The increasing importance of risk management in payment systems has led to the development of an array of sophisticated tools designed to mitigate tail risk in these systems. In this paper, we use extreme value theory methods to quantify the level of tail risk in the Canadian retail payment system (ACSS) for the period from 2002 to 2015. Our analysis shows that tail risk has been increasing over the years, but the pace of growth has been reduced towards the end of our data sample, which suggests a slower rate of growth of collateral required to cover that risk.