Wire Transfer Origination (Fedwire / CHIPS) — impact tolerance ≤ 2 hours — ITOL

Impact tolerance: ≤ 2 hours — Three arms, tiered to the promise the bank actually makes each client class. Promise arm (all clients): the only delivery promise in a standard US wire agreement is same-business-day processing for a wire accepted before the published customer cutoff. So every queued wire must still release before the 18:45 ET customer cutoff — the value date's hard wall. Late in the day the tolerance therefore compresses to the time remaining before the cutoff, minus the time needed to drain the queue. Deadline arm (deadline-bound flow): ≤ 2 hours during Fedwire operating hours (21:00 ET prior day – 19:00 ET) for any wire that must land by a stated hour — a margin call, a scheduled closing, a same-day tax payment. A deadline-bound wire that misses its stated hour is a breach on its own, whatever the volume. The recall/freeze capability on suspected-fraud wires shares this arm, since recovery odds collapse within hours of release. Queue arm (counterparty): > 5% of daily origination volume or ≈ $500M of queued wire value, whichever trips first (illustrative default — trace the derivation, then replace with your own harm analysis)

Definition: Send US-dollar wholesale wire transfers on behalf of retail, commercial, and correspondent customers via Fedwire Funds Service and, where applicable, CHIPS (the Clearing House Interbank Payments System). Includes the fraud-response workflow for outbound wires: initiating recall/return requests, beneficiary-bank freezes, and FinCEN (Financial Crimes Enforcement Network)/FBI kill-chain referrals. That workflow operates on an hours-long window before fraudulent proceeds disperse.

Why it is designated: Wire harm runs on two clocks, and the tolerance now names whose clock each one is. Start with what the bank promises. Published wire agreements at large US banks all make the same promise, because it is the template UCC (Uniform Commercial Code) Article 4A is built around. A wire received before the bank's established cutoff is processed the same business day. One received after the cutoff may be treated as received the next funds-transfer business day (UCC 4A-106). A 2026 survey of published regional-bank agreements found those customer cutoffs in the late afternoon — roughly 4:00 to 6:15 p.m. ET, later for commercial channels — always inside Fedwire's own 18:45 ET customer cutoff. No agreement promises an hour of arrival. Retail deadline events are usually funded a day ahead: a wire sent Tuesday covers a Wednesday closing even at the slow edge of the promise, because incoming electronic payments must be available no later than the next business day (Reg CC § 229.10(b)) and banks generally credit incoming wires the same day. So retail harm concentrates at the value date — the promise arm. The clients who wire same-day against a fixed hour are commercial and treasury clients, sold the later cutoffs for exactly that flow. Their harm ramps within an hour or two of queuing, whatever the hour of day — the deadline arm's ≤ 2 hours protects them, and the fraud-recall workflow rides the same clock. Every other wire faces one shared wall. Fedwire is a systemically important payment system, and its settlement day closes at 19:00 ET — with customer transfers cut off earlier, at 18:45 ET, and Treasury tax payments at 17:00 ET. A wire still queued at the customer cutoff misses its value date, and a delay becomes a failure. So an outage that runs past mid-afternoon converts queued customer wires into failed same-day obligations, such as closings and margin calls. It also strains liquidity at counterparty banks. The promise arm makes the 18:45 cutoff a hard backstop: through the afternoon, the tolerance compresses to the time left before the cutoff, minus the time to drain the queue. Read together, 2 hours is the longest outage that, starting at the worst plausible hour in mid-afternoon, still leaves runway to drain the queue before the customer cutoff. One distinction carries the calibration: harm to a single deadline-bound client can begin inside the first hour, but the tolerance marks where accumulated harm turns intolerable, not where harm first appears. The 2003 Interagency Sound Practices paper sets a 2-hour recovery goal for core clearing and settlement organizations, and asks firms that play significant roles in critical markets to strive for a 4-hour recovery capability. The paper does not size that group down to a regional bank, so the archetype adopts the 4-hour capability as its floor voluntarily. The published ≤ 2 hours is that adopted floor plus a deliberate buffer, sized to the 18:45 customer cutoff and the same-day liquidity at stake. The recalibration at ≈ $500B kept the 2-hour number but changed the severity qualifier. A percentage-of-volume trigger drifts as the book grows, so the line now also trips at ≈ $500M of queued wire value. A channel outage concentrated in large corporate flow can queue half a billion dollars of wires without ever touching 5% of the wire count. One last grounding fact: UCC 4A-305 caps the bank's legal damages for a late wire at interest for the period of delay. The tolerance is calibrated to customer harm, which the law does not compensate — not to legal exposure, which is near zero.