Professional Documents
Culture Documents
1
Content
Definition
2 use of forecasts 5
4 Features & 8
assumptions
5 Selection of 9
forecasting techniques
6 Techniques of 11
forecasting
7 Limitations of 21
forecasting
8 bibliography 23
Environmental forecasting
2
Introduction:
Definition:
Today, changes are rapid and too frequent and in a way quite
necessary for overall growth of economy. There have been quantum
changes from 1970 onwards and today in the business world
anything that is consistent is only change. In the times to come when
changes would predatory, it would be crucial for managers to
invent new ways of surviving in the ever changing business
environment. They would have to build up build up the capacity of
3
the firm to face the onslaught of changes by being more agile and
flexible for adapting themselves to changes. They would have to find
out new ways of creating opportunities of profitability and growth.
New rules of business will have to be written to meet over-
increasing expectations of drivers of business. To be prepared for
such on going eventualities, managers will have to prepare
themselves for really understanding the remote and the immediate
environments of business and mechanisms of changes that affect
their industry.
The changes have not only affected smaller companies but also the
giants of various industries. In fact the organizational models those
are available with us today for giant companies prove as their
handicap in their process of adaptation due to their large inertia
and consequent slower response towards changes in environment.
Hence there is an urgent need for companies shed-off extra inertia
and develop agility since these large corporations will also be
involved in the transition process. For example: the situation that
the automobile industry is facing today is due to incorrect business
environment forecasting for which many giants are paying a price
through under-utilized manufacturing capacity, piling up of
inventories and the locked up capital and operating cash.
4
Use of Forecasts
Plans are often confused with forecasts. Plans are sets of actions to
help deal with the future. Forecasting (or predicting) is concerned
with determining what the future will be. A plan is an input to the
forecasting model. If the forecasts arc undesirable, then one might
change the plan, which, in turn, could change the forecast. The point
to remember is that good plans depend on good forecasts. In practice,
forecasts are sometimes used to motivate people. More properly,
people should be motivated by plans (e.g., “meet this plan. And we will
pay you a bonus of 25 percent”). Decisions have often been made,
before any formal forecasting has been done. In such cases, the
forecast serves little purpose other than to annoy people if it conflicts
with their decision or to please them if it supports their decision. For
the forecast to be used effectively, it should be prepared before
decisions arc made. Not only the expected outcome, but also other
likely outcomes (such as the best and worst outcomes) should be
forecast. If the worst outcome poses too much risk, forecasts should be
made for alternative interventions.
5
Which Area and Why Use Forecasts:
6
firm may be met at the lowest possible cost. Forecasts
also must be made for interest rates, to support the
acquisition of new capital, the collection of accounts
receivable to help in planning working capital needs, and
capital equipment expenditure rates to help balance the
flow of funds in the organization.
7
needs of the community. In order to do this efficiently, a
projection has to be made of: growth in absolute size of
population, changes in the number of people in various
age groupings, and varying medical needs these different
age groups will have.
8
As pointed out, forecasting techniques are quite different from each
other. But four features and assumptions underlie the business of
forecasting. They are:
9
The choice of a forecasting technique is influenced significantly by the
stage of the product life cycle and sometimes by the firm or industry
for which a decision is being made. In the beginning of the product life
cycle, relatively small expenditures are made for research and market
investigation.
After evaluating the particular stages of the product and firm and
industry life cycles, a further probe is necessary. Instead of selecting a
forecasting technique by using whatever seems applicable, decision
makers should determine what is appropriate.
10
What technique or techniques to select depends on six criteria:
11
Techniques of forecasting
Econometric Models:
12
to the dependent variable, one must forecast changes in the causal
variables in order to calculate a forecast for the variable of interest.
(4) Causal factors arc easier to forecast than the variable of interest.
These conditions are often encountered in forecasting
13
Expert Systems:
Extrapolation:
14
forecast will be achieved irrespective of who makes the forecast.
Extrapolation can also be relatively simple and inexpensive. The
opportunities for the introduction of biases in extrapolation are
limited. Perhaps the major potential source of bias is that
extrapolative forecasts can differ substantially depending on the time
period examined. This bias can be reduced by selecting long time
series and by comparing forecasts when different starting and ending
points are used. Another source of bias associated with extrapolative
forecasts involves the selection of the extrapolation method. To combat
this bias, one should use simple, easily understood methods and
preferably more than one method.
Judgmental Methods:
Judgmental forecasting involves methods that process information by
experts, rather than by quantitative methods. The experts might have
access to data, and their approach might be structured, but the final
forecasts arc the result of some process that goes on in their heads.
Before discussing tools that aid judgmental forecasting, it is important
to mention one tool that is widely used and well accepted, but which
typically harms accuracy and leads to an unwarranted gain in
confidence. The culprit is the traditional (unstructured) group meeting.
Besides the biases inherent in unstructured meetings (such as the
influence of the boss), the group’s information is likely to be poorly
used. Judgmental forecasts are susceptible to various biases. To
15
reduce biases, one should select unbiased experts (i.e., those who have
nothing to gain from a forecast that is either too high or too low). In
addition, care should be given to how the forecasting problem is
formulated. Questions should be structured to use the judges’
knowledge most effectively, pre-tested to ensure that the experts
understand them, and worded in different ways to see if that affects the
forecasts. Such procedures are particularly important when
forecasting sensitive issues, such as the effects of global warming. The
use of structured procedures can greatly improve the accuracy of
judgmental forecasts. Structure is easy to apply and involves only
modest costs. I discuss four structured judgmental procedures that
should be of interest for environmental forecasting: (1) role playing,
which uses subjects to act out relevant interactions to determine what
they would do when affected by an intervention; (2) intention surveys,
which use statements by key participants in the system about what they
expect to do given certain trends or interventions; (3) Delphi. Which
uses expert judgment to forecast trends or the effects of intervention:
and (4) analogies, where experts try to generalize from similar
situations. Brief attention is given to conjoint analysis and to
judgmental bootstrapping.
Role Playing
16
interactions. When the interactions of conflicting groups are important
to the outcome, role playing provides a way to simulate this
interaction. If new and important interventions would lead to
behaviors that are dependent upon the interactions among decision
makers, then role playing is likely to be more relevant than intentions.
With intentions, decision makers would have to predict what they
would do initially, how they would modify their decisions in reaction
to the decisions made by others, how others would respond to this
reaction, and so on. This chain of events is often too complex for the
respondent, so it makes sense to act it out.
- Select subjects who can act the role (interestingly, the selection of
subjects does not seem to be a critical aspect for the accuracy of role
playing)
- Subjects should act as they would act if they were actually in such a
role
17
- Subjects should improvise as needed.
Intention Surveys:
Intention studies are surveys of individuals about what actions they
plan to lake in a given situation or, if lacking a plan, what they expect
to do. Such surveys are useful for predicting the outcomes of
interventions. When a situation depends on the decisions of many
people (such as with the trash collection for a community), surveys arc
much more expensive than Delphi. However, they provide the
perspective of those who will actually be making decisions. In
addition, one could have presented this situation to consumers and
asked them how they would respond.
18
Tools for surveys have been improving since the 1936. When
interventions would create large changes and where the behavior of
decision makers is dependent upon decisions by others, respondents
may find it difficult to predict how they would behave. Surveys are of
less value in such cases. Given all the ways that intentions or
expectations may be wrong, it should not be surprising to find that
sampling error alone provides a poor way to estimate prediction
intervals.
Delphi:
19
the data are poor, or experts disagree with one another. As a result,
Delphi is applicable when new interventions arc proposed or where a
trend has recently undergone a shock. Nevertheless, judgments tend to
be too conservative in the face of rapid change. In particular,
judgment underestimates exponential growth and exponential growth
is common in environmental problems.
Analogies:
To forecast the outcome of interventions, it is common for experts to
search for cases where similar Interventions have been conducted at
different times or in different geographic areas and then to generalize
from them.
Conjoint Analysis:
Conjoint analysts can be used to predict what strategy would be
accepted. For example, one could propose different possible plans that
would have various effects. The effects could be varied according to an
20
experimental design. Once a model is developed, predictions can be
made for changes in the design.
Judgmental Bootstrapping:
Business barometers:
21
It involves decomposition of historical series into various components,
viz., trend, seasonal variations, cyclical variations and random
variations. Time series analysis uses index numbers but it is different
from barometric technique, the future is predicted from the indicating
series which serve as barometers of economic change. In time series
analysis, the future is taken as some sort of an extension of the past.
However, time series analysis should be used as a basis for forecasting
when data are available for a long period of time and tendencies
disclosed by trend and seasonal factors are fairly clear and stable.
Regression analysis:
Survey method:
22
Field surveys can be conducted to gather information on the intensions
of the concerned people; for example information can be collected
through surveys about the likely expenditures of consumers on various
items. Both quantitative and qualitative information can be collected.
Such information may throw useful light on the attitudes of the
consumers in regard to various items of expenditure and consumption.
On the basis of such survey, demand for various goods can be
projected. To limit the cost and time, the survey may be restricted to a
sample from the prospective consumers. Survey method is suitable for
forecasting demand both of existing and new products.
Limitations of forecasting
1. based on assumptions:
23
Forecasting is based on certain assumptions. It merely suggests
that if an event has happened this way in the past, it will happen
that way in the future. The basic assumption behind this is that
events do not change haphazardly and speedily but change on a
regular pattern. This assumption may not hold good. In fact there
are various factors which go into determining the occurrence of an
event. The behaviour of all these factors may not be similar. A
change in a particular factor may be so unpredictable and
important that it may affect the total business situation.
Forecasts are not always true; they merely indicate the trend of
future happenings. This is so because the factors which are taken
into account for making forecasts are affected by human factor
which is highly unpredictable. Various techniques of forecasting
suggest the relationship among various known facts. They can
project the future trends but cannot guarantee that this would
happen in future. More is the period of forecasting, higher is the
degree of forecasting, higher is the degree of error. Therefore it has
been commented that “the only thing you can be sure about any
forecast is that it will contain some error.”
24
time and cost factor suggests the degree to which an organization
will go for formal forecasting. For making forecast of any event,
certain information and data are required. Some of these may be in
highly disorganized form; some may be in qualitative form. The
collection of information and conversion of qualitative data into
quantitative ones involves lot of time and money. Therefore
managers have to trade off between the cost involved in forecasting
and resultant benefits. This is the reason why most of the smaller
organizations do not go for formal system of forecasting.
Bibliography:
25
3. Lomash, Sukal and Mishra, P.K. Business Policy and Strategic
Management. New Delhi. Vikas Publishing House, 2007.
4. Internet
26