Time-Varying Risks in Commodity Markets

Commodities are an integral component of the global economy, serving critical functions in food production, manufacturing, and industrial processes. Producers and consumers of commodities are exposed to unwanted price fluctuations that increase the risk of doing business, and they commonly use commodity futures contracts to hedge price risk. Effective hedging necessitates a competent understanding of the behavior of commodity futures returns. Existing work on commodity futures tends to focus on the characterization of risk premia (e.g., Gorton and Rouwenhorst, 2006) or a trade-off between risk and return that restricts the quantity of risk to be static (Szymanowska et al., 2014; Bakshi et al., 2019). That is, while risk factor realizations are allowed to vary over time, factor exposures are typically held constant. There remains limited work on dynamic trade-offs between risk and return through the channel of time-varying factor exposures. The goal of this proposal is to achieve a deeper understanding of how dynamic risk-return trade-offs in commodity markets can arise from time variation in factor exposures, as well as the associated conditional asset pricing implications. Our proposal contains four parts. In the first part, we seek to quantify the number of factors necessary to explain the comovement in commodity premia and to document their behavior. How many independent sources of comovement can we identify? How much variation exists in these common components, and how do they change through business cycles? The second part is concerned with identifying the most informative conditioning variables for the systematic risk exposure of commodity premia. We look for a subset of conditioning variables that contributes maximally to explaining a panel of commodity returns. The identity of the most important variables helps to pinpoint the economic mechanisms that generate commodity risk premia.  The third part pertains to building a conditional asset pricing model to explain the cross section of commodity risk premia. How well do common factors that explain comovement price the cross section of commodity returns? Finally, the fourth part compares our model description of commodity risk premia with alternative models from the literature. We investigate whether proposed asset pricing factors from the literature contribute to the description of commodity risk premia while controlling for the factors our model uncovers, raising the hurdle for economic and statistical significance.