Date of Award

12-2016

Culminating Project Type

Thesis

Degree Name

Applied Economics: M.S.

Department

Economics

College

School of Public Affairs

First Advisor

Ming Chien Lo

Second Advisor

Mana Komai

Third Advisor

David Robinson

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Keywords and Subject Headings

Forecasting, Matlab, Statistics, Agriculture, Price, Corn

Abstract

Of the current techniques used to forecast agricultural commodity prices, none carries as high of a cost as a supply and demand analysis. Because of this expense, firms that have the ability to produce forecasts that rely on supply and demand analysis, do not update their models very frequently. In this paper we will examine if statistical methodologies can provide price forecasts at least as accurate at supply and demand analysis techniques. Both statistical as well as supply and demand models will be evaluated at one, three, six, nine, and twelve month horizons. These horizons are typical for price forecasts that can be used for buying and selling activities, contract negotiations, and production decisions. Of the models we investigated an autoregressive model was found to provide the best price forecasts over a shorter horizon. Whereas a vector autoregressive model was shown to provide the best price forecasts beyond a six-month horizon. This study reveals that a few statistical techniques have the ability to outperformed models that incorporate supply and demand analysis in forecasting the price of U.S. corn.

Comments/Acknowledgements

A special thanks to Dr. Ming Chien Lo who worked on this project with me tirelessly to see it to completion.

Share

COinS