Sending money across a border has historically been slow and expensive for structural reasons that have nothing to do with the difficulty of the task. Stablecoins address those reasons directly, and the results on the part they address are genuinely dramatic: value that took days now moves in minutes, at a cost that rounds to nothing. But an international payment is three legs, not one — getting in, crossing, and getting out — and the technology only replaces the middle one. Understanding where the remaining cost sits is the difference between a realistic view of this use case and the promotional one. It assumes you know what a stablecoin is.
01 — What is actually broken
The reason an international transfer is slow is not that the message takes days to arrive — it is that no single institution has accounts everywhere. A bank in one country typically has no relationship with a bank in another, so the payment travels through a chain of intermediaries, each holding accounts with the next, each taking a fee, each applying its own compliance checks, and each operating on its own business hours and cut-off times.
Everything unsatisfactory about the experience follows from that structure. Cost accumulates because several institutions each take a margin, most visibly in the exchange rate rather than in the stated fee. Delay accumulates because the payment waits at each hop and stops entirely at weekends. Uncertainty is intrinsic, because no participant sees the whole path, which is why senders often cannot say what will arrive or when. And the chains are longest and thinnest precisely where they are needed most — corridors into smaller economies have fewer intermediaries willing to serve them, which is why those corridors are the most expensive in the world.
02 — What the middle leg becomes
A stablecoin transfer replaces that entire chain with a single transaction. Value moves directly from one wallet to another on a public network, settling in seconds or minutes, at a network fee that on an efficient chain is a negligible fraction of the amount. There are no intermediaries taking a cut, no correspondent relationships to maintain, no cut-off times, and no weekends — the network runs continuously.
Two properties matter beyond speed and cost. Settlement is final in a way that a bank transfer is not: once confirmed, the transaction cannot be reversed or recalled, which removes an entire category of uncertainty for the receiver, as our note on crypto settlement describes. And the transfer is transparent and traceable in real time, so both parties can verify status without asking anyone. Measured on this leg alone, the improvement over correspondent banking is not incremental; it is a different order of thing. Which is exactly why the remaining friction deserves the attention.
03 — The last mile is the whole problem
Here is the honest centre of this subject. Almost nobody wants to end up holding a dollar-denominated token. They want local currency in a local account or in their hand. So a real payment has three legs: converting local money into stablecoin at the origin, moving it, and converting stablecoin into local money at the destination. The technology transforms the middle leg and leaves the other two largely as they were.
The consequence is that cost and friction have not been eliminated so much as relocated — from a chain of correspondent banks to a pair of conversion points. And those conversion points can be expensive. Converting into local currency through a local exchange, broker, or agent frequently costs a meaningful percentage, which in some corridors is enough to consume much of the saving from the transfer itself. A transfer that costs a fraction of a percent to send and several percent to cash out is not obviously cheaper than the wire it replaced. This is why claims comparing the network fee to a bank's wire fee are misleading: they compare one leg of the new system with all three legs of the old one.
The blockchain solved the part of international payments that was never really the hard part. Moving value between countries is now trivial; turning it into money someone can spend, where they live, at a fair rate, is exactly as hard as it always was.
04 — Where the remaining cost hides
Four distinct charges make up the real cost of an end-to-end transfer, and only one of them is the network fee. There is the on-ramp spread when local currency is converted into stablecoin. There is the network fee, genuinely small on an efficient chain. There is the off-ramp spread at the destination, typically the largest single component. And there is the exchange rate applied at each conversion, which is where providers most often take their margin quietly rather than in a stated fee.
That last point deserves emphasis because it is the oldest trick in this business and the technology has not touched it. A service advertising zero fees while applying a wide rate on the conversion is charging you as surely as one with a stated commission — the cost is simply invisible unless you compare the rate you received against the prevailing market rate. The only reliable way to evaluate any cross-border service, stablecoin or otherwise, is to compare the amount that arrived in local currency against the amount that left, and ignore every advertised fee entirely.
05 — Business payments are where it already works
The consumer remittance story receives most of the attention, but the clearest adoption is in business-to-business flows, and the reason is that companies escape the last-mile problem in ways individuals cannot. A business receiving payment from an overseas client can hold a dollar balance without converting it, because it has dollar-denominated costs, suppliers to pay, or simply a treasury function that can wait for a favourable rate.
For those users the value is not primarily the fee but the working capital: money that arrives in minutes rather than days is cash that can be deployed immediately, which for a business operating on thin margins across borders is worth considerably more than the transfer cost. Exporters, freelancers and service firms in emerging markets billing clients abroad, marketplaces paying sellers in many countries, and companies managing treasury across subsidiaries are the categories where this has moved from experiment to routine — and surveys of firms in several regions now report majority usage. The pattern is consistent: the use case works best precisely where the recipient does not need to convert immediately.
06 — The compliance layer nobody removes
A common misunderstanding is that moving value on a public network sidesteps the regulatory apparatus that makes conventional transfers cumbersome. It does not, and the reason is that regulation attaches to the on- and off-ramps rather than to the network. Any licensed business converting between local currency and digital assets is subject to identity verification, sanctions screening, transaction monitoring, and reporting obligations — frequently the same ones a bank faces.
Two consequences follow. The compliance cost that made correspondent banking expensive has not disappeared; it has concentrated at the conversion points, which is part of why those points charge what they do. And the regulatory position of the token itself now matters to whether a licensed provider will handle it at all, since frameworks in the major markets restrict what regulated firms may offer — the split described in our note on USDT vs USDC. A corridor is only as usable as the licensed businesses willing to serve both of its ends.
07 — Local liquidity is the binding constraint
Beneath the compliance question sits a more fundamental one: how deep is the market between the stablecoin and the local currency? An off-ramp can only give you a good rate if there is genuine two-sided demand in that pair locally — someone wanting to buy the dollars you are selling, in the size you are selling them.
Where that market is deep, off-ramp costs compress and the end-to-end proposition is excellent. Where it is thin, the local provider is effectively a monopolist, spreads widen, and large transfers move the local rate against the sender — the ordinary consequence of shallow depth described in market liquidity. This explains a pattern that otherwise looks arbitrary: the same technology delivers dramatically different economics in different corridors, and the variable is not the chain or the token but the local currency market at the destination. It also identifies what would actually improve things — not faster blockchains, but more licensed providers competing at the last mile, and integration with the domestic instant-payment systems that many countries now operate, which is where the most credible current work is happening.
08 — An honest assessment
Stablecoins are a genuine and substantial improvement to cross-border payments, and the improvement is largest in exactly the places conventional banking serves worst: corridors between smaller economies, weekend and after-hours transfers, and any situation where the receiver can hold dollars rather than converting immediately. For businesses the case is already made, and adoption reflects that.
The realistic caveats are equally clear. For a consumer sending money to a recipient who needs local cash today, the saving depends entirely on off-ramp competition in that specific corridor, and in some places that saving is modest or absent. Volatility is not a risk here — the token is stable — but the issuer, the network, and the conversion provider each carry their own, and a payment held only briefly is exposed to all of them for that period. The correct expectation is not that international payments have become free, but that one expensive, slow leg of a three-leg journey has been solved, and that the remaining two are ordinary financial-services problems requiring competition and licensing rather than technology. That is a large achievement stated accurately — and stated accurately is the only way it is useful.
"She is like the merchants' ships; she bringeth her food from afar." — Proverbs 31:14
Methodology & Sources
This report describes a use case and its economics as at the date of publication; it deliberately contains no fee percentages, transfer costs, settlement times in precise figures, corridor names, provider names, or adoption statistics — costs and availability vary enormously by corridor, provider, chain, and amount, and readers should compare actual quotes for their own route. The description of correspondent banking — payments routed through chains of intermediary institutions each charging fees, applying compliance checks, and operating on business hours, with margin frequently taken in the exchange rate — reflects the standard structure of international transfers. Characterisations of stablecoin transfer economics are directional: settlement in minutes rather than days, continuous availability, and network fees that are a small fraction of typical amounts on efficient chains, with final and irreversible settlement. The central argument — that the technology replaces only the middle leg of a three-leg transaction, relocating rather than eliminating cost, with on- and off-ramp conversion typically the largest remaining component — reflects widely reported analysis of end-to-end corridor costs. The observation that business-to-business flows show clearer adoption than consumer remittances, because recipients holding dollar balances avoid immediate conversion, reflects reported industry usage. Compliance obligations attaching to licensed conversion providers rather than to the underlying network, and the dependence of off-ramp pricing on local currency market depth, are structural observations. Nothing here is a recommendation regarding any token, provider, corridor, or transfer, and no claim is made that any particular route is cheaper than an alternative.