the chain: a payment held briefly for a compliance query can miss a window it would otherwise have made, turning a two-hour delay into a two-day one.

Rails that never close dissolve the problem at the settlement layer. RTGS operators have been extending hours, and shared-ledger rails settle continuously, with finality at any hour. The remaining cut-offs live at the edges, where continuous rails hand off to institutions that still keep business days. The clock has not disappeared; it has moved to the boundary.