Stablecoin settlement for PSPs: where the margin actually comes from
"Stablecoin settlement improves margin" is true and also nearly meaningless until you can say which line item moves and by how much. This is that breakdown.
The short answer — the three-line margin model: stablecoin settlement recovers PSP margin from three specific places: (1) fewer days of merchant-settlement float the PSP finances — roughly $27k/month at $50M monthly volume, a 2.5-day lag and 8% cost of capital; (2) correspondent and lifting fees on the PSP's own cross-border legs, replaced by sub-cent network costs and one disclosed FX conversion; (3) weekend and holiday settlement gaps, which is why Visa and Mastercard already settle in stablecoins at the network level. It is cost recovery, not a new revenue product.
Every PSP deck pitching stablecoin settlement to a board contains a slide that says some version of "stablecoins improve margin." Almost none of them say where. That vagueness is doing work: it lets the number be as large as the audience wants to believe, because nobody has been asked to trace it to a specific cost line. This article traces it.
To be direct about what this is not: stablecoin settlement is not a new revenue product for a PSP to sell merchants. It is a change to the PSP's own cost of funds and cost of settlement. The margin gain, where it exists, comes from three specific places, and the size of each depends entirely on the PSP's current corridor mix and treasury structure.
1. Float, or the cost of money you are fronting
Most PSPs settle merchants faster than they themselves receive funds from card networks or correspondent banks. A merchant sees "T+1" or "T+2" settlement; the PSP is often carrying the actual receivable for longer, financed out of its own balance sheet or a credit facility, and that financing has a cost. On cross-border volume the gap widens: a payout in a corridor with a two-to-four-day correspondent chain means the PSP is fronting merchant funds for days before its own settlement arrives.
Stablecoin settlement compresses this gap because settlement between the PSP and its liquidity partners happens in minutes rather than days, on a blockchain, at any hour. Every day of float removed is a day of financing cost removed. For a PSP running $50 million a month through corridors with an average 2.5-day settlement lag at an internal cost of capital of 8% annually, that lag alone represents roughly $27,000 a month in carrying cost. Compress the lag to same-day and most of that recurs as margin, not as a new fee charged to anyone.
The formula is monthly volume × (lag days ÷ 365) × annual cost of capital. Run against your own numbers:
| Monthly volume | 1-day lag | 2.5-day lag | 4-day lag |
|---|---|---|---|
| $10M (8% CoC) | ~$2,200/mo | ~$5,500/mo | ~$8,800/mo |
| $50M (8% CoC) | ~$11,000/mo | ~$27,400/mo | ~$43,800/mo |
| $250M (8% CoC) | ~$54,800/mo | ~$137,000/mo | ~$219,200/mo |
Float carrying cost only — before correspondent-fee and weekend-gap recovery. At 5% or 12% cost of capital, scale each cell by 0.625× or 1.5× respectively.
2. Correspondent and lifting fees on the PSP's own rails
PSPs with cross-border settlement legs pay the same correspondent-banking toll that importers and payroll platforms pay, just further upstream and at larger scale: intermediary bank fees, FX spread on internal treasury conversions, and the operational cost of reconciling a chain of banks that each take their own cut. None of this is visible to the merchant; it sits inside the PSP's own cost-to-serve.
A telecom-style cross-border settlement that might carry $20 to $30 in bank and FX fees on a $200 transaction can complete on stablecoin rails for a fraction of a cent in network cost, with the FX conversion happening once, at a disclosed rate, rather than accumulating across a correspondent chain. At PSP scale, run across thousands of settlement legs a month, this line item is usually the single largest and most direct source of margin recovery, because it is a fee the PSP was paying to a bank, not a cost being passed through to anyone else.
3. Weekend and holiday settlement gaps
Wire and ACH rails do not run on weekends or bank holidays. A PSP settling a large volume of Friday transactions is carrying that exposure, unfunded from the network side, until Monday or Tuesday. Card-network settlement between issuers and acquirers has the same gap, which is precisely why Visa and Mastercard moved first here: Visa's stablecoin settlement has reached a $7 billion annualized run rate across nine blockchains, up 50% in a quarter, and in June 2026 Mastercard opened settlement to six stablecoins, USDC, RLUSD, PYUSD, USDG, USDP and SoFiUSD, across eight blockchains including Ethereum, Solana and Base. That is bank-to-bank settlement infrastructure, a layer above merchant-facing PSP settlement, but it is the same mechanism solving the same problem: weekends should not be a cost centre.
Rough sizing, for orientation only: published industry comparisons put stablecoin acceptance costs around 0.5% of transaction value against 2% to 3% for card processing, with settlement in minutes rather than one to three business days. Your own margin recovery will not match a published average; it is a function of your specific corridor mix, current treasury structure and cost of capital, which is precisely why this needs modelling against your numbers rather than borrowed from an industry deck.
What does not show up as margin
A fair accounting has to include what stablecoin settlement does not fix and can add:
- Chargeback exposure does not move. If your card-rail dispute economics are a meaningful cost line, stablecoin settlement on the treasury side does not touch consumer-facing chargeback risk on the acceptance side. Those are different legs of the same business and need separate analysis. See our companion piece on how merchants accept stablecoin payments without holding crypto for the acceptance-side trade-offs.
- New operational cost appears. Custody or provider relationships, compliance monitoring for a new settlement instrument, and staff who understand the rail all cost money. This is usually smaller than the savings at PSP scale, but it is not zero, and early adopters underestimate it.
- Licensing exposure can change. Depending on structure, adding stablecoin settlement can touch money-transmission and, in the US, GENIUS Act considerations. See our piece on GENIUS Act compliance for payment companies for what actually applies to a PSP as opposed to an issuer.
- Provider and corridor quality varies enormously. The margin case above assumes a well-chosen liquidity and settlement partner. A poorly chosen one can reintroduce spread, slippage or counterparty risk that erodes some of the gain. This is the step every vendor's pitch skips, because it is the step where their own product might lose the comparison.
How to size this for your own book
The exercise that actually produces a number, rather than a slide, is this: take your last twelve months of settlement data, segment by corridor, and for each corridor compute three figures: average settlement lag in days, correspondent and lifting fees as a percentage of volume, and weekend/holiday-exposed volume as a share of total. Multiply the lag by your cost of capital, add the fee percentage, and you have a corridor-by-corridor estimate of what stablecoin settlement could recover, before providers get involved.
That number, not an industry average, is what should go in front of your board. Building it, and then choosing which providers and corridors to act on first, is the engagement we run for PSPs, priced fixed and independent of which rail you end up choosing.
Common questions
Where does the margin gain in stablecoin settlement actually come from for a PSP?
Three places: fewer days of merchant-settlement float that the PSP must fund, elimination of correspondent-bank fees on cross-border legs, and settlement on weekends and holidays when card-network and wire settlement are closed. None of these is a new product; each is an existing cost the PSP already carries that shrinks or disappears.
Do card networks themselves use stablecoin settlement now?
Yes. Visa's stablecoin settlement has reached a $7 billion annualized run rate across nine blockchains, and in June 2026 Mastercard opened settlement to six stablecoins including USDC, RLUSD and PYUSD across eight blockchains. This is issuer-acquirer settlement between banks on the card networks, a different layer from merchant-facing PSP settlement, but it establishes the same mechanism at the largest possible scale.
Does stablecoin settlement change what a PSP charges merchants?
It can, but the pricing decision and the cost-recovery mechanism are separate questions. A PSP can pass part of the margin recovery to merchants as a lower blended rate to win volume, keep it as margin, or split it. Which makes sense depends on the PSP's competitive position and corridor mix, not on the technology.
North Settlements provides business advisory services, not legal, tax, accounting or investment advice. Figures cited are from public industry reporting as of August 2026 and illustrative modelling assumptions; verify current pricing and rates before making decisions, and model against your own settlement data.
Want your corridors sized like this?
We build the corridor-by-corridor margin model above using your actual settlement data, then map it to named providers. Fixed fee, no commissions from anyone evaluated.
Book an intro call