Enterprise treasury runs on two clocks. The first is the one treasury controls: forecasts, hedges, netting calendars, payment runs, all planned to the day. The second is the one it does not: how long the money actually takes to arrive once an instruction leaves the building. Settlement infrastructure is the second clock, and for cross-border flows it still runs on days, with arrival times treasury can neither see nor promise. The FSB’s own 2025 assessment concluded the G20’s targets for faster, cheaper cross-border payments are unlikely to be met at the global level by end-2027.
This page is about what treasury should demand from the settlement layer itself, whether the flows are intercompany, supplier, or customer-facing.
The gap between the TMS and the rails
A treasury management system decides and instructs. The rails execute. Between the two sits the gap where problems live: an instruction that left the TMS on Monday becomes value with finality on Thursday, having crossed a correspondent chain that deducted fees in flight, converted currency at a spread nobody approved that morning, and passed cut-offs in three time zones. The TMS records what was intended. The rails deliver what happened. Reconciliation is the department that argues the difference.
Every improvement treasury has built on top, netting programs, in-house banks, payment factories, manages the gap. The newer question is whether the gap itself can close.
What to evaluate in settlement infrastructure
Bank relationships intact. The disqualifying demand is rip-and-replace. A settlement layer should work as an overlay: existing accounts, existing banking group, existing credit lines, with the layer routing value over the rails faster and with better evidence. If adopting the infrastructure means leaving your banks, the cost of the fix exceeds the cost of the problem.
Netting and internal flows first. Infrastructure that helps should compound with what treasury already runs. Multilateral netting cycles leave residual net settlements, and intercompany flows are the natural pilot corridor: both counterparties are yours, and the benefit lands directly on the group’s trapped cash.
Policy control. Treasury policy lives in documents; payments live in systems; the two meet only when someone checks. Ask where the infrastructure enforces policy. The strong answer is inside settlement itself: limits, permissions, jurisdiction rules, and approval conditions evaluated on every transaction, so a transfer that cannot satisfy them does not settle. That is programmable settlement doing governance work, and it turns policy from an audit finding into a property of the rails.
Auditability. Every settled transaction should carry evidence that its conditions were met, evidence a third party can verify without seeing the commercially sensitive detail underneath. Auditors, regulators, and the group’s own controllers all consume this; today they reconstruct it from fragments after the fact.
Rail optionality without rail religion. Fiat rails, tokenized deposits, and stablecoins each fit different corridors and policies, and the map will keep changing. Infrastructure that locks treasury to one rail type, including the newest one, recreates the problem it was meant to solve. The useful property is neutrality: route each payment over whichever rail fits, and change the mix by policy, corridor by corridor, as regulation and counterparties evolve. No stablecoin should be required to start.
Seconds, not days, with visible finality. The working-capital case rests here. Buffers are sized to uncertainty; settlement that completes in seconds with finality both sides can see shrinks the uncertainty that buffers exist to absorb.
Getting there without a program office
The realistic adoption path is incremental: one corridor, one flow type, usually intercompany, run in parallel with the existing process until the numbers argue for more. Because an overlay changes no bank relationships and requires no digital-asset decision on day one, the pilot is a treasury project rather than a group-wide transformation, and the exit cost of being wrong is small. That asymmetry, small downside and compounding upside, is what makes the second clock worth testing now rather than after the next system consolidation.
Where Frame fits
Frame is the settlement layer for modern finance. Through configured integrations, Frame Rules checks the institution's policy before release, and Frame Proof seals the outcome reported by the customer's execution platform.
The customer's platforms execute the transaction. Rail coverage, execution timing and finality depend on the configured integration and the underlying systems. The institution remains responsible for its policies, permissions and operating controls.
Common questions.
What does settlement infrastructure mean for a corporate treasury?
It is the layer that actually moves value once treasury has decided a payment should happen: the rails, the accounts, and the routing between them. Treasury management systems plan and instruct; settlement infrastructure executes. The gap between the two, the hours or days between an instruction leaving the TMS and funds arriving with finality, is where working capital sits idle and forecasts go wrong.
Does adopting new settlement infrastructure mean leaving our banks?
It should not. The test worth applying to any settlement layer is whether it works as an overlay: your existing bank accounts and relationships stay where they are, and the layer routes value over existing rails faster and with better evidence. Treasury gains speed and visibility without renegotiating its banking group, and the banks remain the point of record.
Do we need stablecoins to modernize settlement?
No. A rail-neutral settlement layer can settle over existing fiat rails with no stablecoin involved, and add stablecoin or tokenized-deposit rails later, corridor by corridor, if policy and counterparties call for it. Treating digital assets as an option rather than a prerequisite is exactly what rail-neutrality means, and it keeps the decision with treasury policy rather than with the infrastructure.
What should treasurers ask vendors about compliance?
Where compliance checks actually run. In most stacks, screening and policy checks happen in systems bolted on before or after settlement, which is why a payment can pass every internal check and still be held mid-route by an intermediary applying its own. The alternative design enforces policy inside settlement itself, so a transfer that cannot satisfy its conditions does not settle, and every settled transaction carries evidence that its conditions were met.
How does faster settlement change working capital?
Buffers are sized to uncertainty. When internal and external settlement takes days and arrival times are unpredictable, every entity holds cash for the worst case. When settlement completes in seconds with visible finality, the buffer question changes from how much float to carry to whether to carry float at all, and pre-funded positions across corridors can shrink accordingly.