Payment Coordination and Liquidity Efficiency in the New Canadian Wholesale Payments System
As part of the Canadian payments modernization initiative, the Large Value Transfer System (LVTS) will soon be replaced by a new real-time gross settlement system called Lynx. We identify which of two Lynx settlement mechanisms the Bank of Canada should use for sending its payments: one that allows offsetting and one that does not. We also consider what effect such a choice has on the liquidity usage, payment delays and operational efficiency of the new system.
We use a payments system simulator developed at the Bank of Canada that replicates the functioning of the Lynx system. This tool performs simulation exercises with a large sample of actual LVTS data in several scenarios using the different Lynx settlement mechanisms. We assume the timing of payment submission in the simulated Lynx is the same as in the LVTS. In addition, these simulation exercises contain behavioral rules that simulate Lynx participants’ continuous liquidity monitoring and operational intervention efforts to ensure all critical payments settle on time.
The results show that Lynx achieves significantly higher efficiency, due to liquidity pooling, if all payments are settled in the mechanism allowing offsetting. When only this liquidity-saving mechanism is used, the minimum amount of collateral required to settle all payments by their critical timelines would be approximately $10 billion, or about half the collateral allocated in LVTS (pre–COVID-19), with an associated 10-minute weighted average delay. When the two settlement mechanisms are used in tandem, payment coordination falls, and a higher level of liquidity is required while introducing longer settlement delays. Our results show how important the Bank of Canada’s choice of settlement mechanism in Lynx is, as it affect the overall efficiency of the system operation.