Abstract:

The door to forecasting opens more easily for the layman if he studies the foundations on which the question of forecasting is based. It means looking in to the future. The various methods of forecasting differ basically only in the way this search is carried out, i.e. in its degree of its complexity, not, however, which equations or which statistical aids are used. It is important which ideas concerning the connexion between, and the significance of, the forces are used as a basis, and the requirements as to reliability and accuracy, and last but not least, the degree of difficulty with which the calculation is made, or will be made, for instance in regard to the acquisition of data or the time available for making the forecast. The importance of this part of the techniques of forecasting cannot be too highly emphasised. However carefully a model is conceived, the difficulties of the forecaster really only begin when he is faced with that part of the job in which the model is fed with empirical data from the actuality of the business world which have to be made real to life. If one tries to trace the forecast back to its source, using engineering practice , one encounters the expression 'extrapolation'. Because of its character, each forecast is an extrapolation, however well it may be disguised by clever and complicated model constructions.