Commodity pork price forecasting for Hormel fresh pork sales team

Abstract

To remain competitive in an ever changing pork industry, Hormel Foods required careful evaluation of advertising forecast accuracy. This study determines forecasting accuracy for bone-in loins, boneless loins, butts, and ribs pricing within Hormel Foods and determines the relationship between forecast horizon (how many weeks forward in pricing) and forecasting accuracy of these products. The challenge required the data collection of the advertising pricing quotes for the sale price in comparison to the forecasted price. Several different forecasting combinations were examined to determine the ideal combination. The focus of this research was to determine which forecast or combination of forecasts was preferable for Hormel Foods. Findings include that each commodity and weeks out front have a different preferred forecast or combination of forecasts when analyzing root mean square errors. Four forecasts (three forecast companies and the United States Department of Agriculture actual markets at the time of forecasts) were observed with one forecast company rarely utilized in the preferred forecasting combinations and therefore the potential exists for a cost savings that affect the bottom line profitability of the division. In addition, economic models presented in this study explain the errors (both raw and percentage based) in relation to the forecast companies, weeks out front forecasted, and specific commodity differences.