The main purpose of this study is to build a monthly supply and demand outlook model for four key agricultural and livestock commodities: onion, Chinese cabbage, pork, and broiler. The study was launched as part of a pilot project by the Korea Rural Economic Institute (KREI) to model the market outlook for six agricultural and livestock commodities in 2010.
In order to reduce prediction errors and secure realistic forecasting, onion is subdivided into two categories based on the type of variety: the precocious and the mid-to-late ripening. Chinese cabbage is divided into five categories based on when or where it is grown: Winter, Greenhouse, Spring, Summer, and Autumn. A supply and demand model for pork and broiler was developed after considering biological components and growth process.
The monthly outlook model is made up of MS-Excel for easy operation and simulation under various scenarios. If the data has been changed, the model automatically calculates the outlook results and revises the graph again.
The model file consists of six worksheets: Main, Survey, Total DB, Equation, Table, and Graph. The Main part explains variables and the model's flowchart. Total DB compiles various variables, statistics and research data. The Survey part contains the data obtained by monitors, sample farmhouses, storage firms, and consumers. The Equation part consists of various regression equations. The Table and Graph sections show outlook results in the form of tables and graphs.
There can be some cases where forecast results are different from survey data. To solve this problem, the outlook model is constructed in such a way as to allow forecast results to be reciprocally connected with monthly survey data and enable the model operation chief in charge of a specific commodity item to compare forecast results with survey data. The model system also allows the model operator to choose between outlook results and short-term survey data. In addition, the outlook model is usable for policy simulation.
The survey data and government statistics data should be updated every month to improve the accuracy of outlook results. For the sake of continued utilization of the model, it is also necessary to update data at least once a year and reestimate the parameters of the equation.