Date of Graduation


Document Type


Degree Name

Master of Science in Agricultural Economics (MS)

Degree Level



Agricultural Economics and Agribusiness


Andrew McKenzie

Committee Member

John D. Anderson

Second Committee Member

Eunchun Park


FAS;Futures Comodities;Large Export Sales Reports


This study sought to determine if the Large Export Reports from the USDA Foreign Agriculture Service have any discernible market-moving impact on particular commodity futures markets. The effect of the reports on futures markets is investigated using four different methods: the average return event study approach, the multiple regression with binary variables model, the autoregressive time series model, and the event spline model. The average returns showed that the FAS's Large Export Sales Reports had no influence on the pricing of wheat futures, but did have a considerable impact on the returns from days before and after the release of the sales report. It became evident that the reports' impact on returns on the futures price depended on the magnitude of sales and the sales destination (China, Other, or various destinations, including China). The size of the sale in the large export sales reports positively impacted both corn and soybean returns, according to the preliminary findings of the multiple linear regressions. Yet, only large corn export reports showed statistically significant destination binary variables. Reports having a country or countries other than China as their destination (Other) and reports with numerous destinations, including China as their destination (Combination), were determined to be statistically significant, where they had a positive and negative connection, respectively. The returns for the whole reporting period were the next thing we wanted to examine. In addition to assessing the entire data period, we also produced new time series datasets to evaluate whether reports that fall in the event window of other reports alter their effects. Compared to the first multiple linear regressions, the original time series data from an ARMA-GARCH model showed the same outcomes. The volume of the sale, and more precisely, the volume of the sale with various destinations, are statistically important in corn reports. The influence of the variables on the returns was not changed by the regression findings using reports released within the event window. Furthermore, a linear spline model was employed to illustrate how the returns are influenced by the various sales sizes in reports. For corn, we discovered that larger sale sizes have a far more significant influence on returns, but the exact reverse is true for soybeans. Contrary to the primary goal of this study, there is little data to suggest that China alone has a significant market influence. Instead, price shifts occur as soon as the market has a chance to respond to the announcement of the significant sale report, which is often within the first two minutes of trading. Making the reports available in real-time or at another time that does not coincide with trading hours might be advised to evaluate the implications of these reports more thoroughly so as to determine if the sale was known in the market before publication. The findings of the numerous soybean returns analyses indicate that the futures price returns are mostly unaffected by the magnitude of the sale in reports and are unaffected by the destinations. In addition, different representations of the size of reports and particular destinations changed throughout the research, which affected how the results were scaled.