.jpg)
SWIFT has become a convenient villain in modern finance. When a transaction takes three days to settle, frustration builds. The blame is almost instinctively placed on the messaging network itself. Vendors promise to replace it with blockchain solutions, and crypto pitch decks often speak evangelically of a world without it. However, focusing on SWIFT hides a much simpler and less glamorous reality regarding international finance. SWIFT is not the problem. The correspondent banking chain is.
To understand the delay, we need to separate the instruction from the asset. SWIFT is fundamentally a secure messaging network that allows financial institutions to send payment details. However, it does not move money, it does not hold deposits, and it does not determine when funds are released. A SWIFT message is simply data. The actual capital resides in nostro and vostro accounts, moving via domestic RTGS systems and local clearing schemes.
While the SWIFT message travels across the globe in seconds, the delay lives in how each institution along the path manages its own accounts and risk parameters. When a transaction stalls, the bottleneck is almost always found in that chain of custody rather than in the wire that carried the instruction. Blaming SWIFT for a delayed settlement is like blaming an email provider because the recipient waited two days to open the message and reply.
The hidden cost of every hop
Most international flows continue to rely on the correspondent banking model. In this system, if a bank in one country does not have a direct relationship with a bank in another, the transaction must hop through intermediaries. This is not a seamless relay; each step in the sequence adds resistance.
Every hop adds:
- another balance sheet to fund
- another set of compliance checks
- another set of cut-offs, holidays, and local quirks
- another chance to add margin, fees, or spreads
This operational drag creates measurable latency. SWIFT's own data shows that while roughly 90% of transactions reach the destination bank within an hour, fewer than half are credited to the end customer's account in that same timeframe. The delay is not in the transit; it is in the processing queue of the receiving institution.
Why banks are paid to wait
If technology is not the primary issue, we must look to economic incentives. The correspondent banks facilitating these flows are typically large universal banks. Their core economic engine is lending and balance sheet management rather than clearing. These institutions earn yield by holding deposits and managing liquidity, not by pushing funds out the door as fast as possible.
That creates predictable tensions:
- settlement can be batched or delayed to smooth intraday liquidity
- funds can sit in internal accounts until windows or limits align
- risk teams can slow flows when risk appetite tightens
When operational reality meets these misaligned incentives, the result is a compounding delay that no messaging standard can fix. Legacy cores, manual exceptions, and misaligned time zones all stack delay on top of these economic priorities. Operational teams bounce investigations between institutions through tickets and emails, magnifying the friction. None of this is a feature of SWIFT. It is the consequence of who performs clearing and what their balance sheets optimise for.
The case for unbundling clearing
The traditional network is fracturing under this pressure. According to the Bank for International Settlements (BIS), the number of active correspondent banking relationships has declined by over 20% in the last decade. As universal banks retreat from this low-margin utility, a vacuum has opened for specialist infrastructure.
This helps explain why many fintech projects fail to solve the core issue. Attempts to create new rails often miss the point. If the system still relies on universal banks to hold funds, the underlying friction remains. The critical design question is not how to remove SWIFT, but who should perform institutional clearing and under what incentives.
A truly effective solution requires a structural shift where clearing is separated from lending. Universal banks earn yield by holding deposits. That incentive is incompatible with clearing velocity. The alternative is infrastructure built exclusively for clearing: 100% reserve models with no lending book, direct access to local rails, and economics aligned with moving funds rather than holding them.
This is the model Lorum operates. A neutral clearing layer with direct rail access across 30+ markets, allowing institutions to clear and manage liquidity through one relationship. The result is less trapped capital, lower counterparty risk, and a network that compounds as volume grows.
For treasurers, this shift means:
- local settlement on domestic rails wherever possible
- a single global view of balances and flows
- fewer intermediaries and more predictable timelines
This approach does not replace the messaging layer, as SWIFT already moves messages well. Instead, it redesigns the institutional layer behind those messages. It focuses on the start and end of each transaction where custody, timing, and control actually break. It is easy to blame SWIFT because it is visible and old. It is harder, and more useful, to redesign who holds and releases funds.
