An analysis of monthly wheat, flour, and bread prices in a structural and time series framework

Abstract

Wheat, flour, and bread prices fluctuate at all levels of the market. Accurate forecasts of these prices are valuable to buyers and sellers that trade in the cash and futures markets. Rational distributed lag models of monthly prices from June 1977 to May 1984 for Kansas City No. 1 Hard Red Winter Wheat, Minneapolis Dark Northern Spring Wheat, Portland No. 1 Soft White Wheat, Kansas City flour, and retail bread prices are made to evaluate the economic or structural factors influencing price. Multivariate autoregressive-integrated-moving average error (ARIMA) models are also used to compare with the structural models price forecasting ability. Rational lags are estimated using a nonlinear least squares algorithm, incorporating the specification of nonstochastic difference equations so that the disturbance process is divorced from the systematic portion of the difference equations. Certain economic factors are found to be significant in Influencing the prices of wheat, flour, and bread. Partial derivatives and price flexibilities are calculated to estimate the short, intermediate, and long-run adjustments of prices in the structural models. In the structural models total wheat stocks are the most Influential variable in determining wheat prices and the price of wheat was most influential in the flour price equation. Flour price is highly significant in influencing retail bread price, with the secular effects of income increasing over time. The price forecasting abilities of the structural and ARIMA are found to be relatively close when comparing the Root Mean Square Errors and the adjusted coefficients of determination.