Actions for AN ECONOMETRIC MODEL FOR PREDICTING MONTHLY PRICE AND PRODUCTION IN THE U. S. EGG SUBSECTOR

Email RIS file
Bookmark
Report an Issue

AN ECONOMETRIC MODEL FOR PREDICTING MONTHLY PRICE AND PRODUCTION IN THE U. S. EGG SUBSECTOR

Author
SLANE, THOMAS CHARLES, JR.
Physical Description
92 pages
Additional Creators
Pennsylvania State University
Access Online

Availability

I Want It
I Want It

Finding items...

Summary
This study sought to formalize the underlying structural relationships of the U.S. egg subsector and to develop and validate a quantitative procedure for generating forecasts of future market conditions of this system. The results of the study could provide decision makers information on which to base better informed choices between alternative courses of action.
The structural model of the egg subsector was specified on three bases: economic theory with the concepts of demand, supply and prices; detailed study of data, institutions and technology pertaining to the subsector; and similar models developed in previous studies. The model specification was a recursive system of linear relationships (Model I). A model adapted with shifter variables to account for observed structural changes was specified to improve predictive accuracy (Model II).
The structural parameters of the models were estimated by use of ordinary least squares and Cochrane-Orcutt iterative autoregression least squares techniques with a time series of 120 monthly observations (1969-1978). Using the structural estimates of Model I and Model II, forecasts of the endogenous variables were made for the twelve months of 1979 based on the realized values of the exogenous variables. Validation of the two models was undertaken using the mean absolute deviation of the ex-post predictions as the predictive performance criterion. Based on the predictive performance criterion, it was found that Model II was the more valid structural specification for predicting market conditions of the egg subsector.
In dealing with time series data, estimates of structural coefficients often tend to be related to the time period considered in their estimation. Comparison of the mean absolute deviation of the ex-post predictions in the first six months of 1979 and 1980, based on the provisional structural estimates of Model II, revealed the need to update the model as time passes and new data are obtained. The updating process, for this purpose, was accomplished by introducing data for the most recent full year (1979) and eliminating the data for the first year of the provisional data base (1969).
The structural estimates of Model II were then reestimated and, based on the reestimated structural parameters, forecasts of the endogenous variables of the subsector for the first six months of 1980 were made. The results suggest that the updating procedure appears to be an efficient way of handling structural change since it allows the structure of a model to evolve over time while maintaining the same model specification.
The improvement in the predictive power of the updated model, in the presence of structural change, suggests that the updated structural specification of Model II should be considered a more viable representation of the current structure of the egg subsector. Overall, the updated Model II improved forecasts to the extent that producers and planners may find them useful in decision making or policy considerations.
Other Subject(s)
Dissertation Note
Ph.D. The Pennsylvania State University 1981.
Note
Source: Dissertation Abstracts International, Volume: 42-04, Section: A, page: 1731.
Part Of
Dissertation Abstracts International
42-04A
View MARC record | catkey: 13611324