Forecasting with economic indicators: a study of predictive retail sales devices based on national and local economic indicators of Fresno, California

The purpose of this study is to construct a predictive device for Fresno County retail sales. The major theme discussed is whether local economic conditions or national economic conditions are a better indicator of Fresno retail sales. The selection of appropriate indicators for the forecasts is based on the results of correlation analysis. A predictive device for Fresno County retail sales will be useful to many local organizations. The national forecasting devices, which are available, seldom relate to local conditions. These national devices are averages of situations which may not reflect a particular market's situation. For this reason, a predictive device for Fresno County retail sales is needed. This device will employ national and local indicators which correlate the retail sales of Fresno County. Most retailers lack the time or ability to develop forecasts, or to adapt local and national indicators to predict local sales. There are time lags between the collection, and analysis of independent variables which delay the computation of a forecast. A model which reduces the number of variables to a few reliable indicators will reduce the data input. An analysis showing which indicators are most relevant for Fresno County will further reduce the number of indicators required for forecasts. A forecasting device based on a small number of independent variables, national or local, will be a viable tool for predicting Fresno County retail sales. A forecasting device using nine components of Fresno County retail sales is suitable for this size of market. The retailer will be able to select a predictive device which corresponds to his type of organization from one or more of the nine components.