How vulnerable is the Canadian banking system to fire-sales?

Using a model based on the work of Duarte and Eisenbach (2015), I stress test the Canadian banking system using publicly available data from 1996 to 2015. I find that the Canadian banking system is resilient to all but the most extreme event. This is because (i) strong macroprudential regulation in Canada improves the quality of assets, (ii) given a scenario of severe losses, banks retain a sufficient quantity of liquid assets that these could be used to meet short-term liabilities, and (iii) sufficient equity is available to absorb losses. However there remain some areas of concern. Using aggregate vulnerability (AV), I find that the Canadian banking system has become more vulnerable to a fire-sale episode since 2011 and this raises the possibility of future losses. I focus my study on Canadian assets, but these have grown much faster than the overall economy since 2004. The concentration of loans-to-households, including residential mortgages and consumer loans, should be of some concern to regulators. Indeed, recent government announcements and policy changes reflect an overall concern with the ability of households to repay their loans. In the fourth quarter of 2015, if the government decided to remove its explicit guarantee of mortgage insurance, losses due to a fire-sale event would have been 20 percent higher.