The time-varying linkages between global oil market and China's commodity sectors: Evidence from DCC-GJR-GARCH analyses

Dynamic Spillovers Between International Crude Oil Market and China's Commodity Sectors: Evidence From Time-Frequency Perspective of Stochastic Volatility

“…Previous empirical studies rely mostly on time-series models such as the ARCH and GARCH models (Engle, 1982;Engle and Bollerslev, 1986) to capture volatility structure of commodity prices (Vivian and Wohar, 2012;Efimova and Serletis, 2014;Youssef et al, 2015;Jiang et al, 2019). However, these time-series models do not provide a clear unified methodology to reveal volatility dynamics operating between the involved variables and to identify structural changes (Jebabli et al, 2014).…”

“…Since the global financial crisis (GFC), a growing number of studies are conducted to explore the connectedness between crude oil and commodity markets, and their methods can be broadly classified into several categories: VAR or structural VAR (SVAR) (Wang et al, 2014;de Nicola et al, 2016); GARCH models (Ji and Fan, 2012;Ewing and Malik, 2013;Jiang et al, 2019); Copula models (Koirala et al, 2015); nonparametric causality analysis (Nazlioglu et al, 2013); vector error correction model (VECM); Markov regime switching (MRS) models (Uddin et al, 2018); and forecast error variance decomposition (FEVD) (Diebold et al, 2017;Lovcha and Perez-Laborda, 2020). However, the previous literature generally underestimates connectedness among commodity markets of a particular class or group, while there are few studies focusing on the oil-commodity nexus at the industry level.…”

“…We select a group of nine commodity sectors, including precious metals (NMFI), nonferrous metals (NFFI), coal coke and steel ore (JJRI), nonmetal building materials (NMBM), energy (ENFI), petrochemicals (CIFI), grains (CRFI), oils and fats (OOFI), and soft commodities (SOFI). Compared with the sectors in the existing literature (Chen, 2015;Jiang et al, 2019), these nine commodity sectors are comprehensive selections for the analysis of China's commodity markets. They can not only capture the price fluctuations of the entire commodity markets in China but also reflect the supply and demand relationship of different commodity sectors.…”

“…Using the time-series framework, we adopt the methodological approach that comprises the stochastic volatility (SV) estimation proposed by Kastner and Frühwirth-Schnatter (2014) and the time-frequency volatility connectedness estimation proposed by Baruník and Krehlík (2018) in a multivariate framework. Although previous literature mostly used the implied volatility model (Vivian and Wohar, 2012;Efimova and Serletis, 2014;Youssef et al, 2015;Jiang et al, 2019), this study focuses on the SV model, because the volatility of commodity market price is a stochastic process, and the SV model can perform better than the implied volatility or historical volatility measurements Balli et al, 2019). The motivation for this study is that it is crucial to measure the dynamic connectedness among the commodity or futures markets and the determinants of volatility spillover effect through novel methodologies.…”

“…Second, we combine the SV model and the GFEVD framework of the DY (2012) (DY12) to explore the dynamic spillovers between crude oil markets and China's bulk commodity. Previous studies (Vivian and Wohar, 2012;Efimova and Serletis, 2014;Youssef et al, 2015;Jiang et al, 2019; adopt models that use historical or implied volatility measures. We highlight that the volatility of crude oil and other commodity price is a naturally stochastic process and that the SV model is superior in modeling commodity volatilities due to limited assumptions.…”