The week a payment spent crossing the world

The money left the account on a Tuesday. By Friday, neither the seller nor the two brokers who’d spent four months on the deal could say, with confidence, where it was.

Not lost. Not stolen. Simply — somewhere. Inside the machinery. Held at a node that hadn’t replied, waiting for a compliance officer in a time zone that had already gone dark for the evening, sitting in a queue that would not move until Monday morning when a bank in Frankfurt opened its doors.

The deal had closed. The documents were signed. The congratulatory emails had already gone out. But the money — a seven-figure wire originated in Singapore, destined for accounts in Monaco and Miami — was spending its week the way a letter once spent weeks on a clipper ship: moving, in theory, but untouchable, unstoppable, and invisible to everyone who needed it most.

This is a piece about that week. Not how to prevent it — though prevention is possible — but what it actually is. What’s happening inside those days when a payment is “in transit.” Where it sits, who touches it, what they take, and why the system was built this way in the first place.

The infrastructure beneath the wire

When a professional sends an international wire, what they’re actually sending is a message. Not money — a message. The SWIFT network itself doesn’t actually move money; it sends secure messages between banks about the transfer. The money, in the classical sense, doesn’t travel at all. What travels is an instruction, passed from institution to institution, each of which maintains balances with the next in what are called nostro and vostro accounts — essentially pre-positioned pools of currency held at correspondent banks around the world.

The operational framework of correspondent banking relies heavily on nostro and vostro accounts, which allow the seamless transfer of funds between institutions. These accounts keep track of debits and credits. Think of it as a relay race where each runner doesn’t actually carry the baton across the world, but rather signals the next runner to pull a baton from a bucket they’ve already placed at their station. The original baton never moves. What moves is the authorization.

This architecture was not designed to be slow. It was designed to be global at a time when “global” meant building bilateral trust relationships between thousands of institutions across 200-plus countries — each with its own legal system, its own currency regime, its own compliance requirements, its own business hours. Correspondent banking allows domestic banks to access foreign financial markets through the support of a third-party bank, typically located in another country. This system enables banks to make international transactions without needing bank branches abroad.

The elegance of the architecture is also its burden. SWIFT processes over 53 million messages daily across 11,500-plus financial institutions in 220-plus countries, yet each message can take 1–5 business days to complete its journey. The reason those two facts can coexist — enormous volume, glacial completion — is that speed at the message level and speed at the settlement level are entirely different things. A message can travel in seconds. Settlement, the actual confirmed movement of value from one account to another with legal finality, is another matter entirely.

The hop count problem

Every correspondent bank added to a payment’s route is another processing window, another compliance queue, another business-hours clock. When your bank and the recipient’s bank don’t have a direct banking relationship, the payment routes through one or more correspondent banks — third-party institutions that handle the settlement on their behalf. Each correspondent bank in the chain processes the payment independently, applies its own compliance checks, may apply its own exchange rate, and operates on its own business hours and cut-off schedule.

Every “hop” potentially adds 24 hours to the total timeline. That’s a clean, abstract figure. The reality is messier, because each of those hops doesn’t simply add a fixed increment — it introduces a cutoff event. Miss the cutoff at one node, and you don’t add 24 hours; you add however many hours remain until the next processing window at that institution, which may not open again until the following business day.

A transfer from the US to a small bank in Southeast Asia may pass through three to four correspondent banks, each adding hours or a day to the timeline. In practice, for the professional managing a cross-border deal closing, this means that a wire sent late on a Tuesday afternoon may touch four institutions before it reaches its destination — and only one of them needs to encounter something unusual for the entire chain to pause.

Consider the arithmetic. On average, payments involve 1.31 intermediaries, meaning most use either one or two, with only a few involving up to four. Payments involving intermediaries averaged 1 day 11 hours and 15 minutes, while transactions bypassing intermediaries were quicker at 15 hours and 9 minutes. That differential — roughly 20 hours — is what a single hop costs, on average, in a straightforward transaction under normal conditions. In complex corridors, or when anything at all goes sideways, the number climbs fast. Currency conversion raises average processing time to approximately 4.6 days — 111 hours — compared to same-currency transfers that often settle within a day.

In a cross-border deal where the buyer is in one currency zone and the seller is in another, currency conversion is essentially guaranteed. Which means 111 hours — nearly five business days — is not an outlier. It is, by the numbers, the mean.

The architecture of a delay

To understand why a payment disappears for a week, it helps to trace a specific anatomy. The scenario is neither exotic nor unusual: a buyer in Singapore, a seller in Monaco, a co-brokered yacht sale. The purchase price is $4.2 million. The buyer wires in Singapore dollars. The seller expects euros in Monaco. Two brokers — one based in Miami, one in Antibes — are each owed their commission.

The buyer initiates the wire on a Tuesday morning at his bank in Singapore. The bank acknowledges receipt and tells him it has been sent. What has actually happened is that the bank has queued an outbound SWIFT message. Whether that message has left the building yet depends on whether Tuesday morning in Singapore is inside or outside the bank’s international cutoff window. International cutoff times are typically earlier than domestic cutoffs — banks need extra time to create the SWIFT message, route it through correspondent banks, and account for time zone differences at the receiving end.

Assume the wire was initiated at 9:30 AM Singapore time, well inside the window. The SWIFT message departs. Its first destination is not Monaco. The Singapore bank does not have a direct correspondent relationship with the Monaco bank. It routes through a major US dollar clearing bank — a money-center institution in New York that handles a substantial proportion of global USD-denominated transactions. USD payments, accounting for 63% of transactions, typically route through J.P. Morgan Chase Bank in the US. If the Singapore dollar must be converted to USD before conversion to euros, the payment will touch at least three institutions before reaching the eurozone: the Singapore bank, the New York correspondent, and at minimum one European correspondent before the Monaco beneficiary bank.

Each of these nodes has its own processing window. Time zone differences create a relay effect where an afternoon payment in New York arrives after London banks have closed their processing windows. Singapore is 13 hours ahead of New York. A wire that leaves Singapore at 9:30 AM lands in New York at 8:30 PM the previous evening, local time — well after the international wire cutoff at most US institutions. It sits overnight. It is processed Wednesday morning in New York. It departs New York Wednesday afternoon. It arrives at the European correspondent bank Wednesday evening, Central European Time, after that bank’s cutoff. It sits overnight again. It is processed Thursday in Europe. It arrives at the Monaco beneficiary bank Thursday afternoon.

Four institutions. Two overnight holds. Two full business days of elapsed time — assuming nothing triggered a review flag at any node.

The compliance hold: where payments go dark

A wire doesn’t have to do anything wrong to trigger a compliance hold. It merely has to look, in the momentary judgment of an automated screening system, unusual. If the transaction triggers a review flag — an unusual amount, a new beneficiary, a high-risk corridor — it can be held for manual review. In a deal closing context, almost every wire looks unusual. The amounts are large. The counterparties are often new — a buyer and seller who have transacted with each other exactly once. The corridors are often non-standard: Singapore to Monaco is not a high-volume route.

The review process isn’t arbitrary. Banks operate under strict regulatory obligations and can face significant penalties for non-compliance. But for the sender, a compliance hold looks identical to a technical delay.

This is the brutality of the black box. From the outside, there is no difference between “your wire is processing normally and will arrive Thursday” and “your wire has been flagged for manual review at a correspondent bank and we have no estimated resolution time.” The status reads: in transit. The money is moving. Everything is fine.

Banking compliance delays are no longer exceptional — they are becoming a consistent feature of cross-border transactions, particularly where US dollar payments and intermediary banks are involved. What once may have been viewed as an isolated banking issue has become a structural risk that parties must proactively manage. Left unaddressed, these delays can jeopardize deal timelines, trigger contractual defaults, and in some cases cause transactions to fall apart entirely.

For the yacht broker in Antibes — who has already told the seller the funds are “on their way” — the compliance hold is invisible until Thursday, when the Monaco bank calls to say it hasn’t received anything. The broker calls the Singapore buyer’s bank. The bank says it was sent. The broker calls the Monaco bank again. Monaco says they see nothing. The broker emails. No one responds until Friday morning. Friday morning reveals a wire sitting at a New York correspondent bank, flagged for enhanced due diligence on a transaction originating from a jurisdiction that has seen elevated scrutiny in recent months.

The compliance officer responsible for reviewing it works Monday through Friday. It is now Friday afternoon in New York. The deal, which closed on a Tuesday, will not be funded until the following week.

The cut-off time as a structural mechanism

Cut-off times exist, ostensibly, for operational reasons — banks need time to process a day’s payments, balance their Federal Reserve accounts, manage their correspondent positions. But the cut-off window does something else, something worth naming clearly: it creates a guaranteed minimum delay for any payment that misses it.

Banks impose strict cutoff times for wire transfers — often 2:00–5:00 PM depending on the bank — even when Fedwire operates until 7:00 PM. The earlier cutoff gives banks time to batch-process wires and manage their Federal Reserve account balances. In international corridors, the practical cutoff is often earlier still. SWIFT cutoffs are often around 2:00 PM to allow for manual compliance reviews and correspondent bank messaging.

This means that for a professional managing a deal closure, the window during which a wire can be initiated and arrive on the same day is not “business hours.” It is a narrow band — often 9 AM to 2 PM at the sending bank — that shrinks further when time zones are involved. For international wires, initiating early in the morning — ideally before noon — gives the transfer maximum time to clear through multiple time zones before business hours close at the receiving bank. A wire sent at 3:00 PM for Asia will often sit until the next business morning at the correspondent bank.

The Friday problem compounds everything. The “weekend slip” occurs when Friday afternoon payments don’t move until Monday, creating an automatic 72-hour minimum delay. In a deal that has multiple payments — a split commission structure, for example, where the sale proceeds go to a seller’s account while commission wires go to separate broker accounts in separate jurisdictions — the probability that at least one payment slips the Friday window approaches certainty.

Compound holidays happen when sending and receiving countries observe different holidays in the same week, potentially adding 5–7 business days. Singapore, France, and the United States observe different national holidays on different calendars. A deal closing in a week that straddles a public holiday in any of the three countries can see routine two-day timing extend to seven or eight days with no error, no fraud, no malfunction — just calendar arithmetic multiplied across jurisdictions.

What the system charges for the privilege

The delay is the price of distance. But distance also charges a separate, more literal price: fees, extracted at each node, often invisibly, from the principal of the payment itself.

Each correspondent in the chain is entitled to deduct a fee from the transfer. A wire from the United States to a bank in West Africa might touch three or four intermediaries: a New York money-center bank, a European correspondent, a regional African correspondent, and then the beneficiary’s bank. Each hop can deduct $15 to $30 or more. The sender does not know in advance how many correspondents the payment will pass through or what each one will charge. This is the system’s dirty secret: the fee is variable, unpredictable, and deducted from the payment amount itself. The beneficiary receives less than what was sent, and neither party knows exactly how much will be taken until after the payment arrives.

In a consumer transfer, a $30 shortfall on a $1,000 payment is a nuisance. In a professional deal context, an unexpected deduction creates a reconciliation problem that can hold up final disbursement. If a broker’s commission wire arrives $47 short, someone has to account for the discrepancy. When a supplier says they got less than was sent, an intermediary bank usually took a cut. SWIFT payments often hop through one or more correspondent banks before reaching the destination, and each can deduct a fee from the principal as it passes through. You sent $10,000; your supplier sees $9,930; the missing $70 went to banks you never chose and can’t see on your statement. The practical fallout lands on your accounts payable team. Suppliers chase the shortfall, your team reconciles a payment that doesn’t match the invoice, and someone has to decide whether to top up the difference.

Beyond the correspondent fees, there is the exchange rate markup — the largest single cost in most international wires, and the one least likely to appear on any statement as a line item. For any wire that involves a currency conversion, the exchange rate markup is almost always the single largest cost component. And it is the one least likely to be disclosed clearly. When a bank converts currency for a wire transfer, it does not use the mid-market rate. It applies a markup typically ranging from 1.5% to 5%, depending on the currency pair, the bank, and the client relationship. On a $100,000 transfer, a 2% FX markup costs $2,000.

On a $4.2 million deal, a 2% markup is $84,000. It appears as a slightly unfavorable exchange rate — a number that feels like a market fact rather than a fee. It is not a market fact. It is a margin, captured silently, distributed among institutions the sender never selected.

And then there is the float.

The float economy

Float is the oldest revenue stream in banking. It is also, in the context of international wires, the most structural and the least visible.

International wires present particularly lucrative float opportunities because they involve currency conversion and correspondent banking delays. When you send $50,000 internationally, your bank might debit your account immediately but delay the currency conversion for 24–48 hours to benefit from foreign exchange positions.

The money leaves your account the moment you initiate the wire. It does not arrive in the recipient’s account for days. During the interval, those funds — or rather, the pre-positioned balances that represent them — sit within the banking system, earning. Float income represents 16–22% of total wire transfer revenue for international wires. On the enormous volumes moving through the correspondent banking system daily, this is not incidental — it is structural. This isn’t entrepreneurial financial engineering — it’s infrastructural value extraction built into the architecture of payment systems themselves.

For any individual deal, the float captured on a single wire is small in absolute terms. But for the professional managing a multi-party deal — a sale where the proceeds must flow from one account to three others — float compounds across every leg of every disbursement. The closing happened. The deal is done. But the system extracts a toll measured not just in fees, but in time — and time, in professional deal-making, has a cost that is rarely calculated.

Every extra day of float affects your bottom line. For a broker waiting on commission, that extra day is a receivable that hasn’t arrived. For a seller planning to redeploy capital, it’s a constraint on a decision. For a co-broker in a different currency zone, it’s exchange rate exposure on funds that are technically earned but not yet held.

The fragmentation problem: when a deal has multiple recipients

Most of the existing literature on international wire delays focuses on a single-beneficiary payment: one sender, one recipient. Professional deal closings are almost never that simple. The typical structure — a sale with a buyer, a seller, a listing broker, a co-broker, and potentially a referral advisor — involves multiple disbursements, often to multiple countries, across multiple currencies, initiated in sequence or sometimes in parallel.

Every additional beneficiary multiplies the exposure surface. More wires means more hops, more cutoff windows to hit, more compliance screenings to clear, more correspondent fee deductions, more exchange rate conversions. A deal that has four parties receiving payment across three countries is not one payment with a one-in-ten chance of delay — it is four payments, each with its own independent probability of delay, and a meaningful probability that at least one of them will catch in the machinery.

The coordination burden falls on whoever is closest to the deal — typically the broker or the closing professional. They field the calls when a payment hasn’t arrived. They email the banks. They verify wire instructions that may have changed since they were first exchanged. Even a minor typo in recipient information can cause the transfer to bounce back, requiring the sender to reinitiate it with corrected details. Reinitiation means starting the clock again: new initiation, new cutoff window, new compliance screening.

It is essential for closing documents to be specific as to when payment obligations are considered fulfilled. Is it when the buyer initiates the transfer? When the account holder confirms receipt? Or only once the funds are fully cleared? Ambiguity at this point is a common source of dispute, particularly in transactions with tight delivery windows.

These are not edge-case questions for unusual deals. They are routine ambiguities in the anatomy of every cross-border professional transaction, surfacing only when something in the chain is slower than expected — which, as the numbers make clear, is not a rare event. It is the median outcome.

The geography of speed: why some corridors move and some don’t

Not all international wires are equally slow. The geography of the correspondent banking network creates stark disparities in settlement speed based on which country pair you’re crossing.

Transfers between regions with direct banking connections, such as North America and Europe, are typically settled more quickly. Transfers between Europe and Africa, on the other hand, might take longer due to additional compliance checks and potential delays with intermediary banks. “Direct banking connections” is the operative phrase: it means the sending bank has a pre-established correspondent relationship with the receiving bank, or with a correspondent that does. When that relationship exists, the hop count drops and the processing time compresses.

SWIFT payments to India take about 3 days on average, compared to the overall average of 18 hours. This is because India has strict controls on foreign funds entering the country. Often, the Reserve Bank of India needs to approve these transfers, adding extra time to the process.

This geography has changed over time, not always in favor of speed. A BIS study found a continued decline in the number of correspondent banking relationships; between 2012 and 2019, active relationships in the global network declined by about 20%, though reductions varied across regions. Fewer direct relationships means more indirect routing — more hops — which means slower, less predictable settlement for the corridors that have lost coverage.

The irony is that the corridors most affected by this de-risking trend — markets in emerging economies, jurisdictions with higher compliance overhead — are often the same corridors most important to the kind of cross-border deal professionals who specialize in serving global buyers. The yacht broker with a client base in the Gulf. The commercial real estate advisor connecting Asian capital to European assets. The dealmaker whose whole practice depends on moving money cleanly across jurisdictions that the major correspondent banks have been quietly stepping back from.

The total volume of payment messaging has not fallen, indicating that banks in smaller countries might be seeking out intermediary banks to conduct correspondent banking through additional hops — adding further layers to a chain that was already long enough.

What happens when a wire fails

A delayed wire is a problem. A returned wire is a different category of event entirely.

Even a minor typo in recipient information can cause the transfer to bounce back, requiring the sender to reinitiate it with corrected details. If the account number, SWIFT code, or IBAN is incorrect, the transfer will likely be rejected or rerouted and require manual intervention to complete.

The return journey is not free. A returned status means the wire was rejected and funds will be credited back to your account, less any fees charged by intermediary banks. The correspondent banks that handled the payment on the way out took their fees. They do not give them back because the payment didn’t complete. The sender gets back the principal, minus whatever was charged for the failed attempt, and must start over — often with the same recipient details, which may not be wrong at all, but which triggered a pattern-matching algorithm somewhere in the chain that decided something looked irregular.

There is a class of deal professional — the ones who close complex international transactions routinely — who maintain a private mental lexicon for the different failure modes of international wires. The “bounce-back,” where the wire returns in two to five days without explanation. The “silent hold,” where the wire never officially fails but also never arrives. The “short arrival,” where the amount credited is $100 or $400 less than sent and no one in the chain will explain who took it. The “Monday problem,” where a Friday afternoon wire to Asia sits over the weekend and is then processed into a Monday morning that happens to be a local holiday, extending the delay to Wednesday.

Each of these failure modes carries a common thread: the professional who structured the deal, who shepherded it to closing, who maintained the relationships and coordinated the parties, is now a passenger in someone else’s infrastructure. Their deal closed. Their part is done. But the payment — the final act that gives the deal its meaning — is somewhere in a system they cannot see, cannot access, and cannot accelerate.

The moment the architecture changes

The problem, stated precisely, is this: a deal between known parties, structured by professionals, with agreed-upon terms and consented allocation of proceeds, requires a week of institutional processing, fee extraction, and opacity to execute its most fundamental step — moving value from one account to several others.

None of the delay serves the parties. It serves the correspondent banks that hold the float. It serves the compliance systems that operate on business-day schedules. It serves the architectural inertia of a network built when bilateral relationships and batch processing were the only options available.

The professionals who structure deals are not the problem. They execute with precision. They negotiate the terms, verify the counterparties, manage the documentation, and bring the closing together. What they hand off at the end of that process is an instruction — pay these people, in these amounts, to these accounts — that then spends a week inside infrastructure that was not designed with their clients’ interests at the center.

There are brokers and advisors who have started routing their closing payments differently. Not because the traditional rails have stopped working, but because they’ve learned to distinguish between the closing event — the moment the deal is done — and the settlement event, which can and should happen simultaneously, on terms the professional controls.

When a deal closes through Shaka, the broker sets up the payment link before the transaction completes: the recipient wallets, the split percentages, the commission structure. When the deal executes, every party receives their funds in the same transaction — directly, onchain, without hops, without cutoff windows, without a correspondent bank in a jurisdiction no one selected making a decision that holds the payment over a weekend.

The deal and the settlement are the same moment. There is no week-long interval in which the proceeds of a negotiated transaction are the guest of someone else’s banking infrastructure.

What the delay reveals

The week a payment spends crossing the world is not a failure of the system. It is the system, operating as designed — a design that predates real-time settlement, that was built around bilateral trust relationships between thousands of institutions, that generates substantial revenue for the institutions in the chain, and that has never been optimized for the professionals who stand at either end of it, waiting.

Total processing time equals the sum of sender processing, intermediary bank hops, compliance checks, and recipient verification. Every element in that sum is a genuine function being performed by a genuine institution with genuine costs. None of it is malicious. All of it takes time.

But when you add up the hops, the cutoffs, the overnight holds, the compliance queues, the weekend slips, the FX spreads, the correspondent fees, the float captured at each node, the probability of a short arrival, the risk of a return — what you see is not a payment traveling the world. You see a payment resting in the world, accruing costs and delays for a system that was never in the same room as the deal that generated it.

The professionals who close deals have always understood that their value is in the structure, the relationships, the judgment — not in the plumbing. What changes is that the plumbing now has an alternative. And when the alternative means the closing and the settlement are the same moment, the question isn’t why you’d use it. The question is what exactly you were paying for, all those years, during the week a payment spent crossing the world.