An assessment of the Bank of Canada's term PRA facility

This paper empirically assesses the effectiveness of the Bank of Canada's term Purchase and Resale Agreement (PRA) facility in reducing short-term bank funding pressures, as measured by the CDOR-OIS spread. It examines the behaviour of this spread around both term PRA announcement dates and term PRA operation dates, using an event-study methodology to control for developments in other money markets (i.e., using the U.S. LIBOR-OIS spread) as well as proxies for Canadian banking sector credit risk. Overall, there is robust evidence that the term PRA announcements reduced bank funding costs at both 1-month and 3-month terms, whereas we find no evidence of an impact from term PRA operations. However, given the small number of term PRA announcements in our sample, caution should be taken in attributing the reduction in the CDOR-OIS spread solely to the term PRA announcements, since other concurrent events (including other announcements by the Bank of Canada) may have also contributed to a compression in the CDOR-OIS spread.