An overseas payment described as a $25 wire can reach the recipient $60 short—and the sender may still face a foreign-exchange margin. The discrepancy is not necessarily a processing error. A cross-border payment can involve separate charges from the sending bank, intermediary or correspondent banks, the beneficiary bank, and whoever exchanges currencies. To compare prices fairly, calculate the recipient’s spendable money after the entire chain, not the price printed beside the ‘Send’ button.
SWIFT sends messages; banks move and price the money
SWIFT is a financial messaging network, not a consumer bank and not a vault that receives the payer’s dollars. Its own material explains that the system securely delivers instructions while account entries and cash movements take place at financial institutions [1]. If two banks lack a direct settlement relationship in the relevant currency, one or more correspondent banks may be involved. They can maintain nostro/vostro relationships and apply their own charges under agreements that the retail customer does not see.
This distinction changes who is responsible for a short payment. SWIFT does not decide the retail bank’s advertised fee, the correspondent’s deduction, the destination bank’s inward-wire charge or the exchange rate the bank quotes. SWIFT operates network standards, secure communication and services such as gpi tracking. Calling every charge a ‘SWIFT fee’ hides the commercial decision of the bank or payment provider that actually levied it [1][2].
A typical payment chain contains an originator, sending bank, one or more intermediaries and a beneficiary bank. Some routes have more steps; others use direct relationships or local clearing. There is no universal $15 correspondent charge or guaranteed number of correspondent banks. Wise says its estimates of correspondent deductions commonly fall between $15 and $50, but charges depend on the route and may be higher [3]. The bank pricing the transfer may also retain an FX spread that dwarfs all fixed charges.
OUR, SHA and BEN control the fee allocation—but not always the outcome
| Instruction | Intended fee allocation | What recipient may experience | Caveat |
| OUR | Sender pays sending and associated bank charges | Generally aims for full principal delivery | Extra upfront fees; routing and local systems may still produce deductions |
| SHA / SHAR | Sender pays its own bank; recipient bears others | Correspondent and receiving fees may reduce net amount | Payment amount in message can differ from credit |
| BEN | Beneficiary bears charges | Several bank fees may be deducted from principal | Sender’s bank may restrict use; practices vary by route |
Wise’s help pages say its own provider can receive less than the instructed payment under SHA where intermediaries deduct fees, while an OUR instruction is intended to shift such costs to the sender [4]. A crucial primary-source counterexample comes from HSBC Egypt. Its published tariff warns that even when a transfer is initiated as OUR, US clearing systems such as Fedwire or CHIPS may not recognize that charge designation, and correspondent banks can change it to SHA or BEN during routing. HSBC also publishes a flat USD 25 correspondent fee when clients elect to pay related fees [5].
Therefore ‘select OUR and the beneficiary is guaranteed every cent’ is not a robust universal rule. It is a cost-allocation instruction operating inside a complicated correspondent chain. It is generally appropriate to request OUR when paying an invoice requiring a precise amount, but customers should request the bank’s full quote and any supported full-value or guaranteed-amount service rather than equating the code with a legally universal guarantee.
The $1,000 invoice: which number matters?
Suppose a freelance designer invoices a foreign client for $1,000 and requests receipt in dollars. Three hypothetical routes illustrate the practical consequence. Route A costs the sender $25 for an outward wire, while a correspondent deducts $20 and the beneficiary bank deducts $10; the designer receives $970. Route B charges $25 plus $35 in bank and correspondent charges to the sender and achieves a full $1,000 credit; the sender pays $1,060 in total. Route C uses a provider charging a combined $12, including known payout cost, and lands $1,000 with no receiving fee; the sender pays $1,012.
These figures are controlled assumptions, not simultaneous live quotations from named providers. The correct comparison depends on *who has promised whom what*. If the commercial obligation is a net $1,000 receipt, Route A has failed to fund the invoice in full; the client may need a follow-up $30 payment, which itself may incur charges. If the payer merely promised to send $1,000 with shared bank costs, the contractual treatment can differ. Cost is not just the difference between $25 and $12, but the sum of all sender-paid costs plus any shortfall the sender must make good.
FIGURE 1 | Modeled total cost to satisfy a net $1,000 invoice. Route A requires sender to make up a $30 shortfall; figures are illustrative.
Now add currency exchange. A $10,000 payment converted by a bank at a rate effectively 1.5% worse than a chosen benchmark loses roughly $150 in purchasing-power terms before fixed wire charges. A $25 outward fee plus $20 correspondent fee plus $10 inward fee brings the illustrative total economic drag to $205, or 2.05% of the transfer principal. At $1,000, those same fixed charges plus a 1.5% exchange spread cost $70, or 7% of value. At $50,000, fixed fees largely disappear as a percentage, while the FX spread can remain material at $750.
Those calculations deliberately separate the bank’s exchange-rate markup from independently levied charges. A mid-market reference is not a promise that a retail consumer can trade currency at that rate; it is a transparent benchmark for measuring the price of conversion. A provider that quotes a smaller outward charge but worse FX can be more expensive, particularly for large transfers.
FIGURE 2 | Illustrative $10,000 cross-border cost stack: assumed 1.5% FX margin plus $55 in named fixed bank charges.
Verified examples show how heterogeneous fees really are
Wise publishes receipt fees for certain international Swift payments, including a displayed USD 6.11 and EUR 2.39 receiving charge at the time of review, and explicitly warns that correspondents may deduct charges as well [6]. Its own USD-outside-US help page provides destination-specific payout costs rather than one uniform international-wire rate. For selected routes, its published modeled charges include predicted correspondent costs and can differ considerably by recipient country [7]. Those figures refer to particular Wise flows, account arrangements and dates—not an average for the world’s banks.
HSBC Egypt’s official tariff is another concrete example. Its foreign-currency online outward transfer schedule for several retail segments includes a percentage fee with a floor and cap plus a cable fee denominated in EGP, alongside potential correspondent charges in USD [5]. A direct conversion of the EGP tariff into dollars would require an exchange rate on the execution date, which we do not invent. What matters analytically is that a seemingly simple foreign-currency payment can combine a capped percentage, a fixed local-currency charge and separate dollars taken for correspondent routing.
For most consumers the key distinction is same-currency versus cross-currency routing. USD paid to a beneficiary’s USD account may still incur intermediaries and incoming charges, but it need not involve the beneficiary bank selling the received dollars for another currency. USD sent into a local-currency account creates an additional FX decision and may be automatically converted at the receiving bank’s terms. Wise specifically warns that recipient account currency can create avoidable conversions [8].
Why two identical wires may arrive with different amounts
Correspondent routing is not necessarily fixed for every transaction. Cutoff times, correspondent availability, country, sanctions screening, liquidity, payment urgency, local holidays, beneficiary-bank arrangements and maintenance can influence the path and timing. Wise says correspondent banks on a route can vary, and therefore deductions may change [3]. Even when the bank sends the same currency and amount twice, a different intermediary can affect what the recipient sees.
Returned wires can magnify that uncertainty. A wrong beneficiary name, inaccurate IBAN, compliance hold or account restriction can cause payment rejection. When funds come back, some processing and correspondent fees may not be reversed. Wise acknowledges that returned Swift transfers can arrive back with less money than originally sent because fees were deducted in transit [3]. If the payment was converted, the recipient or sender may face a second conversion at a new rate. The value at risk is more than simply the original outward fee.
Timing is a second layer of cost. Wise gives an estimated one-to-six-working-day window for some Swift payments but warns that time zones and multiple intermediaries can extend processing [4]. For someone paying tuition, rent, a securities margin call or an invoice with a contractual deadline, a cheap but unreliable route has an implied economic cost. It can be rational to pay for a trackable or full-value service even when the headline fee is higher.
Tracking: UETR and gpi make the chain less opaque
SWIFT introduced a Unique End-to-End Transaction Reference (UETR) that functions like a tracking number, allowing banks and participants to identify a payment as it moves through the chain. SWIFT describes the UETR as a 36-character reference present in Swift payment instructions and central to gpi payment tracking [9]. SWIFT says its gpi service offers end-to-end visibility, confirmation and information about deductions [2]. These facilities improve inquiry quality even though they do not eliminate correspondent charges.
If a beneficiary reports that $1,000 was sent but only $964 arrived, ask the sending bank for the transaction’s UETR, SWIFT payment confirmation, charge option and, where available, the gpi trace of deductions and value dates. The receiving bank can identify its own incoming fee. The sender should distinguish deductions made en route from the beneficiary bank’s separate posting fee. Without those details, a complaint that ‘SWIFT stole $36’ is not attributable to any responsible institution.
A key 2026 development is that ISO 20022 is now the cross-border payment message standard. SWIFT says the coexistence period with legacy MT payment instructions ended on 22 November 2025 [10]. Richer structured payment data can help reconcile invoices, reduce manual repair and improve compliance processing. But better data does not automatically mean banks stop charging for liquidity, compliance, FX or correspondent relationships. The material potential gain for retail customers is fewer payment repairs and clearer information, not an abolition of the fee chain.
The scale effect: why cheap wires are rare for small transfers
A fixed $55 total routing burden equals 11% on a $500 payment, 5.5% on $1,000, 1.1% on $5,000 and 0.11% on $50,000. Adding a constant 1.5% FX spread produces effective burdens of 12.5%, 7%, 2.6% and 1.61%, respectively. The result explains why cross-border banking fees can devastate small remittances even when a corporate sending $100,000 barely notices the flat charge. It also explains why price comparisons should be run at the reader’s *actual* transfer size.
FIGURE 3 | Effective percentage cost from $55 fixed routing charges and a 1.5% FX benchmark shortfall across four transfer sizes.
A retailer or freelancer should optimize for net receipt per currency unit, not for the nominal outward fee. A $9 fee with $40 of hidden deductions may be less competitive than a clearly priced $25 service that ensures the correct net amount. However, a non-bank payment company is not always cheaper on every corridor, and some routes still use the Swift rails beneath the interface. Test real quotes with identical inputs: amount funded, funding method, recipient country and bank, destination currency, guaranteed or estimated amount received, and estimated arrival time.
What can a customer actually control?
The customer can control correct recipient details, the charge election offered by the sending bank, destination account currency, whether an alternative domestic or regional scheme exists, sending time relative to bank cutoffs and whether to fund a quoted FX conversion. The customer cannot directly set a third-party correspondent’s tariff. Sending in the beneficiary’s preferred currency can prevent unnecessary recipient-side conversion, but only after comparing the sender bank’s conversion quote with alternatives.
For transfers inside eligible European corridors, SEPA may be the more relevant rail than an international Swift wire. Within one country, domestic account-to-account schemes can avoid correspondent charges entirely. For large or time-critical transfers, a bank’s traceability, creditworthiness, recall procedures and full-value service may matter more than saving a few dollars. The wrong transfer reference or currency instruction can cost more than the difference between fee schedules.
What consumer law changes—and what it does not
For qualifying US consumer remittances, the Consumer Financial Protection Bureau says providers generally must disclose important costs, exchange rates and the expected amount to reach the recipient, with certain permitted estimates and warnings for other receiving-bank charges [11]. A sender commonly has a 30-minute cancellation right before payout or deposit, while other error-resolution protections apply under stated deadlines [12]. That does not automatically make a business-to-business wire or a payment from another jurisdiction subject to the same provisions. Nor does a printed estimate remove the possibility of fees expressly excluded from the disclosure’s coverage.
The CFPB has specifically warned that advertising international transfers as ‘free’ or ‘instant’ may be deceptive when FX spreads, recipient withdrawal charges or actual timing make the claim misleading [13]. The analytical lesson applies well beyond the US legal rule: consumers should ask what the recipient can *spend* after withdrawal and conversion. The difference between ‘no sending fee’ and ‘no transfer cost’ is not semantic. It can exceed the bank’s visible processing charge by several multiples.
For an invoice payer, a more reliable control is to ask the beneficiary to confirm the correct receiving currency, banking instructions and whether intermediary deductions must be grossed up. The contract should say whether the sender’s obligation is to dispatch the face amount or ensure a specified net amount arrives. Otherwise both parties can believe they performed correctly while a correspondent charged $25 in between.
Bottom line
The party responsible for SWIFT-related costs is determined partly by the OUR/SHA/BEN instruction and partly by the commercial and technical realities of the route. For $1,000 sent abroad, do not stop at the sender’s displayed charge. Track origin fees, correspondent deductions, recipient charges and currency conversion, then compare the exact amount delivered—or the total amount needed to satisfy the invoice. ‘SWIFT’ is the communication layer; the economics sit with the banks and payment providers along the path.
Methodology
All public bank, payment-provider and SWIFT material was checked as at 10 October 2026. Dollar scenarios are original illustrations, not fee quotations or regulatory benchmarks. The $1,000 example models three mutually exclusive routes. FX margin is defined as the percentage shortfall from a stated comparison benchmark; it is an assumption and not attributed to a named bank. Percentage drag is total modeled cost divided by payment principal; the $55 fixed-cost sensitivity also assumes a separate 1.5% exchange-cost component. Real quotes vary by account, corridor, date, transfer method and bank. Currency-dependent fees are not converted without an actual exchange rate.
Sources and Primary Documentation
Source list checked 10 October 2026. Bracketed numbers in the article refer to entries below. Hyperlinks open the original publishers.
[1] SWIFT: messaging services and transfer of funds
[2] SWIFT: gpi tracking, speed and deduction transparency
[3] Wise: correspondent-bank fees and return deductions
[4] Wise: paying by Swift, OUR and SHA
[5] HSBC Egypt: Schedule of Services and Tariffs
[6] Wise: receiving money by Swift fees
[7] Wise: costs to send USD outside the US
[8] Wise: sending/FX receiving fees and economics
[9] SWIFT: UETR and payment tracking
[10] SWIFT: ISO 20022 replaced cross-border MT coexistence
[11] CFPB: US consumer remittance disclosures and rights
[12] CFPB: Regulation E §1005.34, cancellation of remittances
[13] CFPB: deceptive marketing of remittance speed and costs
[14] CFPB: sending money abroad and understanding charges
[15] Wise: euro transfers beyond the SEPA region
[16] SWIFT: ISO 20022 transition announcement
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape. You can reach out to him via his social media accounts:
Linkedin: https://www.linkedin.com/in/johan-shamshad-742851262/
X: https://x.com/Yasmine_FX
Investing: https://www.investing.com/members/contributors/279781574