A Quarterly Econometric Model for Short-Term Forecasting of the U.S. Dairy Industry

• 查看报告

英文名称：

A Quarterly Econometric Model for Short-Term Forecasting of the U.S. Dairy Industry

作者：

Roberto Mosheim

工作单位：

USDA's Economic Research Service

来源地址：

https://www.ers.usda.gov/publications/pub-details/?pubid=47590

关键词：

U.S. dairy industry;  forecasts;  simultaneous equations;  vector autoregression; 

年份：

2012

出版地：

Washington, DC, USA

总页数：

38 pp

语种：

English

摘要：

This research evaluates the econometric approaches employed by USDA's Economic Research Service (ERS) to contribute to the dairy sector forecasts published in the monthly World Agricultural Supply and Demand Estimates (WASDE) report. To generate the estimates, a quarterly model of the U.S. dairy industry is specified using data for fourth-quarter 1998 (Q4/1998) to first quarter 2009 (Q1/2009), and it is estimated and validated employing data for Q2/2009 to Q1/2010. Different forecasts are generated using a variety of single equation and system methods, and which are then evaluated in terms of forecast precision or predicting turning points in the data. Different approaches, however, more effectively forecast different variables. Vector autoregression with exogenous variables outperforms structural regression models when forecasting prices, but single and system estimations of structural models are superior to time series models when forecasting some items in farm supply and commodity balance sheets.