Doctoral Dissertations

Joe T. Davis

Date of Award

8-1974

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Agricultural Economics

Major Professor

Raymond Daniel

Committee Members

Irving Dubov, Emmit Rawls, Kieth Phillips

Abstract

The overall objective of this study was to develop a price forecasting model which would give the producers of feeder cattle, feedlot operators, and other segments of the beef cattle industry more than just a hunch as to future feeder cattle price movements. The specific objectives were: (1) develop an econometric model to identify the major factors influencing the quarterly demand and supply of feeder cattle in the United States; (2) develop alternative quarterly feeder cattle price forecasting models using the econometric structural relationships estimated above; and (3) evaluate the interrelationships among the various markets in the beef cattle industry.

An econometric model consisting of eight behavioral equations and two market clearing equations were developed to describe the relationships within and among the feeder cattle, slaughter cattle, and retail sectors of the beef industry. The behavioral equations were fitted to quarterly data for the years 1960-1972 using the two-stage least squares technique.

The farm level demand for feeder cattle was normalized on the current price of feeder cattle. The major factors hypothesized to affect the price of feeder cattle were the current quantity of feeder cattle, the price of com, the number of head on feed, the current price of slaughter cattle, the short-term interest rate, and quarters of the year.

The farm level supply function was normalized on the current quantity of feeder cattle. The major factors hypothesized to affect the quantity of feeder cattle supplied were the current price of feeder cattle, calf crop lagged two quarters, the price of feeder cattle lagged four quarters, a time variable, and quarters of the year.

The demand relationship for slaughter cattle was normalized on the current price of slaughter cattle and the supply relationship was normalized on the current quantity of slaughter cattle. The major factors hypothesized to affect the price of slaughter cattle were the quantity of slaughter cattle, the retail price of beef, cow slaughter, cold storage holdings of beef, wage rate in the meatpacking industry, and quarters of the year. The major factors hypothesized to affect the quantity of slaughter cattle supplied were the current price of slaughter cattle, price of feeder cattle lagged two quarters, the price of corn lagged two quarters, a time variable, and quarters of the year.

A marketing margin was used to connect the prices at the farm level to the prices at the retail level. The factors affecting the farm to retail marketing margin for beef were hypothesized to be the quantity of slaughter cattle moving through the market, the wage rate in the meatpacking industry, the price of slaughter cattle, and time.

Retail level demand equations for beef, pork, and chicken were developed. The major factors affecting the demand for these three substitute meats were their respective prices and quantities, income, and quarters of the year.

The results indicated that the price and quantity of feeder cattle were simultaneously determined. The major factors affecting the price of feeder cattle were the quantity of feeder cattle and the price of slaughter cattle. The demand relationship was found to be significantly higher in the fall quarter. The major factors affecting the quantity of feeder cattle supplied were the price of feeder cattle and the time variable.

The results indicated that the major factors affecting the price of slaughter cattle were the retail price of beef and cow slaughter while the major factors affecting supply were the price of slaughter cattle, the price of feeder cattle lagged two quarters, and time.

Alternative forecasting models were developed to predict the price and quantity of feeder cattle. The most promising model that could be used to predict feeder cattle prices and quantities was a model which included all independent variables in the first stage of the TSLS technique. However, data for variables measured in time period "t" would not be available at the time the prediction is needed. Therefore, a model using all independent variables in the first stage with all variables measured in time period "t" lagged two quarters was used to predict the price and quantity of feeder cattle for the five quarters following the sample period. The predictions were evaluated on the basis of the direction of change and how closely the predicted values approximate the actual value. The model correctly predicted two out of five directions for price and three out of five direction of change for quantity. The largest deviation between the actual price and the predicted price of feeder cattle using this model was $12.05 which occurred in the summer quarter of 1973.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS