Payment coordination and liquidity efficiency in wholesale payments systems

Francisco Rivadeneyra and Nellie Zhang

Abstract

A new wholesale payments system launched in Canada in 2021. This real-time gross settlement system, called Lynx, has two types of settlement mechanisms, one allowing offsetting and the other not. This paper studies the decision problem of the Bank of Canada: which of the two settlement mechanisms it should use to send its payments. Using extensive simulation, we show that, mainly due to the benefits of liquidity pooling, Lynx would achieve its highest liquidity efficiency (even better than that of the current Large Value Transfer System (LVTS)) if all payments (urgent and nonurgent) from all participants were sent to the mechanism allowing offsetting. The minimum amount of liquidity required to settle all payments by critical intraday deadlines is approximately C$10 billion, around half the amount of collateral that LVTS participants allocated pre-Covid-19. Since time-critical payments sent to the offsetting mechanism could experience a delay, the high level of liquidity efficiency is accompanied by an increase in the number of participants’ operational interventions (to pledge more collateral or to alter payment priorities) to ensure that those time-critical payments are never delayed. When coordination does not occur, liquidity efficiency can be far lower than in the LVTS. The results highlight that the Bank of Canada helping with coordination is more important than the specific choice of mechanism.