Cross-Border Remittances: Why Sending Money Home Still Costs So Much
Transfer fees, FX markups and receiving-side charges build a hidden cost stack; here is how it forms, where it is worst, and what is finally pushing it down.
Key takeaways
- The World Bank's Q3 2025 average cost of sending $200 was 6.36%, more than twice the UN's 3% target for 2030
- Fees are only part of the stack: the exchange-rate margin is the line most senders never see
- Banks average 14.99% per transfer against 3.54% for digital-only money transfer operators, so provider choice moves the price more than the destination does
- Instant payment rules, disclosure mandates and stablecoin settlement are squeezing the middle legs, but the cash last mile still sets the floor
Why a remittance rarely costs what the sign says
Migrant workers sent an estimated $685 billion to low- and middle-income countries in 2024, a flow larger than foreign direct investment and official development assistance combined, according to the World Bank. The price of moving that money is not one number but a stack: a sending fee, a currency-conversion margin and whatever the receiving side deducts before the family collects the cash.
The reference dataset for that stack is the World Bank's Remittance Prices Worldwide (RPW). Its cost measure is the charge incurred by the sender — the transfer fee plus the margin between the rate the provider offers and the interbank rate, expressed as a percentage of the amount sent (RPW Issue 54, Q3 2025 report). In the third quarter of 2025 the global simple average was 6.36% of a $200 transfer, or $12.72 lost per transfer. Services that can be started online or in an app and paid into an account averaged 4.59%, against 7.30% for everything else, and digital services accounted for 35% of the services the survey priced (RPW Issue 54).
The trend line is the uncomfortable part. Sustainable Development Goal 10.c calls for the average cost of a remittance to fall below 3% by 2030 and for corridors costing more than 5% to be eliminated (United Nations); an earlier G8 and G20 commitment had set 5% as the interim benchmark (RPW Issue 54). RPW recorded 9.67% in the first quarter of 2009 and 6.35% in the first quarter of 2024 (RPW Q1 2024 report). Q3 2025 came in at 6.36%. Sixteen years, roughly three percentage points, and no movement worth the name in the last two years.
Channel matters more than corridor
Most senders assume the destination drives the price. The data says the type of provider matters at least as much. Global averages for sending $200 in Q3 2025, all from RPW Issue 54:
| Measure (cost of sending $200) | Q3 2025 average |
|---|---|
| Banks | 14.99% |
| Non-digital services | 7.30% |
| Global simple average | 6.36% |
| International MTO Index | 5.52% |
| Global weighted average (weighted by corridor flow size) | 5.04% |
| Digital services | 4.59% |
| Digital-only MTO index | 3.54% |
| SmaRT (three cheapest qualifying services per corridor) | 3.29% |
Three numbers in that table matter more than the rest. Banks at 14.99% are more than four times as expensive as the five digital-only operators in RPW's index, who averaged 3.54% — the same $200, moved by different plumbing. The global weighted average of 5.04% sits almost 1.3 points below the simple average precisely because the largest corridors are the competitive ones, so the headline figure overstates what a typical sender on a busy route pays (RPW Issue 54). And the SmaRT measure — the three cheapest qualifying services in each corridor — is already within a whisker of the 2030 target at 3.29%, which means the average sender is not paying a market floor. They are paying a comparison gap.
The three layers of the cost stack
1. The sending-side fee
This is the visible number: a flat charge or a percentage, applied before anything moves. Regulation in the United States now forces it into the open. Under the CFPB's Remittance Rule, a provider must disclose before the sender pays the fees imposed, any taxes collected, the exchange rate used, any covered third-party fees and the amount the recipient will actually receive in the receiving currency (12 CFR 1005.31). A receipt with the same information follows, together with error-resolution and cancellation rights. The disclosure regime does not cap the price — it makes the price checkable.
2. The FX margin
This is where the money hides. A provider can advertise "zero fee" and still earn several percent by converting at a rate worse than the interbank rate. RPW exists precisely because fee-only comparisons mislead: its headline cost combines the fee with that implied margin. In practice the two are substitutes, and the CFPB has warned that marketing a transfer as "no fee" or "free" when the provider recovers the cost elsewhere can be a deceptive practice (CFPB Circular 2024-02). The practical rule for a sender is mechanical: a fee-free offer at an off-market rate can be strictly more expensive than a 1% fee at a near-market rate.
3. The receiving-side charge
The last leg is the least transparent. The US framework distinguishes fees the provider knows about ("covered third-party fees") from charges imposed by parties it does not control — a correspondent bank, a local bank, a mobile money agent or a payout partner (12 CFR 1005.31). Where the payout is cash through an agent network, the agent's commission is often baked into the rate rather than itemised. That is one reason the same app can look cheap on the sender's screen and expensive at the kitchen table: the household receives less than the sender was shown. Note also what RPW's 6.36% does and does not cover — it is a sender-side measure of fee plus exchange-rate margin, so deductions applied at the payout end sit outside the headline number.
Geography still sets the floor
Regional averages diverge sharply. Sub-Saharan Africa remained the most expensive receiving region in Q3 2025 at 8.46% for a $200 transfer, while the Middle East, North Africa, Afghanistan and Pakistan group was the cheapest at 5.11% (RPW Issue 54). The pattern has a structural cause: corridors that rely on cash distribution, thin correspondent coverage and currency controls carry more cost and more risk per transaction than corridors with deep digital payout networks.
Corridor pricing is two-sided as well. Sending-country costs differ: flows from the United States averaged 5.04% and from Germany 4.49% in Q3 2025, while Italy, with a less competitive outbound market, averaged 7.44% (RPW Issue 54). Comparing by corridor rather than by brand is the more useful habit — which is what the RPW database is built for.
Instant payment rails squeeze the middle legs
The most substantive cost pressure comes from domestic fast payment systems, and the EU has legislated the largest piece of it. The Instant Payments Regulation (IPR) applies to euro credit transfers where the payer's and the payee's providers are both in the EU, and it sets a staggered timetable. Under Article 5a(8) of Regulation (EU) No 260/2012, as inserted by Regulation (EU) 2024/886, euro-area providers had to be able to receive instant euro transfers from 9 January 2025 and to send them from 9 October 2025; providers in non-euro-area member states must receive by 9 January 2027 and send by 9 July 2027; payment institutions and e-money institutions in the euro area have until 9 April 2027 to do both, with their non-euro-area peers receiving from 9 April 2027 and sending from 9 July 2027 (Regulation (EU) 2024/886; ECB timetable).
The pricing rule is the interesting part. Article 5b(1) requires that any charge a provider levies for sending or receiving an instant credit transfer be no higher than the charge for other credit transfers of corresponding type — in force in the euro area since 9 January 2025 and elsewhere from 9 January 2027 (Regulation (EU) 2024/886; ECB). That removes the instant surcharge as a revenue line, but it says nothing about the conversion step, which is where the FX margin is taken.
Read against the three-leg stack, the mechanism is straightforward: the regulation covers euro transfers between EU providers, so it acts on the collection and payout legs when both ends sit inside the scheme, while leaving the conversion leg untouched. That is our reading of how the rule bites, not a World Bank cost estimate — RPW does not isolate the effect of instant payment mandates on corridor prices. It also explains why the domestic-leg story differs from the cross-border one, covered in our analysis of cross-border instant payments and the underlying payment rails. Outside the euro scheme, transfers still ride correspondent chains, where tracking, cut-off times and screening add friction — background we covered in SWIFT gpi and the ISO 20022 migration.
Stablecoin rails: cheaper middle, same last mile
Stablecoin-based corridors remove the correspondent bank from the middle of the chain: value moves on a public ledger in minutes, for a network fee that is a fraction of the amount. The saving, however, is structural rather than automatic, and it lands on the intermediate leg. A consumer still needs an entry point (an exchange or a licensed wallet) and an exit into local currency, and that exit is frequently the same bank, wallet or agent network the legacy corridor uses. The expensive leg — local payout, identity checks and cash handling — does not disappear because the transfer leg got cheaper.
Two further caveats. First, the off-ramp spread is the new FX margin: converting a dollar stablecoin into local currency at a non-bank venue carries its own markup, which can be less transparent than a disclosed bank margin. Second, the regulatory perimeter is tightening rather than loosening. EU issuers of e-money and asset-referenced tokens sit under the MiCA framework, with authorisation, reserve and redemption requirements, and the US has legislated a federal framework for payment stablecoin issuers through the GENIUS Act, Public Law 119-27; both are covered in our comparison of stablecoin rules in the EU and US. Bank-issued tokenised deposits compete on the same corridor with a different risk profile.
For operators, the practical use of these rails is at the treasury layer: pre-positioning liquidity across corridors rather than holding idle balances in local banks can cut settlement time and working-capital cost. That is a gain measured in an operator's balance sheet, and it shows up in the consumer price only if competition passes it on.
Regulation is turning cost into disclosure
Two rulesets now do most of the work of making the stack visible.
In the EU, the cross-border payments regulation requires providers to disclose the total currency-conversion charge as a percentage markup over the latest available European Central Bank reference rates for conversions involving the euro — the same unit of measurement across providers, rather than a rate the customer cannot audit (Regulation (EU) 2021/1230).
In the US, the CFPB's Remittance Rule mandates pre-payment and receipt disclosure of fees, the exchange rate and the amount received, plus cancellation and error-resolution rights (Regulation E Subpart B).
In both jurisdictions, disclosure is separate from marketing: a provider can satisfy every disclosure requirement and still face liability for calling a transfer instant when it is not, or free when it is not.
What this means for fintech builders
The revenue pool is real: at RPW's Q3 2025 averages, the gap between a bank transfer at 14.99% and a digital service at 4.59% is more than ten percentage points on the same $200. But a headline price of 5% is not a margin. From that percentage an operator must fund the payout or prefund it, hedge FX between quote and settlement, pay local payout fees, hold licensing and safeguarding capital, run sanctions and travel-rule screening, absorb fraud losses and chargebacks, and staff support. Two structural facts shape who survives:
- Corridor depth beats headline price. Deep, competitive corridors compress toward 3% because several providers can settle without touching a correspondent bank. Thin corridors cannot, which is why the most expensive routes cluster where cash distribution and bank de-risking dominate.
- "Zero fee" is a pricing choice, not a cost saving. The cost does not vanish; it moves into the conversion margin, where disclosure rules are now shining a light.
Tools that cut operating cost per transaction — routing and orchestration across rails and payout partners, reconciliation, liquidity management — are the ones that translate into a lower consumer price without destroying the unit economics. Our walkthrough of payment orchestration platforms covers that layer in detail.
Risks and costs
For senders. The advertised fee is not the cost: the exchange-rate margin, receiving-side deductions and any local agent commission must be added, and only a full-recipient-amount comparison reveals the real price. Where the provider cannot fix the amount at the time of disclosure — an exchange rate not yet known is the standard case — Regulation E permits it to disclose an estimate rather than a final figure (12 CFR 1005.32), and the same subpart sets out error-resolution rights if what arrives differs from what was disclosed (12 CFR 1005.33). Reversals and recalls can carry additional charges, and sanctions or AML holds can delay funds with no compensation. Informal channels may look cheaper but generally offer no receipt, no legal recourse and no reimbursement route. Person-to-person impersonation fraud targets remittance users specifically, and a completed cash payout is close to irreversible.
For builders. Pricing at 4%-5% leaves limited room for compliance, licensing (payment institution or money transmitter status in each jurisdiction), safeguarding of customer funds, prefunding capital, FX risk and fraud losses. Correspondent bank access can be withdrawn at short notice, which is the single largest concentration risk for a corridor-dependent business. Disclosure is a floor, not a ceiling: rules move at different speeds across jurisdictions, and a model built on an undisclosed margin is exposed to the next transparency rule. Stablecoin rails add technology, custody and redemption risks of their own, and do not remove the licensing obligation at either end.
Bottom line
The cost of sending money home is a stack of three — sending fee, FX margin and receiving-side charge — and the World Bank's Q3 2025 average of 6.36% for a $200 transfer shows how little the headline has moved since 2009's 9.67%. Underneath the average the picture is more encouraging: digital services average 4.59%, the digital-only money transfer operators RPW tracks average 3.54%, and the three cheapest qualifying services in each corridor average 3.29%. The distance between 3.29% and 6.36% is the price of not comparing. Instant payment rules in the euro area, mandatory cost disclosure on both sides of the Atlantic and cheaper settlement on the middle leg are all real pressure on legacy pricing — but the cash last mile and the currency conversion step remain the two lines where a sender's money still quietly disappears.
Follow the shortlist
Following payments, banking and market-infrastructure change? Our global ETF and stock research watchlist is tracked daily. Premium members get the full list, the reasoning behind every entry and updates when entries change.
See premium accessInformation and education only, not investment advice.
Related
Stablecoin Regulation Compared: EU MiCA vs the US GENIUS Act
How MiCA and the GENIUS Act differ on who may issue payment stablecoins, what must back them, how redemption works and where depeg risk still sits.
FinTechTokenised Deposits vs Stablecoins: How Bank-Issued Money Differs
Tokenised bank deposits and reserve-backed stablecoins both move value on programmable rails, but they differ in legal claim, deposit insurance, balance-sheet treatment and risk.
FinTechInstant Payments for Small Business: FedNow, RTP, Zelle Compared
How FedNow, RTP and Zelle differ for US small businesses on speed, settlement finality, limits and cost, plus which emerging options fit SME cash flow.