The latest phase of the Utility Settlement Coin project, which aims to settle trades with a token linked to accounts at central banks, has seen the consortium expand.

Barclays, CIBC, Credit Suisse, HSBC, MUFG and State Street are the latest participants; the project began life with UBS and the other early stage members are BNY Mellon, Deutsche Bank, Santander, NEX. The technology is being built by UK-based start-up Clearmatics.

This third phase of the project is expected to run for about 12 months. Following that we can expect to see the first live collateralised token exchange using the platform, which could occur as soon as the end of 2018, according to Hyder Jaffrey, UBS director of strategic investment and fintech innovation.

Not much has been said about the technology underlying the USC project. Robert Sams, Clearmatics CEO, who favours blockchain technology inspired by Ethereum for the Cleamatics decentralised clearing system, confirmed USC is built on something similar. "It's an Ethereum-like architecture rather than UTXO/Bitcoin-like," he said in an email. Asked if it would comprise a single monolithic chain or a series of cross-connected chains, he explained: "Multiple chains. We will be making some announcements about our approach to interoperability in due course. It will be based on open standards and open source implementations."

Privacy is a big problem for industrial-grade blockchains that handle sensitive information. Clearmatics has leveraged ring signatures to develop a privacy solution called Mobius. Ring signatures are a type of zero knowledge proof for set membership, often associated with things like anonymous voting and whistleblowing.

At a recent Enterprise Ethereum Alliance meet-up, Clearmatics blockchain engineer Matthew Di Ferrante explained how Mobius could be used for payments. He pointed out that it has less in the way of untested cryptographic assumptions than zk-SNARKs and less computational overhead. It can handle tens of thousands of transactions a second, he said.

It is interesting that USC is being developed at the same time as CLS is developing CLSNet, a new automated payments netting service for FX trades based on the Hyperledger Fabric distributed ledger platform. It is clear that regulators and central banks are entertaining the idea of new peer-to-peer infrastructure, including as new services operated by market infrastructure incumbents.

One of the new banks to join USC is Barclays. Dr Lee Braine of Barclays Investment Bank's CTO Office reiterated that Barclays had just joined the USC project and so would begin its architectural reviews of aspects such as data privacy, scalability, interoperability with existing infrastructure, etc.

He said that as well as potential benefits with capital efficiency and risk reduction based around shortening the settlement life cycle, there may also be broader gains from an industry architecture perspective. "The current capital markets ecosystem comprises a highly complex technological landscape with many different systems interacting via many different data feeds. A proportion of the industry processes in place are effectively workarounds to support the fact that there often isn't instant settlement finality. The innovative thinking is that, if you could have a utility token permitting settlement finality, then you may also be able to simplify that ecosystem in order to reduce costs and reduce risk."

The use case roadmap for USC is likely to start with payments and then add other transactions in the longer term.

"The potential value obtained from such utility tokens could eventually be broadened and extended by going beyond payments to also include more complex transaction types where you are looking for an atomic swap of different asset classes," added Braine.

"A good example would be delivery versus payment (DvP), where one leg of the transaction could potentially be the payment as a USC token denominated in a particular currency, and the other leg could be the security. Additional benefits are likely to accrue when both legs can settle instantly with finality, as opposed to just one leg - the payment - and the other leg taking longer."

Harder challenges could include efficient multi-party netting on a blockchain and perhaps even the use of utility token by clearing houses. However, Braine pointed out that efficient securities netting would remain a challenge for distributed ledger technology (DLT) in general for some time to come. He explained: "Remember that no single netting solution is perfect for all scenarios, with one example reason being the unavoidable trade-off between settlement frequency and settlement efficiency."

Braine said that while some have proposed that efficient securities netting could be performed in a completely decentralised manner, he remains sceptical. "I think for certain netting optimisations, you need to effectively temporarily centralise an aspect of the algorithm in order to perform the equivalent of a batch exercise. That's not necessarily incompatible with a distributed ledger, it just means you may need to pass control to a trusted entity for a short while – whether that's a broadly-trusted node on the distributed ledger or perhaps a trusted oracle."

Regarding blockchains and digital currency, the central banking world has been quite vocal. There has been research by the Bank of England (including some work in association with UCL), Bank of Canada, MAS for Singapore, while the People's Bank of China has opened a Digital Currency Research Institute.

But despite this enthusiasm, issuing a central bank digital currency would probably take years to happen, and so something like the USC, which is focused on major currencies, has the potential to be deployed in the meantime.

But of course to move this project forward may mean loosening a Gordian Knot: a whole layering of agreements, whether they be network agreements, counterparty agreements, and jurisdictions under which these agreements are governed and supervised by regulators.

Peter Randall, CEO of rival blockchain settlement solution SETL, queried: "Would a deposit in the name of a USC account at a central bank be sufficient for global regulators to agree that any bank using this route would be allowed to relax the capital treatment of their obligations towards each other by an equal and opposite amount?"