Reinvestigating international crude oil market risk spillovers to the Chinese financial market via a novel copula-GARCH-MIDAS model

In order to accurately and reasonably investigate risk spillovers from the international crude oil market to the financial market, we develop a copula generalized autoregressive conditional heteroscedasticity mixed-data sampling (copula-GARCH-MIDAS) model to estimate the joint probability distribution of multivariate variables, and we then derive conditional-value-at-risk-type (CoVaR-type) risk measures. Our method has three main steps. First, we formulate a GARCH-MIDAS model with a long-run volatility component driven by macroeconomic fundamentals such as gross domestic product, consumer price index and money supply to fit the marginal distribution of a single market. Second, we apply the copula technique to model dependence among multiple markets. Third, we derive the joint distribution using the fitted marginal distribution and the estimated dependence structure, and we also calculate CoVaR-type risk measures. Our empirical studies on risk spillovers from the international crude oil market to the Chinese financial market show that the copula-GARCH-MIDAS model is promising and that it is superior to the standard copula-GARCH model. We find that macroeconomic fundamentals are very important to improve the accuracy of the CoVaR measure. In addition, the effects of more severe distress events in the international crude oil market on the Chinese financial market are huge.