Globally Distributed Production and Asset Pricing:the Rise of Latin America in CME Soybean Futures

The mechanism of aggregation of various sources of fundamental information into a single price is a central question in asset pricing. In this paper I investigate how information about local supply shocks in the globally distributed production of commodities is incorporated into the prices of future contracts traded at the Chicago Mercantile Exchange. I take advantage of three facts: first, the global production of soybeans is heavily concentrated in the US and Latin America. Second, the market shares of these dominant regions have varied over the last fifteen years as Latin America’s output has surpassed that from the US. Third, local daily rain precipitation is a source of exogenous local supply shocks because it tends to increase output. I exploit these facts and estimate the sensitivity of CME soybean futures prices to local rain precipitation, in the US Midwest and Argentina, for the periods 1996/7-2002/3 and 2003/4-2010. I find economically significant negative rain impact on prices. I also find that, across periods and regions, the magnitude of local rain impact on the CME price has a strong linear relationship with the share of global output produced locally, with correlation in excess of 93%. This suggests that the CME price reflects a globally integrated market for soybeans, in which distributed supply shocks are aggregated linearly.