SWIFT: Why the Plumbing of International Payments Matters More Than It Looks

There’s a moment in most finance education when SWIFT stops being a vague acronym you’ve heard on the news and becomes something you actually have to understand. For most people, that moment arrives via geopolitics, a headline about a country being “cut off from SWIFT” rather than via a textbook. But the geopolitical framing, while accurate, tends to obscure what SWIFT actually is: a messaging network, not a payment system, and the distinction between those two things is where most of the confusion about international transactions begins.

SWIFT does not move money. It moves instructions about money. Understanding that one sentence properly is most of the battle, and it’s worth building the rest of the explanation on top of it rather than around it.

What SWIFT Actually Is

SWIFT the Society for Worldwide Interbank Financial Telecommunication is a cooperative founded in Brussels in the 1970s by a consortium of banks that needed a standardized, secure way to send payment instructions to each other across borders. Before SWIFT, cross-border instructions travelled by telex slow, inconsistent in format, and prone to error because there was no common template for “pay this amount, from this account, to this account, for this reason.”

SWIFT solved a coordination problem, not a payments problem. It gave every bank in the network a unique identifier the Bank Identifier Code, or BIC, sometimes called a SWIFT code and a standardised set of message formats so that an instruction sent from a bank in Mumbai would be unambiguously interpretable by a bank in Frankfurt. The genius of the system was never the money-moving; it was the standardisation that made money-moving between unrelated institutions in different legal and linguistic environments reliable enough to trust.

This matters for understanding everything that follows. When a bank is “removed from SWIFT,” it doesn’t lose the ability to move money in some mechanical sense money can still move through physical channels, correspondent relationships outside the network, or alternative messaging systems. What it loses is the ability to communicate payment instructions through the one channel that essentially every other bank in the world is already plugged into. That’s a crippling loss in practice, but it’s a communications loss, not a plumbing loss in the most literal sense.

How a SWIFT Transaction Actually Moves

Take a concrete example: an Indian importer needs to pay a supplier in Germany. The importer’s bank in Mumbai does not have an account relationship with the supplier’s bank in Frankfurt most bank pairs in the world don’t. So the payment has to travel through a chain of correspondent banks, each of which does have a relationship with the next link in the chain, and SWIFT is the messaging layer that carries the instructions at each link.

The Mumbai bank sends a SWIFT MT103 message, the standard message type for a single customer credit transfer to its correspondent bank, typically a larger bank with a presence in multiple currencies and jurisdictions. That correspondent, in turn, either has a direct relationship with the Frankfurt bank or routes through a second correspondent that does. At each hop, the message specifies the amount, currency, ordering customer, beneficiary, and any charges, and each bank in the chain debits and credits the appropriate nostro and vostro accounts the correspondent accounts banks hold with each other specifically to settle these obligations.

The important part is that the actual movement of value happens through these correspondent account relationships, settled typically through the currency’s home clearing system a euro payment ultimately clears through a payment system in the eurozone, a dollar payment through Fedwire or CHIPS in the United States. SWIFT’s MT103 message is the instruction that tells each bank in the chain what to do; the settlement itself happens outside SWIFT, in the underlying correspondent banking and domestic clearing infrastructure.

This is why cross-border payments, even routine ones, can take one to several business days and can pick up unexpected fees along the way every additional correspondent in the chain is another bank taking its cut and another point where the message has to be manually reconciled if something doesn’t match.

Message Types Worth Knowing

SWIFT messages follow a structured numbering system, and while the full catalogue runs into the hundreds, a handful are worth recognising because they come up repeatedly in trade finance and cross-border payment contexts.

MT103 is the single customer credit transfer — the workhorse message for one-off cross-border payments, the kind used in the importer-supplier example above.

MT202 is a financial institution transfer, used when banks are moving money for their own account or on behalf of another financial institution rather than an individual customer; it’s the interbank equivalent of an MT103 and is common in the correspondent banking chain itself.

MT700 and its series are the letter of credit family, used extensively in trade finance to establish, amend, and confirm documentary credits critical for exporters and importers who want payment guaranteed against the presentation of shipping documents rather than trusting a counterparty’s creditworthiness directly.

MT760 relates to guarantees and standby letters of credit, another trade finance staple.

SWIFT has also been migrating its messaging standard from the older MT format to a newer, more structured format called ISO 20022 (MX messages), which carries richer, more granular data fields useful for anti-money-laundering screening and straight-through processing, since it reduces the ambiguity that comes from cramming structured information into older, more free-text-heavy fields.

The Role of Correspondent Banking

SWIFT messaging only works because of the correspondent banking relationships underneath it, and it’s worth being explicit about why those relationships exist rather than treating them as a black box.

No bank can hold direct account relationships with every other bank in the world the compliance burden and capital cost would be enormous. Instead, banks concentrate their foreign currency relationships in a smaller number of large correspondent banks, typically institutions with a genuine global network and deep balance sheets in major currencies. A mid-sized Indian bank might hold a dollar correspondent account with a large US bank, a euro account with a eurozone bank, and so on, and route the vast majority of its foreign currency payment traffic through those relationships rather than trying to maintain direct links everywhere.

This concentration is efficient, but it also creates a structural vulnerability: a small number of large correspondent banks sit at the centre of an enormous share of global cross-border payment flows, and disruption to any one of them whether through sanctions, de-risking decisions, or operational failure has outsized effects on the banks and countries that depend on it. This is part of why “de-risking,” where large correspondent banks pull back from relationships with smaller banks in jurisdictions they consider higher-risk from a compliance standpoint, has been a recurring concern for banks in emerging markets, including in South Asia losing a correspondent relationship can functionally cut a bank off from significant parts of the global payment system even though nothing has happened to its SWIFT membership itself.

SWIFT and Sanctions: Why Disconnection Is So Consequential

The geopolitical salience of SWIFT comes almost entirely from its near-universality. Essentially every bank of any significance globally is a SWIFT member, which means that removing an institution or a country’s banking sector from the network doesn’t just inconvenience it; it removes the one channel that connects it to the rest of the global financial system in a standardised, mutually trusted way.

When banks are excluded from SWIFT as part of a sanctions regime, they don’t lose the physical ability to move money, but they lose the practical ability to instruct that movement through the infrastructure the rest of the world relies on. Workarounds exist manual correspondent instructions via other channels, alternative messaging systems, barter and commodity-for-commodity arrangements, or third-country intermediaries but all of these are slower, more expensive, and carry higher operational and compliance risk than SWIFT-mediated transfers. This is precisely why SWIFT exclusion has become a favoured tool of financial statecraft: it degrades a target economy’s connectivity to global trade and capital flows without requiring military or even direct economic intervention.

It’s worth noting that alternatives to SWIFT do exist and have been actively developed, partly in response to the geopolitical weaponisation of the network. Russia’s SPFS and China’s CIPS are the most prominent examples of national or regional payment messaging systems built partly as insurance against SWIFT dependency. None of these alternatives currently approaches SWIFT’s global reach or interoperability, which is precisely the network-effect advantage that makes SWIFT so valuable and so hard to displace a messaging standard is only useful to the extent that the counterparties you need to reach are also using it.

Risks and Limitations Within the SWIFT System Itself

Even setting aside sanctions and geopolitics, the SWIFT-mediated correspondent banking model carries structural inefficiencies worth understanding.

Chain length is the most obvious one. Every additional correspondent in a payment chain adds cost, delay, and a point of potential failure a mismatched reference number or an intermediary bank’s compliance hold can leave a payment in limbo for days while the originating bank tries to trace where it went, a process banks refer to informally as payment “tracing” and one that historically has been genuinely difficult given the lack of end-to-end visibility in the older MT-based system.

Fee opacity is a related problem. Because multiple correspondent banks may each deduct a handling fee before passing a payment along, the beneficiary can receive less than the sender intended without either party having full visibility into where the deductions occurred the “OUR/SHA/BEN” charge conventions in the MT103 message (specifying who bears the correspondent fees) exist precisely to manage this ambiguity, though disputes over deducted amounts remain common in practice.

SWIFT itself has also been the target of security incidents most notably the 2016 Bangladesh Bank heist, in which attackers used compromised SWIFT credentials to send fraudulent payment instructions and successfully moved tens of millions of dollars before the fraud was detected, a reminder that SWIFT’s security model depends heavily on the security practices of its member banks rather than being centrally bulletproof by design.

Why This Matters Beyond the Mechanics

For anyone working in eye banking, corneal tissue logistics, or any field adjacent to international trade and finance, the SWIFT framework is a useful mental model even outside pure payments it’s a case study in how standardisation of communication, rather than the underlying transfer of value itself, becomes the actual chokepoint and source of both efficiency and fragility in a global system. The same logic that makes a bank’s SWIFT connectivity so consequential shows up in other standardised international networks, customs documentation formats, trade finance messaging, logistics tracking standards anywhere multiple independent parties need to trust a common protocol rather than negotiating bilateral arrangements from scratch.

From a markets and macro perspective, SWIFT data itself has become a genuine analytical resource SWIFT publishes aggregated data on currency usage in global payments, and shifts in the relative share of currencies used in SWIFT-routed transactions are watched as one (imperfect) proxy for shifts in global trade and reserve currency preferences, including ongoing questions about the pace of any move away from dollar dominance in trade settlement.

Final Thoughts

The instinct to think of SWIFT as “the system that moves international money” is understandable but slightly wrong, and the correction is worth internalising. SWIFT is the standardised language banks use to tell each other what to do with money; the actual movement happens through correspondent account relationships and domestic clearing systems that sit underneath the messaging layer. That distinction explains why SWIFT exclusion is so damaging it doesn’t destroy a country’s banks’ ability to hold or transfer value, but it does destroy their ability to communicate about it in the one language virtually everyone else in the global financial system already speaks. And it explains why alternatives to SWIFT, however well-engineered, struggle to gain traction: a communication standard is only as valuable as the number of parties already using it, which is exactly the kind of network effect that’s easy to describe and very hard to displace.