Forecasting Chinese crude oil futures volatility using dynamic volatility spillover CARR-MIDAS model

This paper proposes a dynamic volatility spillover conditional autoregressive range-mixed data sampling (DVS-CARR-MIDAS) model to forecast volatility in the Chinese crude oil futures market by incorporating dynamic volatility spillovers from the dominant US crude oil futures market to the emerging Chinese crude oil futures market. Empirical results based on West Texas Intermediate (WTI) crude oil and Shanghai International Energy Exchange (INE) data revealed significant and time-varying spillover effects from the US to the Chinese market. In addition, the DVS-CARR-MIDAS model consistently showed that the proposed model consistently outperforms benchmark models in both in-sample fitting and out-of-sample forecasting. These findings were robust to the Diebold-Mariano (DM) test, $R_{oos}^2$ test, alternative dominant market, and different out-of-sample forecast windows. Furthermore, the economic value analysis demonstrated that the proposed model provides meaningful benefits for portfolio management.