Realized correlations, betas and volatility spillover in the agricultural commodity market: What has changed?

This article provides new insights on the changes in the dynamics of price correlations and spillover effects in the commodity market. Using US-traded futures price data at a 1-min frequency over the 2002–2017 period, we consider the interaction within soft and grain commodities and between these commodities and oil. We rely on a recently introduced volatility model – the realized Beta GARCH model of Hansen et al. (2014). Our results reveal that soft commodities were segmented prior to 2008 and became correlated thereafter. The nature of the increase in correlation is only temporary. The correlations within grains – already significant and positive – increased only marginally, indicating that this group has been less affected by recent events. The correlation between oil and agricultural commodities, which reached its peak in 2008, has also reverted to pre-crisis level. Spillover effects between oil and commodities have become more prominent prior to the commodity price crash. However, this increase in volatility transmission tends to precede the increase in correlations. Finally, the impact of these findings on the performance of hedging strategies and optimal portfolio weights is discussed. Our results are important for investors exposed to the commodity market as they show that while the diversification benefits of investing in this market have decreased, volatility transmission risk and hedging costs have increased.