The Bank of Canada's experimental effort to build a wholesale payment system using a DLT-based settlement asset has thrown up concerns about whether the technology can meet international standards for systemically important payments infrastructure.
Last summer, the central bank revealed it is working with the country's biggest banks, Payments Canada and the R3 consortium on a DLT experiment called Jasper that uses a digital fiat currency dubbed Cad-Coin.
The system sees participants pledge cash collateral into a special pooled account held by the BoC, which then converts it to generate the Cad-Coin, which is transferred to fund participants' accounts. Then identifiable counterparties can exchange assets on the Cad-Coin platform and redeem the digital currency for cash collateral, with the BoC destroying the redeemed Cad-Coin.
As the experiment moves towards its conclusion in the spring, BoC senior deputy governor Carolyn Wilkins has written an article for CoinDesk in which she says that it has "taught us a lot about how the technology would have to improve to win a horse race with our current Large Value Transfer System".
With the bank's oversight role central, Wilkins says that a major objective of the work has been to see how the test system could meet the Principles for Financial Market Infrastructure (PFMIs).
While Jasper has proved itself up to the task of meet the PFMIs concerning collateral, credit risk, money settlement and liquidity risk, "there are some important gaps, through, with respect to settlement finality, operational risk, as well as access and participation requirements".
A new platform build is underway to "see if these gaps can be addressed".
Meanwhile, Wilkins says that the initial design is "quite collateral intensive" and that means that any cost savings are likely to come not through the system itself but through bank reconciliation efforts.
The big savings seem likely to come through building other applications - clearing and settlement, trade finance - on top of the core cash payment distributed ledger system.
Wilkins also warns that "while DLT may aim to reduce concentration of risk, a substantial amount of centralisation would still be required (eg permissioning of nodes and setting of operational standards) if applied to wholesale payments systems".