|
Accession Number
|
PB2012-114848
|
|
Title
|
Quarterly Econometric Model for Short-Term Forecasting of the U.S. Dairy Industry.
|
|
Publication Date
|
Jan 2012
|
|
Media Count
|
38p
|
|
Personal Author
|
R. Mosheim
|
|
Abstract
|
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 USDA World Agricultural Supply and Demand Estimates (WASDE) report. To generate the estimates, a quarterly model for 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 then evaluated in terms of forecasting 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 on farm supply and commodity balance sheets.
|
|
Keywords
|
Agricultural economics
Commodities
Dairy industry
Economic forecasting
Economic models
Estimates
Prices
Regression analysis
Supply and demand
Vector autoregression
|
|
|
Source Agency
|
Economic Research Service
|
|
NTIS Subject Category
|
98B - Agricultural Economics
96A - Domestic Commerce, Marketing, & Economics
|
|
Corporate Author
|
Economic Research Service, Washington, DC.
|
|
Document Type
|
Technical report
|
|
Title Note
|
N/A
|
|
NTIS Issue Number
|
1226
|
|
Contract Number
|
N/A
|
