Using Synthetic Data to Measure the Impact of RTGS on Systemic Risk in the Australian Payments System

This paper examines the possibility that financial contagion may be spread from one bank to another via the Australian payments system. The initial study of payments system risk was undertaken by Humphrey (1986) who found significant risk in the U.S. Fedwire system in the mid 1980s. Subsequent studies by Angelini, Maresca & Russo (1996), Kuussaari (1996), Northcott (2002) and Furfine (2003) have found, however, little evidence of systemic risk in the payments systems of Italy, Finland and Canada, and in the U.S. inter-bank market. Given that the implementation of real time gross settlement (RTGS) systems in many countries, including Australia, at significant cost, has been designed to reduce payments system risk, the finding that this risk is small is significant. While detailed payments system data for Australia is not available to researchers outside the Reserve Bank, this study constructs a synthetic data set based on available information and uses this data to simulate the failure of each financial institution operating in the Australian payments system. We find little evidence of systemic risk in the Australian payments system using this approach and conclude that the introduction of RTGS in the Australian system in 1996 had only a marginal effect on risk.