Forecasting
Forecasting is an
important component of Business Management.
It is essentially a
technique of anticipation and provides vital information relating to the
future. It is the basis of all planning activities in an organisation. It
involves collecting valuable information about past and present and estimating
the future. Forecast is an estimate of what is expected to happen in some
future period.
According to
Fayol-the father of modern management— “Forecasting is the essence of
management. The success of a business greatly depends upon the efficient
forecasting and preparing for future events.”
The techniques of forecasting can be grouped under:-
1. Qualitative Techniques
2. Quantitative Techniques
3. Time Series Techniques of Forecasting
4. Causal Modeling
5. Technological Forecasting.
Some of the qualitative techniques of
forecasting are:-
(i) Market Research Techniques (ii) Past
Performance Technique (iii) Internal Forecast (iv) Deductive Method (v) Direct
vs. Indirect Methods (vi) Jury of Executive Opinion (vii) Historical Analogy
(viii) Delphi Technique (ix) Market Survey (x) Judgemental Forecasting (xi)
Sales Force Composite Method (xii) User’s Expectation Method (xiii) Brain
Storming.
Following are the important quantitative
techniques used for the purpose of forecasting:-
(i) Business Barometers Method (ii) Trend
Analysis Method (iii) Extrapolation Method (iv) Regression Analysis Method (v)
Economic Input Output Model Method (vi) Econometric Model (vii) Expectation of
Consumer (viii) Input and Output Analysis.
The factors to be considered for making the choice of
techniques for forecasting are as follows:
(a) The purpose of forecast.
(b) The degree of accuracy desirable.
(c) The time period to be forecast.
(d) Cost and benefit of the forecast to the
company.
(e) The time available for making the
analysis.
(f) Component of the system for which forecast
has to be made.
Basic forecasting techniques may be classified as:
(1) Qualitative and
(2) Quantitative.
1. Qualitative Techniques:
A qualitative forecasting technique relies on
individual or group judgment. When quantitative data are not available, the
use of ‘informed experts’ can be made. Sometimes the opinions of many “experts”
are analysed to predict some future occurrences.
i. Panel of Executive Opinion:
It is also called as a jury-of-expert-opinion
approach. It consists of combining and averaging top management’s views about
the future event. In this approach, generally the executives from different
areas such as sales, production, finance, purchasing are brought together.
Thus, a varied range of management viewpoints can be considered. Forecasts can
be prepared quickly without elaborate data.
ii. Historical Analogy:
This method is most commonly used. It is based
on the belief that future trends will develop in the same direction as past
trends. It assumes that the future will remain as in the recent past. Hence,
past trends are plotted on a graph or chart to show the curve.
Three forms of this method are in use:
(a) Taking the current years’ actual
performance as base for future prediction;
(b) Increasing certain percentages with the
last year’s actual performance to predict the future events; and
(c) Averaging the actual performance of the
previous few years.
iii. Delphi Technique:
This is another judgmental technique. It polls
a panel of experts and gathers their opinions on specific topics. The
forecasting unit decides the experts whose opinions it wants to know. Each
expert does not know who the others are. The experts make their forecasts and
the coordinator summarizes their responses. Here, the experts express their
views independently without knowledge of the responses of other experts.
On the basis of anonymous votes, a pattern of
response to future events can be determined. His technique is used to reduce
the “crowd effect” or “group think” in which everyone agrees with “the experts”
when all are in the same room.
iv. Market Survey:
Another type of qualitative forecast is the
market survey. In this approach, the forecaster can poll, in person or by
questionnaire, customers or clients about expected future behaviour. For
example- people can be asked about their probable future purchases of cars.
This method is effective if the right people are sampled in enough numbers. It
asks a set of “experts”—consumers or potential consumers—what they will do.
(v) Market Research Techniques:
Under this technique, polls and surveys may be
conducted to find out the sale of a product. This may be done by sending
questionnaires to the present and prospective consumers. In addition, this may
also be interviewed personally, though questions and interviews, the manager
can find out whether the consumers are likely to increase or reduce their
consumption of- the product and if so, by what margin. This interviews etc.,
and hence this method is somewhat costly and time consuming.
(vi) Past Performance Technique:
In this technique the forecasts are made on
the basis of past data. This method can be used if the past has been consistent
and the manager expects that the future will resemble the recent past.
(vii) Internal Forecast:
Under this technique indirect data are used
for developing forecasts. For Example—For developing sales forecasts, each area
sales manager may be asked to develop a sales forecast for his area. The area
sales manager who is in charge of many sub-areas may ask his salesmen to
develop a forecast for each sub-area in which they are working. On the basis of
these estimates the total sales forecast for the entire concern may be
developed by the business concern.
(viii) Deductive Method:
In the deductive method, investigation is made
into the causes of the present situation and the relative importance of the
factors that will influence the future volume of this activity. The main
feature of this method is that it is not guided by the end and it relies on the
present situation for probing into the future. This method, when compared to
others, is more dynamic in character.
(ix) Direct vs. Indirect Methods:
In the case of direct method, the different
subordinate units on departments prepare estimates and the company takes the
aggregate of these departmental estimates. This method is also called bottom up
method of forecasting.
On the other hand, in the case of indirect
method of forecasting, first estimates are made for the entire trade or
industry and then the share of the individual units of that industry is
ascertained. This method is also called as “top down” method of forecasting.
(x) Jury of Executive Opinion:
In this method of forecasting, the management
may bring together top executives of different functional areas of the
enterprise such as production, finance, sales, purchasing, personnel, etc.,
supplies them with the necessary information relating to the product for which
the forecast has to be made, gets their views and on this basis arrives at a
figure.
(2) Quantitative Techniques:
Quantitative techniques are known as
statistical techniques. They focus entirely on patterns and on historical data.
In this technique the data of past performance of a product or product line are
used and analysed to establish a trend or rate of change which may show an
increasing or decreasing tendency.
Following are the important quantitative techniques
used for the purpose of forecasting:
(i) Business Barometers Method:
This is also called Index Number Method. Just
as Barometer is used to measure the atmospheric pressure similarly in business
Index numbers are used to measure the state of economy between two or more
periods. When used in conjunction with one another or combined with one or more
index numbers, provide an indication of the direction in which the economy is
heading.
For example—a rise in the amount of investment
may bring an upswing in the economy. It may reflect higher employment and
income opportunity after some period.
Thus, with the help of business activity index
numbers, it becomes easy to forecast the future course of action projecting the
expected change in related activities within a lag of some period. This lag
period though difficult to predict precisely, gives some advance signals for
likely change in future.
The forecasts should bear in mind that such
barometers (index numbers) have their own limitations and precautions should be
taken in their use. These barometers may be used only when general trend may
reject the business of the forecasts. It has been advised that different index
numbers should be prepared for different activities.
(ii) Trend Analysis Method:
This is also known as ‘Time Series Analysis’.
This analysis involves trend, seasonal variations, cyclical variations and
irregular or random variations. This technique is used when data are available
for a long period of time and the trend is clearly visible and stable. It is
based on the assumption that past trend will continue in future. This is
considered valid for short term projection. In this different formulas are used
to fit the trend.
(iii) Extrapolation Method:
Extrapolation method is based Time series,
because it believes that the behaviour of the series in the past will continue
in future also and on this basis future is predicted. This method slightly
differs from trend analysis method. Under it, effects of various components of
the time series are not separated, but are taken in their totality. It assumes
that the effect of these factors is of a constant and stable pattern and would
also continue to be so in future.
(iv) Regression Analysis Method:
In this method two or more inter-related
series are used to disclose the relationship between the two variables. A
number of variables affect a business phenomenon simultaneously in economic and
business situation. This analysis helps in isolating the effects of various
factors to a great extent.
For example- there is a positive relationship
between sales expenditure and sales profit. It is possible here to estimate
sales on the basis of expenditure on sales (independent variable) and also
profits on the basis of projected sales, provided other things remain the same.
(v) Economic Input Output Model Method:
This is also known as “End Use Technique.” The
technique is based on the hypothesis of various sectors of the economy industry
which are inter-related. Such inter-relationship is known as coefficient in
mathematical terms. For example—Cement requirements of a country may be well
predicted on the basis of its rate of usage by various sectors of economy, say
industry, etc. and by adjusting this rate on the basis of how the various
sectors behave in future.
As the data required for this purpose are
easily available this technique is used in forecasting business units.
(vi) Econometric Model:
Econometric refers to the science of economic
measurement. Mathematical models are used in economic model to express
relationship among various economic events simultaneously. To arrive at a
particular econometric model a number of equations are formed with the help of
time series. These equations are not easy to formulate. However, the
availability of computers has made the formulation of these equations
relatively easy. Forecasts can be solved by solving this equation.
2. Time Series Techniques of Forecasting:
These techniques are based on the assumption
that the “past is a good predictor of the future.” These prove useful when lot
of historical data are available and when stable trends axe apparent. These
techniques identify a pattern representing a combination of trend, seasonal,
and cyclical factors based on historical data. These methods try to identify
the “best-fit” line by eliminating the effect of random fluctuations.
This category includes the following:
i. Trend Projection:
This method projects past data into the
future. This can be done in a table or a graph. This method fits a trend line
to a mathematical equation and then projects it into the future by means of
this equation.
ii. Moving Average:
In this method, the average of a limited
number of significant results is calculated and updated as new results become
available by adding the latest result and dropping off the oldest.
iii. Exponential Smoothing:
This technique is similar to the moving
average, except that it gives more weight to recent results and less to earlier
ones. This is usually more accurate than moving average.
