Demand Forecasting

Prof. A.N. Sah

What is demand
Demand: In economics, demand refers to the amount
of quantity demanded for a product at a particular price
during a given period of time.
Market Demand: Sum of individual demands. It
refers to total volume that would be bought by a
defined customer group in a defined geographical area
during a given time period.
Market Potential: Market potential refers to the
upper limit of market demand.
What is forecasting?
Forecasting is defined as “Estimating in unknown
situations. In the context of a company, it refers to the
estimates of future sales of the company’s products.
“A successful business manager is a forecaster first;
purchasing, producing, marketing, pricing, and
organizing all will follow it.”
Significance of Forecasting for Business
The area of business forecasting revolves around
demand forecasting, for marketing and production for
shorter horizons.
Marketing department forecasts sales for new product
lines to give strategic information about their sales later
at maturity.
It also forecasts sales for existing product lines to give
feedback on whether the current sales techniques are
working well or not.
Such forecasts are called short range strategic forecasts
which are done usually for 1 to 2 years.
Cont..
Besides this, there is tactical and requirements
forecasting that is done for a very short to short
horizons of 1 week to 12 months with a typical
horizons of 6 to 12 weeks.
Actually, for producing short strategic and tactical
forecasts, we make use of objective forecasting
methods such as smoothing techniques, causal
methods etc.
Cont..
Finally, long range strategic forecasting, usually for a
period comprising from 2 to 10 years is done. Some of
the techniques used under this are surveys, panel of
experts and the Delphi method. It is this strategic
forecasting which tends to make or break companies.
Cont…
In the forecasting process, the identification of the
problem is the most difficult part. It involves
identifying the exact variable of interest, that is, to be
forecasted which requires a deep understanding of the
system. It raises some critical questions such as:
 how the forecasts will be used in the organization,
who requires the forecasts
 and for what purpose, and how the whole forecasting
exercise fits within the organization.
Demand Forecasting Methods
There is a wide variety of forecasting techniques availabe.
They range from relatively informal qualitative methods
to highly advanced quantitative methods of forecasting. It
is important to note that advanced and sophisticated does
not necessarily mean more accurate forecasts.
Sometimes guess estimate by experienced people in the
industry has a very high accuracy. However, forecasters
use wide range of techniques, recognizing that some
techniques might perform better than others depending
upon nature of data.
 
Qualitative Methods
Qualitative forecasting technique is based on
judgements about future. These are often called
judgemental or nonextrapolative techniques.
In this type of forecasting, dependence on numbers are
not only small but they also lack rigourous
specification of the underlying assumption.
Judgemental forecasting is one of the important
techniques of forecasting. In this method, an individual
or small group of people prepare forecasts regarding
likely future conditions.
Cont…
When used by experienced persons keeping in mind
historical trends, current economic situations and other
relevant factors, this method can produce good estimates.
This method, however, tend to work best when
environment changes very rapidly. When economic and
political conditions are very stable, quantitative methods
may yield very good estimates.
However, when economic and political environment are
in flux, quantitative methods may not capture important
clues which have the ability to change the historical
patterns in a country.
 Quantitative Methods
Quantitative methods of forecasting rely on numerical data.
Data is the bread and butter of quantitative methods. There are
two general types of quantitative methods.
First, the time series methods that consists of numerous
techniques that use past trends to predict future conditions.
Second, causal models that use the variables assumed to
influence another variable. In general, quantitative methods
perform better job than qualitative methods in predicting future.
 Finally, time series methods outperform causal models at least
in near term. Time series methods of forecasting are discussed
below:
1. Naïve Model
A naive model simply assumes the value in previous
period as the forecasted value in the current period. In
other words, the value of a variable at time t is the
same as the value at time t-1. This model is also called
random walk model which is specified as follows:
Y(t) =Y(t-1)
Cont..
where Yt is the forecasted value at time t, and Yt-1 is
the actual value at time t-1.
A variant of naive model involves taking average of
two prior values to generate forecast.
Another variation of naive modelling involves
incorporating trend that may be present in data.
Data 1.xlsx
2.Moving Average
Moving average is a very common time series
technique of forecasting. It is very frequently used by
technical analyst in predicting stock prices, interest
rates, foreign exchange rates etc.
In moving average old data points are excluded as new
ones are added. As a consequence, the most recent
information is utilized in obtaining each new moving
average.
Cont…
The optimal time period to include in the computation of
moving average depends on the degree of variation in
the data.
If there is high degree of random component present in
the data then a longer period is used.
Similarly, the presence of cyclicality and seasonality in
the data also requires a longer period to smooth the data.
In practice, trial and error method is used to arrive at the
best fitting model that minimizes forecast error.
Data2.xlsx
3. Exponential Smoothing
Another popular time series forecasting technique is
exponential smoothing.
While moving average technique gives equal weight to
each data points, exponential smoothing assigns
exponentially declining weights as the observation get
older.
In other words, exponential smoothing technique gives
relatively more weight to recent observations and less
weight to older observation. Symbolically, exponential
smoothing is expressed as follows:
Ft+1=α.Xt+ (1-α) Ft
Cont…
where
Ft+1= the forecast for the next time period (t+1)
Ft= the forecast for the present time period (t)
Xt = the actual value for the present time period
α = a value between 0 and 1 referred to as the
exponential smoothing constant.
Cont…
The value of α is determined by the forecaster. If α is
chosen to be less than 0.5, less weight is placed on the
actual value than on the forecast of that value. If α is
chosen to be more than 0.5, more weight is placed on
the actual value than on the forecast of that value
A high smoothing coefficient could be more
appropriate for new products or items for which the
underlying demand is shifting (dynamic or unstable)
Cont…
An α of .7, .8 or .9 might be best for these conditions.
If demand is very stable and believed to be representative
of the future, the forecaster wants to select a low value of
α to smooth out any sudden noise that might have
occurred.
Under these stable conditions, an appropriate smoothing
coefficient might ne .1, .2 or 0.3.
When demand is slightly unstable, smoothing
coefficients of 0.4, 0.5 or 0.6 might provide the most
accurate forecast.
Data3.xlsx
Linear Regression
Regression analysis is a statistical tool for analyzing the nature of
relationship between two variables. It is an important tool
frequently used by economists to understand the relationship
among two or more variables. When our interest lies in
explaining one variable in terms of another variable i.e. y in terms
of x, we use regression.
Regression model studies the relationship between two variables
when one is the dependent variable and the other is an
independent variable. For example, an agriculture economist
might be interested in explaining crop yield (y) with the help of
amount of fertilizers (x) used; change in inflation (y) in terms of
change in money supply (x).
Data4.xlsx
Linear Trend Model
Trend refers to the general tendency of an economic
variable over a long time.
According to Simpson & Kafka, “ Trend also called
secular or long term trend is the basic tendency of
production, sales, income, employment, or the like to
grow or decline over a period of time.
The concept of trend does not include short –range
oscillations but, rather, steady movements over a long
time.” data5.xlsx
Seasonality: Dummy Variable
Seasonal variations in production and sales is a well
known fact in business.
Seasonality refers to regular and repetitive fluctuation
in a time series which occurs periodically over a span
of less than a year.
The main cause of seasonal variations in time series
data is the change in climate. For example, sales of
wollen clothes generally increase in winter season.
Besides this, customs and tradition also affect
economic variables for instance sales of gold increase
during marriage seasons.
Cont…
A variable which takes only two values such as 0 and 1
are called dummy variables. They are also called
categorical, indicator or binary variables in literature.
While 1 indicates the presence of an attribute, o
indicates the absence of an attribute.
data6.xlsx
Thank You