Summarise the article with your AI
Here you can read:
International bank transfers pass through correspondent banks that hold pre-funded nostro accounts in foreign currencies before it reaches the recipient's account, which is why it rarely settles instantly. On the slowest SWIFT corridors, transfers take over two days end to end, according to a 2022 study by the Bank for International Settlements. Even within the euro area, a third of retail cross-border payments still took more than one business day to fully settle in 2024.
What actually happens between "send" and "received"?
When a corporate treasurer authorises an international payment outside of the SEPA area, the money does not typically move directly from one bank to another. The sending bank issues an instruction through the SWIFT network to one or more correspondent banks, which are institutions that hold accounts in foreign jurisdictions and can complete the payment leg the originating bank cannot. Each correspondent in the chain then debits and credits accounts of the participating partner banks in sequence, forwarding the instruction until it reaches the ultimate destination beneficiary bank, acting like a chain of instructions.
However, adding more intermediaries to the chain usually compounds the cost and increases the processing time. Payments requiring three or more correspondent banks represent fewer than 1% of volume but disproportionately slow the system.
Furthermore, SWIFT is a messaging network rather than a payment system. It is a messaging service rail that tells participating banks what to do and does not fund the payment, meaning a SWIFT message can cross the globe in seconds. Yet, the money still sits on the balance sheets of each correspondent bank and with each hop needs to be transferred from one bank to the next which costs time.
Research by the Bank for International Settlements analysing roughly 20 million transactions across 141 countries found something counter-intuitive: 78% of payments clear intermediary banks in under five minutes, yet only 33% clear the beneficiary bank leg in the same timeframe. The last mile, where the receiving bank credits the end customer's account, is where most of the delay happens.
Nostro and vostro accounts: the mechanism behind the delay
A nostro (“ours” in Latin) account is a bank's own account held in a foreign currency at another bank abroad. From the second bank's perspective, the same account is a vostro (“yours” in Latin) account. These labels describe the same account from opposite sides of the relationship.
A correspondent bank can only complete a payment straight away if its nostro account already holds sufficient pre-funded balance in the relevant currency. If the balance is insufficient, perhaps because the incoming payment flow exceeds expectations or the bank has not topped up its nostro position, the payment queues. This is the mechanical root cause of international transfer delays, existing alongside the necessary layers of compliance checks and fraud screening. The more hops involved, the more likely there is some underfunding somewhere in the chain, leading to multiple currency conversions and ultimately more time required.
How a nostro account can create a bottleneck
To participate in a currency corridor, every correspondent bank must park working capital in nostro accounts across a multitude of currencies, just in case demand arrives. This idle money is expensive. Funds sitting in a USD nostro account in New York earn some interested that’s typically not equal to the opportunity cost if the bank invested this money somewhere else. That cost contrasts with FX spreads and intermediary fees the bank earns for participating in the correspondent network to enable international transfers; yet, these costs are often visible to the corporate payer only after the payment lands, if at all.
The cost burden of maintaining these balances is why the global correspondent banking network has been shrinking. Between 2011 and 2018, active correspondent banking relationships declined by approximately 20%, and continued to decline thereafter. Banks have retreated from lower-volume corridors where the capital cost of maintaining nostro positions outweighs fee revenue.
Why some currency corridors are faster than others
Payment speed can vary enormously by corridor, ranging from a few minutes on the fastest routes to two or more days on several of the slowest. A country's wealth and financial interconnectedness is typically the strongest predictor of speed. Wealthier and more interconnected destination economies tend to have more developed payment infrastructure. For example, East-to-West payments tend to settle faster than West-to-East flows, and payments between North America and major parts of Western Europe typically settle quickly.
Payments destined for Northern Africa, Southern Asia, and Central Asia can take hours or in some cases surpass two days, because beneficiary bank infrastructure in those markets is less automated and nostro pre-funding requirements are more complex.
The cost of slow transfers for B2B treasury teams
Six in ten payments are credited to end beneficiaries within 30 minutes on the SWIFT gpi network, and nearly 100% within 24 hours. Yet these statistics measure the bank-to-bank messaging leg, not the moment the customer's account is actually credited.
For treasury teams, the problem is compounded by a visibility deficit. A significant proportion of companies still manage FX exposure manually, and many lack in-house banking or payment centralisation, meaning they have no real-time visibility into whether a payment is stuck and where. This translates into concrete pain: cash forecast error when expected receipts don't land as modelled; delayed supplier payments in slower corridors that damage relationships; and manual overhead chasing payments across time zones. As the industry saying goes: SWIFT can move a message in seconds, but the money still waits on whichever bank in the chain has not pre-funded its nostro account.
At the macro level, costs remain stubbornly high. The global average cost of sending $200 stood at 6.36% in Q3 2025, , according to World Bank data, more than double the UN Sustainable Development Goal target of 3%. Cross-border payments can cost up to ten times more than equivalent domestic transactions.
Correspondent banking vs. modern payment rails
The landscape is changing, albeit unevenly. SWIFT gpi, live since January 2017, overlays tracking and service-level agreements onto the correspondent banking network without replacing it. ISO 20022, whose cross-border coexistence period ended in late 2025, introduces richer structured payment data that reduces errors and screening false positives.
Fast payment system interlinking, which connects domestic instant payment networks across borders, is the approach championed by central banks. Where such links exist, they can bypass the nostro pre-funding step entirely for covered currencies.
| Dimension | Traditional correspondent banking | Modern payment infrastructure |
|---|---|---|
| Settlement time | Hours to 2+ days depending on corridor | Minutes to same-day on covered corridors |
| Fee visibility | Often opaque until funds land | Fees typically shown before send |
| Intermediaries | 1 to 3+, each adding time and cost | Fewer or none for corridors with local rails |
| Tracking | Manual bank enquiry | End-to-end status via SWIFT gpi tracker |
| Pre-funding requirement | Bank must hold nostro balance in currency | Varies; some models hold local accounts directly |
A practical framework for reducing transfer delays
Whilst infrastructure transitions play out over years, treasury teams can take concrete steps today:
1. Map your correspondent chain by corridor
Treasury teams should ask their relationship bank whether their most-used payment corridors are direct or multi-hop. Payments through three or more intermediaries carry higher delay risk and greater fee leakage, even though the vast majority of SWIFT payments today route directly or through a single intermediary.
2. Request SWIFT gpi tracking on all outbound payments
Rather than accepting basic transit statuses, gpi-enabled banks provide real-time, end-to-end visibility of where a payment sits in the chain.
3. Evaluate local-currency accounts for recurring corridors
For routes where a business makes frequent high-value payments, holding a local-currency account, or working with a bank that does, can remove the correspondent chain and the nostro pre-funding bottleneck entirely. This is particularly relevant for companies with regular payments into Southern or Central Asia, or Africa, where delay is structural rather than exceptional. Multi-currency account infrastructure increasingly offers this without the overhead of opening traditional bank relationships in each jurisdiction.
4. Build corridor-specific settlement time into your cash forecast
The most common treasury forecasting error on international payments is assuming same-day settlement when the corridor's typical settlement time is 24 to 48 hours. Assigning a corridor delay factor per payment destination converts an unknown into a managed variable, strengthening the overall cash forecast.
Today, 90% reach the beneficiary bank within an hour, but the last-mile crediting step still accounts for around 80% of total end-to-end time.
International bank transfers pass through correspondent banks that hold pre-funded nostro accounts in foreign currencies before it reaches the recipient's account, which is why it rarely settles instantly. On the slowest SWIFT corridors, transfers take over two days end to end, according to a 2022 study by the Bank for International Settlements. Even within the euro area, a third of retail cross-border payments still took more than one business day to fully settle in 2024.
What actually happens between "send" and "received"?
When a corporate treasurer authorises an international payment outside of the SEPA area, the money does not typically move directly from one bank to another. The sending bank issues an instruction through the SWIFT network to one or more correspondent banks, which are institutions that hold accounts in foreign jurisdictions and can complete the payment leg the originating bank cannot. Each correspondent in the chain then debits and credits accounts of the participating partner banks in sequence, forwarding the instruction until it reaches the ultimate destination beneficiary bank, acting like a chain of instructions.
However, adding more intermediaries to the chain usually compounds the cost and increases the processing time. Payments requiring three or more correspondent banks represent fewer than 1% of volume but disproportionately slow the system.
Furthermore, SWIFT is a messaging network rather than a payment system. It is a messaging service rail that tells participating banks what to do and does not fund the payment, meaning a SWIFT message can cross the globe in seconds. Yet, the money still sits on the balance sheets of each correspondent bank and with each hop needs to be transferred from one bank to the next which costs time.
Research by the Bank for International Settlements analysing roughly 20 million transactions across 141 countries found something counter-intuitive: 78% of payments clear intermediary banks in under five minutes, yet only 33% clear the beneficiary bank leg in the same timeframe. The last mile, where the receiving bank credits the end customer's account, is where most of the delay happens.
Nostro and vostro accounts: the mechanism behind the delay
A nostro (“ours” in Latin) account is a bank's own account held in a foreign currency at another bank abroad. From the second bank's perspective, the same account is a vostro (“yours” in Latin) account. These labels describe the same account from opposite sides of the relationship.
A correspondent bank can only complete a payment straight away if its nostro account already holds sufficient pre-funded balance in the relevant currency. If the balance is insufficient, perhaps because the incoming payment flow exceeds expectations or the bank has not topped up its nostro position, the payment queues. This is the mechanical root cause of international transfer delays, existing alongside the necessary layers of compliance checks and fraud screening. The more hops involved, the more likely there is some underfunding somewhere in the chain, leading to multiple currency conversions and ultimately more time required.
How a nostro account can create a bottleneck
To participate in a currency corridor, every correspondent bank must park working capital in nostro accounts across a multitude of currencies, just in case demand arrives. This idle money is expensive. Funds sitting in a USD nostro account in New York earn some interested that’s typically not equal to the opportunity cost if the bank invested this money somewhere else. That cost contrasts with FX spreads and intermediary fees the bank earns for participating in the correspondent network to enable international transfers; yet, these costs are often visible to the corporate payer only after the payment lands, if at all.
The cost burden of maintaining these balances is why the global correspondent banking network has been shrinking. Between 2011 and 2018, active correspondent banking relationships declined by approximately 20%, and continued to decline thereafter. Banks have retreated from lower-volume corridors where the capital cost of maintaining nostro positions outweighs fee revenue.
Why some currency corridors are faster than others
Payment speed can vary enormously by corridor, ranging from a few minutes on the fastest routes to two or more days on several of the slowest. A country's wealth and financial interconnectedness is typically the strongest predictor of speed. Wealthier and more interconnected destination economies tend to have more developed payment infrastructure. For example, East-to-West payments tend to settle faster than West-to-East flows, and payments between North America and major parts of Western Europe typically settle quickly.
Payments destined for Northern Africa, Southern Asia, and Central Asia can take hours or in some cases surpass two days, because beneficiary bank infrastructure in those markets is less automated and nostro pre-funding requirements are more complex.
The cost of slow transfers for B2B treasury teams
Six in ten payments are credited to end beneficiaries within 30 minutes on the SWIFT gpi network, and nearly 100% within 24 hours. Yet these statistics measure the bank-to-bank messaging leg, not the moment the customer's account is actually credited.
For treasury teams, the problem is compounded by a visibility deficit. A significant proportion of companies still manage FX exposure manually, and many lack in-house banking or payment centralisation, meaning they have no real-time visibility into whether a payment is stuck and where. This translates into concrete pain: cash forecast error when expected receipts don't land as modelled; delayed supplier payments in slower corridors that damage relationships; and manual overhead chasing payments across time zones. As the industry saying goes: SWIFT can move a message in seconds, but the money still waits on whichever bank in the chain has not pre-funded its nostro account.
At the macro level, costs remain stubbornly high. The global average cost of sending $200 stood at 6.36% in Q3 2025, , according to World Bank data, more than double the UN Sustainable Development Goal target of 3%. Cross-border payments can cost up to ten times more than equivalent domestic transactions.
Correspondent banking vs. modern payment rails
The landscape is changing, albeit unevenly. SWIFT gpi, live since January 2017, overlays tracking and service-level agreements onto the correspondent banking network without replacing it. ISO 20022, whose cross-border coexistence period ended in late 2025, introduces richer structured payment data that reduces errors and screening false positives.
Fast payment system interlinking, which connects domestic instant payment networks across borders, is the approach championed by central banks. Where such links exist, they can bypass the nostro pre-funding step entirely for covered currencies.
| Dimension | Traditional correspondent banking | Modern payment infrastructure |
|---|---|---|
| Settlement time | Hours to 2+ days depending on corridor | Minutes to same-day on covered corridors |
| Fee visibility | Often opaque until funds land | Fees typically shown before send |
| Intermediaries | 1 to 3+, each adding time and cost | Fewer or none for corridors with local rails |
| Tracking | Manual bank enquiry | End-to-end status via SWIFT gpi tracker |
| Pre-funding requirement | Bank must hold nostro balance in currency | Varies; some models hold local accounts directly |
A practical framework for reducing transfer delays
Whilst infrastructure transitions play out over years, treasury teams can take concrete steps today:
1. Map your correspondent chain by corridor
Treasury teams should ask their relationship bank whether their most-used payment corridors are direct or multi-hop. Payments through three or more intermediaries carry higher delay risk and greater fee leakage, even though the vast majority of SWIFT payments today route directly or through a single intermediary.
2. Request SWIFT gpi tracking on all outbound payments
Rather than accepting basic transit statuses, gpi-enabled banks provide real-time, end-to-end visibility of where a payment sits in the chain.
3. Evaluate local-currency accounts for recurring corridors
For routes where a business makes frequent high-value payments, holding a local-currency account, or working with a bank that does, can remove the correspondent chain and the nostro pre-funding bottleneck entirely. This is particularly relevant for companies with regular payments into Southern or Central Asia, or Africa, where delay is structural rather than exceptional. Multi-currency account infrastructure increasingly offers this without the overhead of opening traditional bank relationships in each jurisdiction.
4. Build corridor-specific settlement time into your cash forecast
The most common treasury forecasting error on international payments is assuming same-day settlement when the corridor's typical settlement time is 24 to 48 hours. Assigning a corridor delay factor per payment destination converts an unknown into a managed variable, strengthening the overall cash forecast.
Today, 90% reach the beneficiary bank within an hour, but the last-mile crediting step still accounts for around 80% of total end-to-end time.
The friction is a measured, closing gap, but it remains a real cash flow planning variable for treasury teams until it closes.The friction is a measured, closing gap, but it remains a real cash flow planning variable for treasury teams until it closes.
