Adoption of DLT presents significant operational challenges for Swift member banks

Adoption of DLT presents significant operational challenges for Swift member banks

Banking co-operative Swift has published the final results from its DLT proof of concept for Nostro reconciliation. While the PoC proved a resounding success, just don't expect a working solution anytime soon.

In total, 34 global banks joined Swift's DLT sandbox to validate whether blockchain-based technology can help banks reconcile their international nostro accounts in real time.

The PoC, which is based on Hyperledger Fabric v1.0 technology and has been conducted in concert with six founding banks, is part of the Swift gpi initiative to re-arm the correspondent banking system for a new age of technological disruption.

During the tests, Nostro Account owners and their servicers could share a private confidential ledger recording transactions related to their Nostro accounts.

“The PoC went extremely well, proving the fantastic progress that has been made with DLT and the Hyperledger fabric in particular”, says Damien Vanderveken, head of R&D at Swift. “The DLT sandbox enabled us to control access, to define and enforce user privileges, to physically segregate confidential data and store it only with the relevant parties while supporting a strong identity framework by linking all participants to their BIC, and having all keys signed by a Swift certification authority."

Despite the positives, translating the proof-of-concept into a workable product would require significant re-engineering - for instance, all account servicers would first need to migrate from batch to real-time liquidity reporting and processing, and back office applications would need to be upgraded to feed the platform with real-time updates.

States Vanderveken: “Although the PoC demonstrated DLT could improve Nostro liquidity management and reconciliation processes, it also revealed that the pre-requisites will have to be met before banks can enjoy the full benefits of switching to a DLT process."

The tests also showed that further progress is needed on the DLT technology itself. For example, just 528 channels were required in the PoC to ensure Nostro accounts would only be stored on the nodes of their account servicers and owners. In order to move to full production, more than 100,000 channels would need to be established and maintained, covering all existing Nostro relationships, presenting significant operational challenges.

In the meantime, Swift says it will continue to migrate more banks to towards real-time liquidity reporting and processing through the gpi project, and consult with its member banks on a timeline for full-scale uptake of ISO 20022 as a potential means to reduce integration costs.

“It is a strategic priority for Swift to work with new technologies like DLT and incorporate them into key solutions like gpi”, says Stephen Gilderdale, chief platform officer, Swift. “We are already working on new PoCs and will continue our R&D efforts to ensure that Swift customers will be able to leverage their existing Swift infrastructure and connectivity to benefit from blockchain services, whether offered by Swift or by third parties, on a secure and trusted platform.”

Comments: (1)

Behzod Sabirov
Behzod Sabirov - Sanscrit LLP - Almaty 12 March, 2018, 03:52Be the first to give this comment the thumbs up 0 likes

To me Nostro Reconciliation in connection with DLT is just meaningless. What is the purpose of NR? To make sure that a bank's internal ledger is balanced with the mirrored external statement. The non-real-time and non-transparent nature of bank transfers required this, right until the DLT-powered RTGS is presented. If you have a DLT system, that means that your ledger is fully synchronized with you counterpart's ledger at any given time. Of course there must system checks ensuring the integrity and raising flags in case of discrepancies. But the NR in it's traditional meaning and form becomes obsolete, if of course SWIFT augments its 'gpi' product with DLT properly.