ECONOMETRIC AND ARIMA MODELS IN PREDICTNG CATTLE AND HOG PRICES: AN EVALUATION

Forecasting has been very important in decision making at all levels and sectors of the economy. In agriculture, where the decision environment is characterized by risks and uncertainty largely due to uncertain yields and relatively low price elasticities of demand of most commodities, decision makers require some information about the future and the likelihood of the possible future outcomes. Forecast information serves many users. Farmer's production and marketing decisions, for instance, are often based on some perspective of the likely pattern of price movements over the coming year. Likewise, production prospects for the season are used by market intermediaries in coordinating their resources. Outlook information about production and consumption are also important in developing government price support programs. Along with the variety of forecast users, there exists a wide range of available forecasting techniques. They range from a very simple trend extrapolation to a very complex system of structural equations and simulation models. Very often, these alternative approaches provide different forecasts of the same event. The forecast user, then must determine which of the set of forecasts to use, given his/her decision environment. After a choice has been made, the other forecasts are often discarded. This paper aims to test the hypothesis that forecast performance is proportional to the information contained in the forecasting method. This is carried out through the evaluation of a forecasting approach which is a combination of annual and quarterly ratio econometric models for cattle and hog prices. Such a combination is done to draw from the advantages of both annual models in estimating structural parameters and quarterly models which provide needed information for decision making.