FedNow Opens to Cross-Border Messaging: Why Logistics CFOs Should Rethink What They Pay for International Settlement
On April 8, 2026, the Federal Reserve Board proposed amendments to Regulation J that would permit FedNow participants to use intermediaries, including correspondent banks, to send funds transfers through the service. The change is narrow but significant: it converts FedNow from a domestic-only rail into infrastructure that can settle the U.S. portion of a cross-border transaction in real time.
Under the current framework, FedNow payments may include only two U.S. banks, with no ability to route through correspondents. That structure has kept the service effectively domestic since its 2023 launch. The Fed's proposal would allow a FedNow participant to send a payment to another participant acting as an intermediary, which could then route the payment onward through correspondent banking networks for the international portion. The change would align FedNow with the Fedwire Funds Service, which has permitted intermediaries for decades.
The proposal does not alter immediate funds-availability requirements for U.S. beneficiaries. A payment received by a U.S. bank acting as beneficiary's bank would still settle instantly. But for outbound cross-border transfers, the beneficiary's bank would sit outside the U.S., and Regulation J would not oblige it to make funds available immediately. The domestic leg settles in real time; the international leg depends on what comes after.
This asymmetry matters because it reveals where the friction actually lives. Swift data shows that the international leg of a cross-border payment, the portion moving between financial institutions across borders, accounts for less than 20% of total processing time. Around 80% of total processing time is spent in the last mile, after the payment reaches the beneficiary bank, where domestic clearing systems, local regulations, and market practices introduce delays. A payment arriving in the U.S. at 7 p.m. Eastern might wait until morning for domestic settlement. One arriving Saturday could sit until Monday.
FedNow's proposed rule change addresses exactly this problem for inbound payments: real-time settlement eliminates the queue. But the fix is regulatory, not technical. The infrastructure already exists. The Fed simply prohibited its use for cross-border flows. Now it is reconsidering.
For logistics operators managing supplier payments across multiple corridors, the implications are direct. The 2-4% you lose on cross-border settlements, wire fees, FX markups, correspondent bank deductions, and capital locked in transit, compounds across every payment. Traditional banks typically apply markups of 1.5% to 4% above the mid-market rate on currency conversion, and that cost rarely appears as a line item. Correspondent banks charge $15 to $50 per hop, with one to three banks processing a single transfer. By the time funds arrive, the recipient receives less than expected, with no clear accounting for where the value went.
These costs are structural, not arbitrary. Pre-funded nostro accounts tie up capital. Duplicated AML checks add friction at every node. Business-hours-only processing extends settlement windows. But the FedNow proposal demonstrates that at least part of this structure, the part under U.S. regulatory control, can be changed with a rule amendment. The question is what other parts are equally malleable.
The Clearing House's RTP Network is moving in the same direction. It has set a September 2026 target for enabling domestic correspondent bank activity on the network, the first step toward supporting "one leg out" transactions involving foreign counterparties. RTP currently handles roughly 97% of instant payment volume in the U.S. and processed $1.3 trillion in 2025, real scale that could reshape how cross-border settlements clear on the American side.
Outside the U.S., domestic instant payment systems are already linking directly. Project Nexus, a BIS Innovation Hub initiative, connects the instant payment systems of India, Malaysia, the Philippines, Singapore, and Thailand through a hub-and-spoke model. A Thai small business can receive payment from a Malaysian buyer in seconds, settled in baht, with no correspondent banking chain. Singapore's PayNow linked with Thailand's PromptPay in 2021, and similar bilateral connections now span much of Southeast Asia. The infrastructure is being deployed faster than most observers anticipated.
The Financial Stability Board's 2025 consolidated progress report on the G20 cross-border payments roadmap acknowledged that the global targets for cost, speed, and transparency are unlikely to be met by the 2027 deadline. Policy work is largely complete, but real-world impact remains limited. The bottleneck is implementation, not design. Disjointed regulation, inconsistent AML requirements, sluggish infrastructure upgrades, and continued reliance on correspondent banking are slowing progress.
For a CFO at a mid-sized freight forwarder, this creates a strategic tension. The Fed has just demonstrated that instant settlement is achievable through rule changes. The correspondent banking model you rely on for cross-border payments is not a technical necessity, it is a policy choice that other jurisdictions are actively reconsidering. The fees you pay are not fixed costs of moving money; they are artifacts of infrastructure that is being bypassed elsewhere.
The challenge is operational. Treasury teams built for overnight funding do not function in a world of always-on settlement. Money has to be in the right place, in the right currency, around the clock. The FedNow proposal does not solve FX conversion, compliance checks that hold funds for review, or payment data that gets stripped between systems. Speed on the U.S. side is the easy part. Until those adjacent frictions move at the same speed as the rails, the end user will not feel the difference.
But the direction is clear. Domestic instant payment systems are no longer isolated national utilities. They are being connected, corridor by corridor, regulation by regulation, into something that looks less like SWIFT's correspondent model and more like a network of local rails with bilateral and multilateral linkages. The G20 targets may slip, but the infrastructure is being built anyway.
The 3-day settlement windows, mystery correspondent fees, and FX spread leakage that logistics operators have normalized are not immutable features of cross-border commerce. They are characteristics of a specific infrastructure layer that is now being reengineered. The question for treasury teams is not whether alternatives will emerge, but whether you will recognize them when they do, and whether you will still be paying the settlement tax by the time they become the default.
References
[2] Swift, "Unlocking last mile speed in cross-border payments," July 2026
