Forward Pricing Behavior of Corn and Soybean Producers
Davis, Todd D., Patrick, George F., Coble, Keith H., Knight, Thomas O., Baquet, Alan E., Journal of Agricultural and Applied Economics
Forward pricing behavior of random samples of Indiana, Nebraska, and Mississippi crop producers was analyzed using Heckman's two-step limited information maximum likelihood estimation procedure. Producers who forward priced during the 1995-1998 period generally expected to forward price in 1999 using similar techniques. Probit models were estimated for cash forward contracts and taking a direct position in futures or options separately and combined. Results provide limited support for the hypothesis that forward pricing should be analyzed as an adoption decision. Variables reflecting risk attitudes do affect the decision to use forward pricing, while variables related to economic position affect the level of forward pricing.
Key Words: forward contracts, futures, grain marketing, Heckman procedure
JEL Classifications: Q130, Q120
Understanding of the factors that influence producers' marketing behavior, including the choice of pricing alternatives, is an ongoing area of research for agricultural economists. Marketing education programs commonly encourage grain producers to use preharvest forward pricing as a way to manage price risk by avoiding selling at harvest. However, producers differ in their attitudes toward forward pricing and their willingness to use futures, options, and cash forward contracts. Empirical studies have found that an increasing percentage of producers have adopted forward pricing techniques over time. For example, Hill reported that 4% of surveyed Kansas grain farmers hedged and that 12% used forward pricing in the early 1970s. Asplund, Forster, and Stout found that 42% of Ohio grain producers surveyed in 1986 used forward pricing, while only 7% of the respondents used hedging. Goodwin and Schroeder, in a 1992 survey of Kansas producers, found that 45% used forward contracting, 11% used hedging, and 19% used options to manage price risk in their farming operation. Sartwelle et al. found that 70% of Kansas, Iowa, and Texas producers surveyed in the late 199Os used forward contracts and that 52% used futures and options contracts.
Sartwelle et al. used multinomial logit models to categorize producers' grain marketing orientation and analyze the associated socioeconomic characteristics. Other studies have used Tobit models (Goodwin and Schroeder; Musser, Patrick, and Eckman; Sartwelle et al.; Shapiro and Brorsen) to analyze the effect of socioeconomic variables on the adoption and use of forward pricing. Tobit models account for censoring of observations at zero (no forward pricing) but also impose a restriction on economic behavior. Park and Florkowski indicate that any factor that determines the probability of an individual forward pricing also has the same impact on the amount forward priced. Transactions costs, such as the costs that Townsend and Brorsen have identified as being associated with using futures, options, and forward contracts, are not reflected in a Tobit model. An alternative model, the Heckman two-step procedure, is used in this study to analyze the possible effects of these transactions costs.
Previous studies of marketing behavior have used nonrandom samples of producers (Goodwin and Schroeder; Musser, Patrick, and Eckman; Sartwelle et al.; Shapiro and Brorsen). Those studies in which random samples have been used (Asplund, Forster, and Stout; Hill) have been limited to a single state. This study uses large, stratified random samples of producers in multiple states to determine if socioeconomic variables have separate effects on the decision whether to forward price and on the level of preharvest forward pricing of soybeans and corn for the 1999 crop year. Forward pricing through the use of cash forward contracts or by taking a direct position with a futures or options contract is first considered separately, and then all forms of forward pricing are combined.
Previous studies of the use of marketing techniques have treated the decision to use a particular technique as an adoption of technology decision (e. …