MSpace - DSpace at UofM >
Faculty of Graduate Studies (Electronic Theses and Dissertations) >
Manitoba Heritage Theses >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1993/7232

Title: Short term hog price forecasting models for the Manitoba hog industry
Authors: Mbabaali, Shakib
Issue Date: 1992
Abstract: The Manitoba hog industry operates under uncertain and changing circumstances. Manitoba Department of Agriculture cites some of the factors responsible for the uncertain and changing environment. Such factors include high costs of new facilities, rising energy costs, variable prices for feed grains and protein supplements, and uncertainties about future hog market prices. The research reported in this study concentrates on the uncertainties about future hog market prices by identifying the factors responsible for hog price fluctuations both on a weekly and monthly basis. The identified factors are used to generate knowledge about future hog market prices by using univariate time series, econometric and compsite models as forecasting toos. The forecasts generated using those models are evaluated against the naive or no change model for their quantitative and qualitative forecasting performance. Evaluation measures used include Mean Squared Error, Mean Absolute Percentage Error and Theil's U1 inequality coefficient for quantitative evaluation. The qualitative evaluation measures include the Naik and Leuthold 4 x 4 contingency table method and the Henriksson-Merton probability-based method. Under certain circumstances the Naik and Leuthold 4 x 4 contingency table method is shown to be inappropriate and a 9 x 9 contingency table is suggested. Overall, the models developed do not perform very well quantitatively but the univariate time series model performs well at predicting turning points. The study demonstrates how producers could benefit from the turning point information generated by the univariate times series model. Keywords: forcasting, time series, econometric model, composite model, naive model, quantitiative forecast evaluation, qualitative forecast evaluation.
URI: http://hdl.handle.net/1993/7232
Other Identifiers: ocm00014907
Appears in Collection(s):FGS - Electronic Theses & Dissertations (Public)
Manitoba Heritage Theses

Files in This Item:

File Description SizeFormat
Mbabaali_Short_term.pdf4.33 MBAdobe PDFView/Open
View Statistics

Items in MSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0! MSpace Software Copyright © 2002-2010  Duraspace - Feedback