Volatility in U.S. dairy futures markets

Abstract

Essay 1: Despite its sheer market size and volatile prices, the study of US dairy commodity prices and volatility is limited in recent literature. Compared to other commodities, dairy products show large price swings and unpredictable price volatility, making it more difficult for dairy farmers, dairy manufacturers, and other market participants to manage price risks. This essay investigates dairy commodities futures returns using five different return metrics, and corresponding realized volatilities at weekly and monthly levels. We apply GARCH models to consider the conditional heteroskedasticity in dairy futures returns. Our results show that dairy commodities returns are correlated positively, but only cheese and milk have strong correlations (over 85% coefficient for returns and over 90% for volatility.) We also report strong evidence that volatilities of all dairy commodities are significantly impacted by short-term return surprise (innovation) and past volatility. However, only dry whey exhibits dominant volatility persistence (long memory.) Our GJR GARCH results suggest that "good" news in return impacts butter and cheese volatility greater than "bad" news, while "bad" news impacts milk and dry whey greater than "good" news. Using multivariate GARCH DCC model, we show the return comovement are time-varying and volatility are interdependent among dairy commodities. The GARCH BEKK model indicates asymmetric (bi-directional) volatility spillover effects between dry whey and other commodities, especially between dry whey and cheese. We also demonstrate that dairy commodities return and volatilities, especially milk and cheese, are not closely related to macroeconomics variables possibly because of the unique characteristics of dairy products and heavy government regulation. Essay 2: The classified dairy pricing scheme under the Federal Milk Marketing Order (FMMO) and the cash settlement feature in the dairy commodity futures markets generate unique volatility patterns. This study shows that the USDA's weekly wholesale price announcement, commonly called the National Dairy Products Sales Report (NDPSR), influences futures prices and reduces volatility. Our results demonstrate that the volatility diminishes when a futures contract approaches expiration for all dairy commodities, including class III milk, cheese, butter, and dry whey. We provide further evidence that the dairy futures market responds to the USDA announcement by showing that trading volume and open interest are highest on the announcement days, and volatility peaks, for all dairy commodity futures, in-between the announcement days. Furthermore, we develop an information uncertainty measure to capture the information released by the USDA announcement. Using OLS regression models at the individual commodity level and VAR models considering the interdependences among dairy commodities, we show that the diminishing volatility is significantly correlated with the information uncertainty for all dairy commodities. Trading activities, such as trading volume and open interest, also contribute to the diminishing volatility. The short-term interest rate impacts the volatility of butter, cheese, and milk, while dollar index impacts dry whey volatility. Our VAR results show that volatilities of butter and dry whey are more independent, while volatility of cheese shows a moderate impact from butter volatility. Our results are robust to different models and seasonality effects.