Stablecoin rails vs money transfer operators: how each one cuts the cost of a cross-border payment

A money transfer operator cuts cost by netting customer flows and paying out from local balances it pre-funds in each country. A stablecoin rail cuts cost by moving value as a token, with no pre-funded local balance, then converting at an off-ramp. The BIS found MTOs already compressed costs to 40 to 150 basis points on larger transfers. Stablecoins compete on settlement and pre-funding, not on magic.

Both models exist because the old route is slow. The Federal Reserve's March 30, 2026 FEDS Note counts over 60% of wholesale payments routed through one or more intermediaries, and about 30% fewer active correspondent banks over the decade to 2022. Both models sit inside the same basic shape: money in, value across, money out. The stablecoin sandwich explains the stablecoin version layer by layer. This page puts it next to the fintech model most businesses already know.

Key takeaways

• MTOs such as Wise, Remitly, and Western Union's digital arm bypass correspondent banks by holding local accounts and netting flows internally (BIS Working Paper 1385, October 2026).
• MTO all-in costs fall from about 200 to 300 basis points at $200 to about 40 to 130 basis points at $10,000, because fixed fees shrink as a share of the transfer.
• Stablecoin rails need no pre-funded local currency, but they move the liquidity job to exchanges, market makers, and off-ramps.
• Stablecoin to fiat liquidity is thin outside a few currencies, which caps how much can move at a good price.
• The two models combine. MTOs can run stablecoins in the back end.

What is a money transfer operator, and how does it avoid correspondent banks?

A money transfer operator is a non-bank that pays recipients from local money it already holds in the destination country. The BIS paper by Wenxin Du, Catherine Huang, and David Scharfstein (Working Paper 1385, published October 2026) describes the model in three moves.

First, the MTO keeps local accounts in each country, usually through partner banks. Second, when a customer sends from country A to country B, the MTO pays out of its balance in B. Nothing crosses a border at that moment. Third, it nets flows: money going one way offsets money going the other, and only the leftover imbalance gets settled through wholesale FX markets or a bank transfer.

That is why MTOs work best in high-volume corridors. More flow in both directions means more netting and cheaper pre-funding.

The catch, per the paper: MTOs still rely on correspondent banks for rebalancing, local banks for payout, and traditional FX markets for hedging. They re-optimize the old architecture. They don't leave it.

How does a stablecoin rail move the same payment?

A stablecoin rail converts the sender's money into a dollar stablecoin, moves the token on a public blockchain, and converts it into local currency on arrival. The BIS writes it as fiat A to stablecoin, on-chain transfer, stablecoin to fiat B.

The difference that matters for a finance team: no local currency waits in the destination country. The paper states it plainly, the stablecoin rail "does not require pre-funding of local currency liquidity." The conversion happens at the off-ramp, priced at that moment.

What it adds instead is a set of new intermediaries. Exchanges, OTC desks, or payout providers run the on-ramp and off-ramp, and the BIS says those exchanges "effectively perform functions analogous to correspondent intermediaries." The step-by-step payment flow shows where each one sits.

How do MTOs and stablecoin rails compare?

They solve the same problem with different balance sheets. This table summarizes the BIS paper's comparison in section 7.

Which one is cheaper, according to the BIS?

It depends on the size and the currency. The BIS measured all-in cost as FX markup plus explicit fees, benchmarked against Bloomberg rates.

For MTOs, the authors collected quotes from three major providers every 30 minutes during January 23 to 30, 2026, on US-outbound corridors:

Source: BIS Working Paper 1385, section 5.2 and Figure 1. The paper also notes that PLN, PHP, and several Latin American currencies stayed above 150 to 250 basis points even for large transfers.

For stablecoins, the FX leg for major currencies tracks benchmark rates closely. The costs show up elsewhere. Trading fees on the leading centralized exchanges in the sample ranged from 0 to 60 basis points for low-volume traders, with one charging 100 basis points under $10,000 of monthly volume. Fiat withdrawals often carry fixed fees, such as €1 for a SEPA withdrawal in the paper's example. Gas fees averaged well below $1 even on Ethereum.

So the honest answer: for a large transfer between major currencies, a good MTO and a well-run stablecoin route can land close to each other. For a small transfer, fixed fees decide it. For a Latin American currency, the corridor's liquidity decides it. The cost per delivered dollar method is how you compare two real quotes.

Which one is faster?

Both can deliver in minutes in a major corridor. The reasons differ.

MTOs are fast because "the underlying funds do not need to move cross-border at the time of the transaction," in the BIS's words. Settlement happened earlier, through liquidity positioning.

Stablecoins are fast because the token settles on-chain in minutes. But the paper stresses that end-to-end speed "depends critically on the efficiency of fiat on- and off-ramping at both ends." It gives one example: Circle charges an expedited fee for same-day USDC redemptions, while standard redemptions settle T+1. On the other side, the World Bank's Remittance Prices Worldwide database, which the BIS also uses, tracks what traditional providers charge per corridor.

After either model, the local rail sets the final clock. Pix and SPEI land in minutes. ACH takes about 1 to 2 business days. Settlement times by rail has the full table.

Where does each model struggle?

MTOs struggle where netting is hard. A corridor with flow mostly in one direction forces the MTO to keep refilling local accounts through banks and FX markets. That costs money, and emerging-market corridors stay more expensive as a result.

Stablecoins struggle on depth. The BIS found the USD stablecoin to euro pair, the most liquid one, averaged about $186 million a day on centralized exchanges over 2021 to 2025. EUR FX instruments turned over about $2.7 trillion a day in 2025. Beyond the euro, only KRW, TRY, BRL, GBP, and THB averaged more than $10 million a day. Thin markets mean volatile spreads and slippage on large orders. Onchain liquidity vs off-ramp liquidity explains why a deep token pool doesn't fix this.

Stablecoins also ask more of the user. Exchange accounts, wallets, gas tokens, the right network. The paper calls this "the greatest operational complexity on retail users." That is the gap payment providers fill.

Can a money transfer operator use stablecoin rails?

Yes, and the BIS expects it. The paper says stablecoin rails "can be used by retail users but also by fintech money transmitters themselves," and that fintechs can build on stablecoins "while abstracting away much of the underlying operational complexity."

In practice that looks like an MTO keeping its app, pricing, and support, and swapping the back end for specific corridors. It can fund payouts per transaction instead of parking local balances, or use stablecoins to rebalance accounts it already has. Building a remittance app on an on/off-ramp API walks through that architecture.

How should a business choose?

Start from the corridor and the size, not the technology.

1. High-volume corridor between major currencies, retail sizes: an MTO is hard to beat on price and experience.
2. Corridor where pre-funding local balances ties up capital: a stablecoin rail removes the pre-funded balance. See what no pre-funding means.
3. Weekends, holidays, or 24/7 treasury needs: the on-chain leg doesn't keep bank hours, though the local rail might.
4. Very large tickets in a thin currency: test liquidity first with a quote ladder.
5. You are the MTO: compare your cost of pre-funding a corridor with per-transaction conversion on a stablecoin rail.

How does BlindPay handle this?

BlindPay is the stablecoin rail behind the payout. A business or an MTO sends USDC or USDT and BlindPay pays local currency over Pix, PIX Safe, TED, SPEI, ACH Colombia, Transfers 3.0, ACH, RTP, domestic wire, SEPA, Faster Payments, and SWIFT (POBO/COBO) with UETR tracking and MT103 confirmations. No local balance has to be pre-funded. Each payout quote locks the rate and fees for 5 minutes, and BlindPay combines a monthly plan with per-transaction fees that vary by rail and network, disclosed in the live quote before execute. KYC and KYB run inside the same API.

What to do next

Pick your largest corridor and price one real transfer both ways at the same minute: your current MTO quote and a stablecoin payout quote for the same amount. Compare what the recipient receives. On BlindPay, the payout quotes docs show the request.