Build your settlement strategy

Most used by

Any business with cross-border payables: supplier payments, merchant settlement, worker payouts, seller disbursements.

Merchant settlement and scheme-fee flows, where payout timing is contractual and every day of delay is merchant churn risk.

The core product. The send leg is the business, and the operator's cost to settle is the margin.

Client payment legs after conversion, where the settlement leg's speed bounds the desk's exposure window.

Salary disbursement to workers and contractors, where the arrival date is a promise the platform made.

Seller and creator payouts, where payout reliability is a retention feature of the marketplace itself.

What drives the choice

Corridor cost and speed, counterparty preferences, and how much failure the flow can tolerate.

Settlement finality before payout, corridor coverage matching the merchant base, and reconciliation that scales with volume.

Corridor cost against the 6.36% global average, payout-network reach, and float reduction in payout markets.

Finality speed to close exposure, and payment-versus-payment settlement where both legs move at once.

Date certainty first, then landed cost per worker, then coverage of employment markets.

Coverage of the seller base, cost at small ticket sizes, and failure rates that drive support load.

Example

A $250,000 supplier payment from the US to Mexico: on the correspondent chain it carries wholesale fees and lands when the domestic leg clears; routed over an instant rail or a stablecoin leg with local conversion, it can land the same hour. The right answer depends on the counterparty's bank, and a rules-driven router can make that choice per payment.

Merchant settlement to Brazil: the US to Brazil corridor averages 6.4% to send $200 on surveyed services. An acquirer settling wholesale volumes pays far less in percentage terms, but the same chain structure sets its floor. Routing the Brazil leg over a modern rail moves settlement from days to hours without changing the merchant contract.

A payout operator adds a stablecoin leg for the UK to Nigeria corridor, where surveyed services average 3.1% and the digital channel is the cheap one. Local conversion happens at the payout partner, and the operator's pre-funded float in Lagos shrinks with each hour cut from settlement.

A broker settling the dollar leg of a EUR/USD trade wants payment versus payment: both legs settle together or neither does. Where a shared venue exists, that is the rail; where it does not, the broker prices the exposure window into the spread.

India payday for 4,000 workers: the US to India corridor averages 4.2% to send $200 on surveyed services, and wholesale rates are far tighter, but the date matters more than the basis point. The platform routes the bulk leg early over fiat rails and keeps an instant rail as the fallback for late adjustments.

Payouts to Indonesian sellers: Malaysia to Indonesia averages 5.0% on surveyed services and the region's instant rails are linking up. The platform routes payouts through a local partner and holds the option to move legs to newer rails as they prove out.

Most used by

Businesses collecting from customers or partners abroad: acquirers, marketplaces collecting buyer funds, operators receiving partner settlements.

Collecting buyer funds and receiving settlement from schemes and partners across markets.

Receiving sender funds in origination markets, often across dozens of collection partners.

Receiving client funds pre-conversion, where crediting speed sets how fast the desk can quote.

Collecting employer funds ahead of payroll runs, where a late collection becomes a late payday.

Collecting buyer payments globally before the split to sellers.

What drives the choice

Data quality on arrival, screening before credit, and reconciliation cost per payment.

Matching collections to merchants automatically; the reconciliation break is the expensive event.

Collection coverage in origination markets and the cost of holding collected float.

Credit speed and certainty of the received leg before releasing the converted one.

Predictable collection timing from employers across entities and currencies.

Collection cost at scale and clean attribution of every inbound dollar to an order.

Example

A payment arrives with a beneficiary name that does not match the account. On legacy rails that is a repair: about $12 in handling cost, on 2023 data, and a day of delay. Received over a rail with structured data and pre-credit screening, the mismatch is caught before value moves.

An acquirer receives scheme settlement across 14 currencies. Each arrives as an unstructured credit to reconcile against expected batches. Structured settlement data turns a manual matching queue into an exception report.

Sender funds collected across three markets settle into the operator's account with the originating reference intact, so the payout leg can release the moment the collection clears rather than on a batch schedule.

A desk releases the converted leg only when the received leg is final. Deferred finality means quoting wider; settlement finality in minutes means quoting tighter and winning flow.

An employer's funding wire arrives the morning of a payroll run. With pre-credit screening and instant crediting, the run releases the same day instead of rolling to the next.

Buyer funds arrive tagged to orders, so the seller split runs automatically and the platform's suspense account stops growing.

Most used by

Anyone who pays out on a schedule: the float sits wherever settlement is too slow to trust.

Payout float per merchant market, sized to the slowest expected settlement day.

Payout-market float, the largest balance-sheet item in most payout networks.

Currency balances at clearing partners so either leg of a trade can settle on demand.

Payroll float per employment market, sized to payday peaks rather than average flow.

Payout float across seller markets, growing linearly with each new market opened.

What drives the choice

Days of cover held, the cost of capital on those balances, and how quickly a faster rail lets them shrink.

Settlement speed per corridor sets the float; every hour cut compounds across markets.

Corridor settlement speed and payout-partner terms set the days of cover.

Netting across flows and payment-versus-payment settlement cut the balances a desk must hold.

Payday concentration: float sized for the 1st and 15th, idle in between.

Market count: each new seller geography adds another parked balance.

Example

There is no audited public figure for global trapped liquidity, which is itself part of the problem. The one primary scale signal: a full-migration scenario to a shared settlement network of the kind J.P. Morgan and Oliver Wyman model would cut commercial banks' overnight balances by about $10 billion. The calculator in chapter 5 prices your own version of this cost.

A PSP holding 2 days of cover on $120M monthly flow parks about $8M across payout markets. At a 5 percent cost of capital that is roughly $400K a year spent on the slowness of rails.

An operator holding 4 days of cover on $60M monthly flow parks about $8M in payout markets. Cutting settlement from days to hours releases most of it back into the business.

A desk pre-funding both sides of its top pairs can net exposures only as fast as settlement confirms. Faster finality means smaller parked balances for the same quoted liquidity.

Five days of cover on $40M a month is about $6.6M parked. Payroll floats are sized to payday peaks, so the true figure is higher on the 1st and the 15th.

Three days of cover on $80M a month is about $7.9M parked across seller markets, a balance that grows every time the marketplace opens a new country.

Most used by

Every cross-border flow with two currencies in it, and increasingly flows that change form within one currency.

Multi-currency merchant settlement and treasury rebalancing across collection currencies.

The FX leg of every transfer, where the margin is the least visible part of the price.

The product itself. The question is what the settlement leg costs the desk around the conversion.

Converting employer funding currencies into dozens of payout currencies each cycle.

Converting buyer currencies into seller payout currencies at marketplace scale.

What drives the choice

Visible spread against the mid-rate, netting before converting, and freedom to choose the conversion venue per corridor.

Netting collections against payouts before converting the residual.

Transparent pricing per corridor; the bank channel discloses least, per Frame corridor analysis.

Venue choice and settlement speed on the converted leg.

Rate consistency across a payroll cycle, so workers in one country are not priced differently by day.

Automating conversion at order time versus payout time, and netting across the marketplace book.

Example

Where a corridor is served mainly by banks, the same $200 transfer costs about three times what money-transfer and mobile operators charge, 15.0 percent against 5.0 percent on surveyed services, per Frame corridor analysis of World Bank data. The premium is mostly conversion margin, and it is the clearest public evidence that conversion pricing depends on the channel, not the currency pair.

A PSP collecting in 12 currencies and paying out in 8 nets flows first, converts the residual at a chosen venue, and settles both legs on rails picked per corridor. Each of those three choices is a margin lever the correspondent chain never offered.

On bank-dominated corridors the intermediary premium reaches 70 percentage points over the non-bank alternative, per Frame corridor analysis. Operators win those corridors by carrying the conversion themselves on a cheaper rail.

The desk already owns the conversion. The settlement layer's job is to make the legs around it instant and final, so the spread stops subsidizing the rails.

Converting once per cycle at a netted rate, rather than per payout at retail spreads, is often the single largest recoverable cost in a payroll flow.

A marketplace that converts at payout time, netted across all sellers in a currency, captures margin that per-order conversion gives away.