Abstract:

This paper deals with problems encountered in trying to fit mathematical line projections to the trends of available data and in combining these projections into an objective forecast of the most probable developments which may effect retail trade over the next ten, twenty and thirty years. The paper describes how various problems were overcome and how progress was made towards developing the projections into a form suitable for a computer programme and continuous revision. This paper deals only with the development of a method to produce a long term forecast of the grocery and provision market, with two specific uses: 1. To discover what economic levels will result if all projected trends continue unchanged; and 2. to be modified as soon as there is a change in the current trend of any of the projected component series.