The choice of infrastructure model shapes everything from settlement speed to capital efficiency to regulatory exposure. This comparison examines where the two architectures diverge and what each means in practice for platforms operating across multiple markets.
How correspondent banking works for platforms today
When a regulated platform needs to move money across borders, it typically relies on a chain of correspondent banking relationships. The platform's bank sends a payment instruction to a correspondent bank in the destination country. That correspondent holds funds in a nostro account, converts currency if needed, and credits the beneficiary through the local clearing system.
If the platform operates in fifteen or twenty markets, it may depend on a dozen or more correspondents, each with its own onboarding requirements, settlement schedules, and fee structures. This model was designed for bank-to-bank flows, not for platforms processing thousands of transactions daily across multiple corridors.
The structural tension is straightforward: correspondent banks earn revenue from their balance sheets. They lend deposits, capture FX spreads, and retain funds for as long as it is economically rational. Clearing, the act of releasing funds quickly and predictably, competes with those economics. Settlement queues form not because the technology is slow, but because releasing funds competes with retaining them.
For the platform sitting at the end of this chain, the consequences are tangible. An estimated $5 trillion sits trapped globally in nostro and vostro prefunding, according to Oliver Wyman and SWIFT joint research. Customer funds often sit in pooled omnibus accounts where individual end-customer visibility is limited.
Banking access itself is becoming fragile. BIS CPMI monitoring data shows a 20% decline in active correspondent banking relationships since 2011, with steeper declines affecting PSPs in particular.
What a specialist clearing model looks like
A specialist clearing institution is designed exclusively for clearing, settlement, and custody. It does not lend. It does not operate a trading desk. It does not manage a retail deposit base. The entire business model is built around moving funds predictably and maintaining segregated client accounts, not around balance sheet yield.
In practical terms, this means a platform can replace a network of bilateral correspondent relationships with a single clearing partner that provides direct connectivity to local payment rails in each market. Account opening, payment initiation, FX conversion, and balance reporting all run through one API.
The platform does not need to manage separate onboarding processes, negotiate individual fee schedules, or reconcile across multiple banking portals. One integration surface replaces a patchwork of bilateral connections.
The custody model is structurally different as well. Instead of pooling client funds into a single omnibus account under the platform's name, a specialist clearing institution can provision named accounts in the customer's name, each with its own KYC profile and direct contractual relationship with the custodian.
This eliminates the counterparty risk that pooled structures introduce and satisfies safeguarding requirements in most regulated jurisdictions without the platform needing to apply for its own banking licence.
A structural comparison
The differences between the two models are not incremental. They are architectural. The table below summarises the key dimensions where the models diverge.
The direction of travel is clear. The FSB's 2025 progress report on cross-border payments highlights the need for greater transparency and end-to-end visibility across payment chains, reinforcing the structural advantages of models where every end customer is individually identified at the point of clearing. For platforms operating at scale, the cumulative difference across these dimensions, from settlement certainty to capital efficiency to compliance readiness, compounds with every market added.
When correspondent banking still applies
Not every institution should move away from correspondent banking. The model remains well suited to large global banks that already hold correspondent relationships as part of their core operations. Institutions that process low volumes of high-value interbank transfers may find the existing model adequate.
The structural case for specialist clearing is strongest when a platform:
- Operates across multiple markets with high payment volumes requiring predictable settlement
- Needs named or segregated account structures for regulatory or commercial reasons
- Finds existing banking relationships increasingly difficult to maintain as correspondents de-risk their portfolios
- Wants to reduce operational overhead from managing ten or more bilateral banking integrations
For platforms in that position, the question is not whether correspondent banking works in theory. It is whether the model's incentive structure serves their operational requirements in practice.
How Lorum's clearing infrastructure is designed
Lorum operates as a specialist correspondent institution focused exclusively on clearing, custody, and cash management. The infrastructure provides direct access to local rails across 30+ markets through a single API integration.
Platforms connect once and clear in USD, EUR, GBP, AED, and additional currencies without establishing local entities or managing bilateral banking relationships in each jurisdiction. The Named Account Custody model provisions segregated accounts in each end customer's name, with full KYC at the individual level.
There is no conflicting lending book, no commingling of client funds, and no balance sheet incentive to delay settlement. Cash management services include wholesale FX at institutional rates, automated liquidity sweeps across currencies, and real-time position visibility across all markets.
For platforms evaluating whether to continue building on correspondent banking or consolidate through specialist infrastructure, the comparison is not between two equivalent options with different pricing. It is between two fundamentally different architectures, each optimised for a different objective. One is designed to maximise balance sheet yield. The other is designed to clear.
