Dynamic volatility connectedness of commodity markets: unravelling global crisis impacts and key drivers of connections

Absztrakt (kivonat)

This study investigates the static and dynamic relationship between all six categories of commodities (energy, industrial metals, precious metals, grains, softs, and livestock) encompassing 33 highly traded commodity assets. The network of six aggregated groups is analyzed beforeand after major global events and crises. It also aims to assess whether different macroeconomic, financial, and geopolitical factors have a significant effect on various connectedness measures across the six groups. The network of commodity volatilities is explored by employing LASSO regularization and the connectedness measures based on the Diebold and Yilmaz framework (Diebold and Yılmaz [2014]). This framework, utilizing a parameter vector autoregression (VAR) model, offers an insight into the dynamic connectedness among the daily volatilities of commodity futures. The sample period ranges from 25th August 1998 to 1st of April 2024, including several relevant global events and crises such as the Global Financial Crisis, the European Debt Crisis, the Crude Oil Collapse, the COVID-19 Pandemic, and the Russo-Ukrainian War. The result shows rather low inter-group connectedness and high intragroup connectedness. The dynamics of the total and the aggregated net connectedness measures display time-varying behavior, with an increased level of connectedness during crisis periods. The Baltic Exchange Dry Index (BDIY), the VIX and the S&P500 are shown to have a significant role in shaping the dynamics of various connectedness measures, alongside the GFC and the COVID-19 Pandemic. However, after testing the significance of several previously identified changes in the connectedness network graphs due to crisis events, - employing the moving-block bootstrap method - the results show only a few key differences.