Predictability of tail risks of Canada and the US over a Century : the role of spillovers and oil tail risks

Abstract

We examine the tail risk spillovers between Canada and U.S. stock markets using over a century data, and also account for the roles of tail risks of other advanced economies (France, Germany, Japan, Italy, Switzerland, and the UK) and oil-market tail risk. We use the “best” tail risk measure obtained from different variants of the Conditional Autoregressive Value at Risk (CAViaR) model developed by Engle and Manganelli (2004) in the predictive model and compare its performance with that of an AR(1) benchmark model. We find strong evidence of risk spillovers between the two stock markets. We find contrasting evidence for the predictability of oil-market tail risk, with positive predictability in case of the net oil exporter and negative in case of the net oil importer. Further results using tail risks of other advanced economies (combined) support possible diversification potential for Canadian stocks in the presence of market risks of advanced economies other than the U.S. Our findings have implications for investors and are robust to various out-of-sample forecast horizons, alternative data frequencies, data splits, and 1% and 5% VaRs.