After months of testing, Swift, the interbank payments platform that connects 11,000 global banks, has formally launched the system it believes will be the future of its cross-border payments services.

And notably absent was any integration with blockchain or distributed ledger technology.

As part of the formal rollout, Swift (or the Society for Worldwide Interbank Financial Telecommunication) revealed that 12 of the world's largest banks have been conducting real-time transactions via its Global Payments Innovation (GPI) project for some months.

Although Swift plans to continue its experiments with blockchain for other applications, GPI program director Wim Raymaekers told CoinDesk that its cross-border payments solution demonstrates the efficiencies possible using more traditional alternatives.

He said:

As of early this morning, initial GPI users – a group that includes the Bank of China, BBVA, Citi, ING Bank and Standard Chartered – will have conducted "tens of thousands of transactions" between 60 country corridors, Raymaekers said.

Instead of rebuilding Swift's cross-border payments infrastructure from scratch, the GPI has been constructed as a set of business rules encoded on top of the existing infrastructure with an eye to increasing speed, transparency and the traceability of transactions.

On top of those rules, Swift has further created a transaction tracking mechanism supported by cloud computing, with a graphical user interface designed to show exactly where a payment will be, should it get held up.

Participating banks will be able to add a widget to their websites to activate the feature, Raymaekers said.

Other banks that are now using the system – which has been live since January – include ABN AMRO, DBS Bank, Industrial and Commercial Bank of China, Intesa Sanpaolo, Nordea Bank and UniCredit.

Although the cost of the service is not being made public, Raymaekers shared with CoinDesk the potential savings attributed to another of the GPI's possible future services, which is yet to be implemented.

In January, Swift revealed it had commenced work on a blockchain proof-of-concept (PoC) to test whether the 'nostro and vostro' accounts common to banking could be simplified using technology from the Linux-led Hyperledger project.

To measure whether the blockchain PoC will be successful, Swift is analyzing the security and privacy of transaction data, and whether new and deeper types of account liquidity data can be captured.

"That’s why we’re looking for banks with a strong liquidity practice to come to the proof of concept," Raymaekers said.

Swift’s push to modernize cross-border payments, however, is part of a bigger movement around the world to make it easier and cheaper to send money. While Swift's service is aimed at making it cheaper for banks, rivals are looking to blockchain to help cut the banks out altogether.

At stake is nothing less than a slice of global remittance market estimated by the World Bank to be worth $552bn for the past two years.

To capture a piece of that transaction volume, some well-funded bitcoin startups have set their sights square on cross-border payments, including Align Commerce and Abra, two startups that have raised $20.25m and $14m, respectively.

But perhaps no blockchain startup has more at stake than Ripple, which has raised $93m from Standard Chartered, Accenture Ventures and others to streamline cross-border payments.

But even as Raymaekers positions his technology as a means to modernize cross-border payments without blockchain, he said Swift is looking to see how the system might further explore the technology.

This October, as part of Swift's annual Sibos conference, blockchain developers will be given access directly to the GPI as part of a hackathon, Raymaekers revealed to CoinDesk.

"We're going to open those APIs for fintech and blockchain designers to come up with ... new ideas," he said.

By the time the hackathon kicks off, Raymaekers expects that more of the 100 banks that have so far signed up for GPI will go live with real transactions, some as early as May. Further, he indicated that if the nostro-vostro blockchain PoC is successful, it is scheduled to go live sometime in 2019.

Overall, while Raymaekers is optimistic about the possibility that blockchain might improve some products, he ultimately sees the need for the tech as limited.

He concluded:

Images via Swift