Do economic variables forecast commodity futures volatility?

Abstract

This paper explores empirically whether the supply or the demand uncertainty, the time to maturity, and the slope of the term structure (storage), explain the realized volatility of nearby commodity futures 5-min returns. I find support for the "uncertainty resolution" and the "theory of storage" hypotheses while the "time to maturity" hypothesis is rejected.These results are robust to the inclusion of autoregressive terms in the baseline model. Next, I evaluate the in-and out-of-sample forecasting ability of models including these economic variables and find mixed results. Finally, I test the validity of these forecasts in expected shortfall modeling.

Performance comparison of alternative stochastic volatility models and its determinants in energy futures: COVID‐19 and Russia–Ukraine conflict features

“…A plausible motivation for this interest is the need for accurately measure the volatility of energy futures prices. In fact, this variable plays a key role in the price connection between spot and futures markets—see, for example, Silvapulle and Moosa (1999), Lin and Tamvakis (2001), Hammoudeh et al (2003), Hammoudeh and Li (2004), Huang et al (2009), and Balcilar et al (2015)—on the linkage with economic and financial variables—as documented by Maréchal (2021), Prokopczuk et al (2021), Kupabado and Kaehler (2021), and Xu and Wang (2021)—and on risk management problems—see, for instance, Sadorsky (2006), Aloui and Mabrouk (2010), Yin et al (2021), and Ammann et al (2022). The time‐varying dynamics and contagion in commodity futures are explained in detail in Mehlitz and Auer (2021) and Gong, Jin et al (2022), respectively.…”