Causality and Price Discovery in U.S. Corn Markets: An Application of Error Correction Modeling and Directed Acyclic Graphs

This study investigates dynamic relationships among U.S. corn cash prices for the years 2006-2011. With daily data from 182 spatially separated markets spreading across 7 states, Iowa (IA), Illinois (IL), Indiana (IN), Ohio (OH), Minnesota (MN), Nebraska (NE), and Kansas (KS), we apply an error correction model and directed acyclic graphs to identify the contemporaneous casual relationships among prices of different states, where the price of each state is calculated as the average of the prices of observed markets in it. Our empirical results show that 3 states, IA, OH, and MN, can dominate the corn cash prices. Depending on the way the causal flows are assigned among them, OH and MN together, MN, OH, or IA is the most important in pricing. If additional empirical evidence becomes available, more precise results can be expected. We also divide the data into a store period and a harvest period, and adopt a VAR in differences to model the price relationships. While IA and OH dominate the corn prices during the store period, IL, IN, OH, and MN can be essential in pricing without more evidence to narrow our attention.