Techniques of Business Forecasting

The various techniques of forecasting may be classified into two broad categories – (1) quantitative, and (2) qualitative.  Quantitative techniques involve the use of various statistical tools to data for predicting future events.  These include time series analysis, regression analysis, extrapolation and econometric models. Qualitative techniques employ human judgement to predict future. Business barometers, panel consensus, delphi technique, morphological research and relevance tree are the main qualitative techniques. These are used where data are not readily available. A brief description of all these techniques is given below:

1. Time series analysis. In this method a historical series of data is decomposed into various components, viz., trend, seasonal variations, cyclical variations and random variations. After the original data are adjusted for seasonal and cyclical variations, a trend line can be fitted by using the method of least squares. When the various components of a time series are separated, the trend for the variable under study, e.g., demand, can be known. Time series may give misleading results when the future does not reflect the past. Moreover, this method can be used only when data are available for a long period of time.

2. Extrapolation. This method is also based on time series but it does not isolate the effect of various types of changes. It is assumed that there is a constant and stable pattern of movements in the data. On the basis of past behaviour of data, a trend curve is established. For example, if the sales of a company during the last three years (1992, 1993, 1994) have been Rs.’ 10 lakhs, 12 lakhs, and 14 lakhs, it is assumed that in future too the sales will increase by Rs. 2 lakhs per annum. This past trend is projected into the future so that the sales for 1995 are estimated at Rs. 16 lakhs. This method is used to forecast industry growth, national income, population, etc.

3. Regression analysis. This method attempts to find out the relative movements of two or more interrelated series. It is used to estimate the changes in one variable as a result of specified changes in other variables. For instance, if it is known that there is a correlation between advertising expenditure and sales volume, future sales can be estimated on the basis of changes in advertising expenditure. In business several factors often simultaneously influence a particular variable. Regression analysis can be used to isolate the effects of several factors. When there is one dependent variable and several independent variables, regression equations become complex and computer programming is used to estimate the dependent variable.

4. Input-output analysis. Under this method, input requirement or output is estimated on the basis of known relationship between input and output. This technique is based on the well-established interrelationships between different sectors of the economy. For example, the total quantity of petrol required in the country can be estimated on the basis of its usage rates in various sectors say transport, industry, household, etc. Input-output analysis fields sector-wise forecasts. It is widely used because the required data are easily obtained.

5. Econometric models. In this method, mathematical models are used to express in quantitative terms the inter-relationship among different variables. These models take the form of a set of simultaneous equations. A large number of equations may have to be formed to arrive at one econometric model because the variables affecting a business phenomenon are several.. These equations are not easy to formulate. The construction of an econometric model is a technical and expensive job. Electronic data processing is required. Small individual firm can ill afford the models of their own. They have to- rely on aggregate or macro level models developed fey specialized institutes or agencies.

6. Historical analogy. Here forecast is based on some analogous conditions elsewhere m the past. According to Rostow an economy passes through certain stages in its development Therefore, the economic situation of a country can be forecasted by making comparison with- another country which has already gone through that stage. For example, if the Indian economy is at the take-off stage demand for various products may be forecasted on the basis of the conditions prevailing in USA around 1940. However, this method is more useful for estimating qualitative changes, e.g., attitudes of workers, social values, etc.

7. Business barometers. Business barometers are the index numbers used to predict the directions in which the economy is mowing. The assumption here1 is that past patterns tend to repeat themselves in future. Gross national product^ wholesale price index, consumer price index, index of industrial production, volume of money supply stock exchange index are common business barometers used in forecasting. Continuous rise in gross national product, for example; indicates an upswing in the economy. However, indo numbers have their own-limitations and should be used carefully;

8. Panel consensus method. Under this method, a panel of experts in the area is prepared. The opinions these expert are combined and averaged. For example, to forecast the profitability of a new computer model, the opinions of departmental heads-out probable sales and profit-margins may be used. Similarly, the- opinion polls are used predict the outcome of elections, to formulate demand projections, to estimate the life of technology, etc.

9. Delphi technique. In this method, also the minds of experts in the concerned area are probed systematically. But there is no face-to-face contact between them. They are kept apart and their identity is kept secret from one another. Their opinions are solicited through written answers to a carefully prepared questionnaire. The experts who have differed with the majority opinion are fed back the results and are requested to communicate the reasons for their divergent opinions. The process is repeated until a better convergence of opinions emerges. The final results are taken as the forecasts. This method applies a scientific approach but may not give the same answer in all cases.

10. Morphological analysis. This method is used mainly to forecast technological changes. It consists of identifying the relevant dimensions of the object, listing all varieties and combinations of these dimensions and finding practical applications for them. For example, managers have: successfully used this technique to find multiple uses for transistors, lasers and microcircuity.

The choice of a method of forecasting depends upon several factors : (a) time period to be covered, (b) cost of the forecast, (c) time available for forecasting, (d) content of the forecast, (e) desired degree of accuracy, (f) the relevance and availability of historical data.

Techniques of Business Forecasting Techniques of Business Forecasting Techniques of Business Forecasting

Techniques of Business Forecasting Techniques of Business Forecasting Techniques of Business Forecasting

Techniques of Business Forecasting Techniques of Business Forecasting Techniques of Business Forecasting Techniques of Business ForecastingTechniques of Business Forecasting

By Hassham

One thought on “Techniques of Business Forecasting – POM – Short Notes”

Leave a Reply

Your email address will not be published. Required fields are marked *