Cross-border payment costs: banks, fintechs, stablecoins

A new BIS paper, published on 2026-10-19, puts a price on sending money abroad through three channels: retail bank wires, fintech money transfer operators and stablecoins. Its answer is that fintech costs are much lower than bank wires, while stablecoins carry their own charges. This article looks at the plumbing behind each rail and at where the money actually leaks out.

What the BIS paper found

The BIS researchers built what they call harmonized all-in cost measures. They combined intraday pricing from fintech providers, World Bank remittance figures, bank wire transactions and crypto market data, so that costs could be compared by currency and by transfer size. The bank channel in the study runs through correspondent banking and SWIFT. The fintech group includes firms such as Wise, Remitly and the digital business of Western Union.

The headline results are plain. In a small sample of bank wires, the foreign exchange markup alone sat at approximately 2 to 4.7%, even for very liquid currencies, and before any wire fee. Fintech transfers were much cheaper: for $10,000 sent from the United States, median costs for developed market currency groups were around $40–$60, though some destinations stayed far more expensive. On stablecoins, the authors are cautious. In their words, "stablecoins are not a panacea for frictions in cross-border payments".

How this works: the cost stack of a transfer

A transfer abroad is not one payment. It is a chain of steps, and each step can take a slice. The useful habit, and the one the BIS paper follows, is to split the price into two kinds of charge. An explicit fee is the number shown on the receipt. An FX markup is the gap between the exchange rate you get and the market rate at that moment. The markup is quiet, because it is folded into the rate rather than shown as a fee.

• Bank wire: your bank converts the currency and passes the payment along a chain of correspondent banks, which hold accounts with each other and exchange instructions over SWIFT. The markup and the wire fee both sit here.
• Fintech operator: the BIS found its cost falls sharply as a share of the amount sent when transfers get bigger. The paper does not break down the firm's internal steps, so treat the exact route as firm-specific.
• Stablecoin route: dollars go into a crypto exchange, become a stablecoin, move on a blockchain, then are sold for local currency and withdrawn to a bank. Each hop can carry a trading fee, a blockchain transaction fee or a withdrawal fee.
• On-chain FX: one version swaps a dollar-linked coin for a coin tied to another currency. The price then depends on how much trading depth that pair has.

Read this way, the question for any rail is simple: how many hands touch the money, and how much does each one keep? The blockchain leg is only one step. The BIS points to the conversions before and after it as a source of extra cost.

Why stablecoins are not a shortcut

The BIS points to three frictions. First, the on-ramp and off-ramp. Moving from bank money into a stablecoin and back again still relies on crypto exchanges and other intermediaries, and their fees can push the total up substantially. The recipient may also wait while the coin is converted and the bank processes the deposit.

Second, liquidity. The paper finds very thin trading depth when US dollar stablecoins are converted into currencies other than the dollar. Thin markets mean a large order moves the price against the sender, which is a cost even when no fee is charged. Shallow depth and volatile prices also cap how big these transfers can get. Because prices can move, a sender can lose money on a stablecoin transfer, not only pay fees.

Third, the cases where stablecoins look cheaper. On some routes from developed to emerging economies, stablecoins showed better FX pricing. The authors say this reflects distortions from capital controls, not an advantage of the technology. Put simply, the stablecoin price can differ from the official rate because the official channel is restricted, and that gap is a policy effect rather than an engineering one.

What changes, and for whom

For people sending money home, the practical lesson is to compare the all-in cost, not the advertised fee, and to remember that small transfers are hit hardest. The paper notes that sending money abroad remains expensive, especially for small amounts, even after decades of better domestic payment systems.

For policymakers, the paper shifts the debate. It points attention to currency conversion, intermediary fees and market liquidity, which the study measures separately from the transfer itself. The authors conclude that stablecoins can compete only with market liquidity, dependable fiat access, regulatory compliance and trust in intermediaries, not blockchain alone.

Open questions

The bank sample is small, as the authors say, so the markup range may not describe every bank or corridor. It is also unclear which destinations make up the more expensive fintech routes, and how stablecoin costs would change if exchange fees fell or local currency pairs gained depth. Finally, the capital-control effect raises a harder question: if a cheaper route depends on stepping around a restricted official channel, how long can that price last, and who carries the legal risk?

Sources

• BIS, 19 Oct 2026

How this post was made: an AI model drafted it from the sources listed above, and software checks ran before it went live without a human edit. Every figure has to match a listed source and no sentence is copied. It explains how the mechanism works; it is not investment advice.