Abstract:
In this dissertation a study is made of the long-range forecasting problem and the role of forecasting in corporate planning in a South African chemical manufacturing company. Planning is identified in this dissertation as a special kind of decision making, which led to a study of decision making with the objective of identifying the role of forecasting in decision making. The forecasting problem is viewed as comprising two subproblems: the statistical problem and the business problem. Both of these elements are considered necessary for arriving at a forecast. A study of corporate planning leads to the conclusion that a wide diversity of forecasts are needed in any organisation. These include forecasts of technological, social, economic and political developments as well as forecasts of sales volume and prices of existing and planned company products. A survey is made of known forecasting techniques, which are classified into the following three categories: qualitative methods; time series analysis and projection, and; causal methods. It is found that there is a wide range of techniques that could be used in any forecast. The identification of the need for systematic approach to the forecasting problem led to the development of a six-step process for this purpose. A study is made of the methods used for long-range forecasting in African Explosives and Chemical Industries (AE and CI). The approach taken is based on the use of teams whose members are drawn from various parts of the organisation. A forecast of the demands for caustic soda and chlorine and the imbalance between the two is used as an illustration of the problems encountered in forecasting in the Chemical Industry. Finally, it is concluded that the methods used in AE and CI constitute a reasonable approach to the long-range forecasting problem. It is recommended that the use of forecasting teams should be continued and that the suggested six-step procedure should be used for all forecasts.