Project Jasper is an experiment being done by the Bank of Canada, Payments Canada and R3 to test the viability and feasibility of using Distributed Ledger Technology (“DLT”) as the basis for wholesale interbank payment settlements. This project was launched in March 2016 and has completed two phases. Phase 1 of Project Jasper employed the Ethereum platform as the basis for the DLT, while Phase 2 employed the custom-designed R3 Corda platform. In June 2017, the Bank of Canada issued a report on its preliminary findings from Project Jasper, which were summarized in our previous article. On September 29, 2017, the Bank of Canada, Payments Canada, and R3 released a white paper outlining their detailed findings from Project Jasper. This article elaborates on our previous article based on the findings from the white paper and discusses the next steps for Project Jasper.
Key Merits and Considerations of Project Jasper
Project Jasper was premised on the idea that payment settlement is the final leg of most economic transactions, but also that other areas of the contract chain have the potential to be supported by DLT. For example, “smart contracts” can be used to codify the terms and conditions of an agreement and can be automatically executed once certain conditions are met. Based on the experience from Project Jasper, the primary benefit of a DLT interbank cash payment platform would be an “end-to-end” settlement, meaning that the DLT arrangements for payment settlement would be aligned with other DLT arrangements within the same economic contract.
Principle 8 of the Principles for Financial Market Infrastructures (“PFMIs”) requires that a settlement must be final and irrevocable. Settlement finality in Phase 1 was “probabilistic” because of the possibility that a payment could fail to remain in the blockchain and be recorded under a proof-of-work consensus. To address this and improve settlement finality, Phase 2 introduced a notary node to be managed by a trusted third party. The Bank of Canada served as the notary node and was responsible for confirming the uniqueness of a transaction to avoid double spending. The second requirement under PFMIs is that there be a full and irreversible transfer of an underlying claim in central bank money. To meet this, Project Jasper created a digital depository receipt (“DDR”) as a digital settlement asset, which represented a claim to central bank deposits. The strength of the legal basis for settlement finality remains to be tested.
Operational Resilience and Efficiency
Project Jasper used a “permissioned” DLT, meaning that only those approved could use the exchange. This allows regulation of users and allows consensus to be achieved more quickly than with a public ledger. DLT solutions can also reduce the number of errors and duplications compared to the incumbent, manual systems in Canada because parties are required to reach a consensus before a transaction is posted. However, due to the limited implementation of the project, it is difficult to assess whether DLT is more operationally efficient than the current system.
The white paper also considered the operational resiliency of Project Jasper and noted the following:
Phase 1: The maximum processing capacity was 14 transactions per second, which is similar to incumbent systems, meaning there are constraints for future volume increases.
Phase 2: There is capacity for volume increases, in part because only the transacting parties, a supervisory node and the notary node are required to validate and record transactions (vs. the requirement for majority consensus in Phase 1).
Availability and Cost
Phase 1: The proof-of-work consensus allows for high availability at a lower cost. This is because of the sharing of databases across all participants in the proof-of-work consensus and the back up of ledgers by all participants.
Phase 2: To increase data privacy, each participant had a proprietary ledger. This creates challenges for data replication across the network.
Phase 1: The consensus protocol requires agreement of a majority of R3 members, meaning there could not be a single point of failure.
However, this does not eliminate the need for participants to back up their data. Due to the confidentiality of the information, in the event of a failure, participants would be unlikely to share data.
Phase 2: Both the notary and supervisory nodes are needed for consensus, therefore increasing the risk of a single point of failure. To mitigate this risk, participants will need to back up their data.
Potential Applications & Benefits of DLT to the Payments Industry
Reduction of Disputes and Errors
A single payment or file transfer can involve many participants, and therefore may be recorded by multiple financial institutions. This can lead to errors and duplication, and inevitably, disputes. DLT technology requires multiple parties to reach an agreement on the legitimacy of a transaction before it can be posted. While this is a recognized benefit of DLT, the overall operational efficiency of this benefit compared to the incumbent system has not been measured.
Improved Back-Office Efficiency
After the parties reach a consensus, a single record of the transaction is recorded. This eliminates the need for internal record keeping of each party. Project Jasper found that DLT is not necessarily more efficient on a domestic level than the current LVTS system, however the analysis did not account for the back-office work that might be avoided by the individual financial institutions if DLT is used. Significant resources are expended in back-office reconciliations; therefore there may be significant cost savings that have not yet been considered.
DLT has the potential to assist with regulatory compliance, particularly with anti-money laundering (“AML”) and anti-terrorism financing (“ATF”) regulations for cross-border transactions where counterparty risk can run high. In the current system, false positives in relation to AML/ATF are a problem as they can take weeks or months to resolve. DLT has the potential to allow for easier reconciliation of such payments in order to legitimize a transaction because of the trusted ledger created. These benefits could extend to other regulatory compliance as well.
Transparency vs. Privacy
In the traditional clearing and settlement process, there is a central database. The DLT used in Phase 2 allows for privacy between the financial institutions, with each only being able to view their own proprietary ledgers. However, those with the supervisory or notary nodes can view all transactions and therefore have the ability to monitor and perform the traditional function of a central database. In Phase 2, the Bank of Canada held the supervisory and notary nodes.
Improved Automation through use of Smart Contracts
As previously discussed, significant benefits can be obtained where DLT can be used for end-to-end settlement through the use of smart contracts. The solution system created by Project Jasper could be the basis upon which other DLT platforms can be built for a variety of transactions, such as the settlement of financial asset transactions, managing syndicated loans, and supporting trade finance.
Conclusions from Phase 1 and Phase 2
The key conclusions from Phase 1 and 2 of Project Jasper are that DLT platforms that employ a “proof-of-work” consensus protocol, as used in Phase 1, do not deliver the required settlement finality and low operational risk. While Phase 2 was able to address improvements in settlement finality, scalability and privacy, it did not adequately address operational risks requirements. Further evaluation and enhancements will need to be done to satisfy PFMIs. On a global scale, the white paper recommends that the focus should be on developing protocols for interoperability between DLT platforms.
Overall, Project Jasper is an example of the benefits of collaboration within the payments industry. Such collaboration is particularly conducive in the concentrated Canadian financial industry. This collaboration is being extended for Phase 3.
On October 17, 2017 Payments Canada, the Bank of Canada and TMX announced the third phase of Project Jasper. This phase will build on the first two phases and involve developing a proof of concept for the clearing and settling of securities. Phase 3 hopes to explore an end-to-end settlement process by integrating the securities and payment infrastructure and the ability to settle multiple assets on the same ledger. The objectives of this phase are to reduce the cost of securities transactions, increase efficiency, and reduce settlement risk. The results of this phase are expected to be released at the Payments Canada Summit in May 2018.