Open Access

Amelie Schischke, Andreas Rathgeber

• In this study, we examine the evolution of commodity markets, focusing on the role of fundamentals and cross-market co-movement in shaping commodity price dynamics over time. To capture these time-varying spillover effects, we apply a Markov-switching global vector autoregressive (MS-GVAR) model to industrial metal markets in the period from January 1995 to December 2020. By distinguishing between calm and volatile regimes, our results reveal significant cross-commodity responses, emphasizing the importance of modeling commodity markets jointly. Spillover effects from supply and demand fundamentals within and across markets further highlight the critical role of fundamental drivers in commodity price formation. While the significance of spillover effects remains consistent across regimes, their magnitude intensifies during volatile periods, indicating heightened spillover risks. Additionally, we demonstrate that our time-varying model provides a superior representation ofIn this study, we examine the evolution of commodity markets, focusing on the role of fundamentals and cross-market co-movement in shaping commodity price dynamics over time. To capture these time-varying spillover effects, we apply a Markov-switching global vector autoregressive (MS-GVAR) model to industrial metal markets in the period from January 1995 to December 2020. By distinguishing between calm and volatile regimes, our results reveal significant cross-commodity responses, emphasizing the importance of modeling commodity markets jointly. Spillover effects from supply and demand fundamentals within and across markets further highlight the critical role of fundamental drivers in commodity price formation. While the significance of spillover effects remains consistent across regimes, their magnitude intensifies during volatile periods, indicating heightened spillover risks. Additionally, we demonstrate that our time-varying model provides a superior representation of interdependencies between commodity prices compared to a time-invariant benchmark model, as validated through an out-of-sample forecasting analysis.…