Chapter-Four

Demand Forecasting
4.1 Meaning of Forecasting
Forecasting is the process of estimating the occurrence, timing or magnitude of future
event. Forecasts are statements about future specifying the volume of sale to be
achieved, material demand required, equipments and other inputs needed to meet the
sales. They give operation managers a rational basis for budgeting, capacity planning,
sales, production and inventory, Personnel and material management.
Forecasting is the basis of planning ahead even though the actual demand is quite
uncertain thus, it involves estimation of the future, and of particular interest here is the
expected demand of company’s product. Therefore, forecast of future demand is the
link between company’s internal expectations with outside environment that permits
planning function to commence activities. A popular definition of forecasting is that it is
estimating the future demand of a product, service and the resources necessary to
produce an output.
4.2 Characteristics of forecasts
The following are the characteristics of forecasts:-
 Forecasting techniques generally assumes that the same underling causal
system that assisted in the past will continue to exist in the future.
 Forecasts are rarely perfect; actual results usually differ from predicted values.
 Forecasts for a group of items tends to be more accurate than forecasts for
individual item, because forecasting errors among items in a group usually are
smaller than that of individual items.
 Forecast accuracy decreases as the time period covered by the forecast-time
horizon increases.
4.3. Types of forecasts
There are long-term forecasts as well as short-term forecasts. Operations managers
need long range forecasts to make strategic decisions about the products, processes
and facilities. They also need short-term forecasts to assist them in making decisions
about production issues that span only the next few weeks. Since forecasting forms an
integral part of the planning and decision-making process, production managers must
be clear about the horizon of the forecasts. Additionally, they must also be clear about
the method of forecasting and the units of forecasting. Forecasting is an important

Page 1
planning tool for the organizations.

Long range forecasts

Long range forecasts provide operations managers with the information to make
important decisions such as following-
1.Selecting a product design.
2. Selecting a product processing scheme for a new product
3. Selecting a plan for the long-term supply of scarce materials
4. Selecting a long-term production capacity plan
5. To build new buildings and to purchase new machines
Short term forecasts
Short-term forecasts provide, operations managers with the information to make
important decisions such as following-
1.How much inventory of a particular product should be carried next month?
2. How much of each product should be scheduled for production next week?
3. How much of each raw material should be ordered for delivery next week?

4.4. Benefits of Forecasting

Good forecast of material, labor and other resources for operation are essentially
needed by the managers. If good projection of future demand is available, the
management may take suitable action regarding inventory. Similarly, if production
activities are accurately forecasted, then balanced work-load may be planned. Good
labor relations may be maintained, as there would be lesser hiring and firing activities by
the management with better manpower planning. Therefore, forecasting is useful due to
following benefits:

1. Effective handling of uncertainty

2. Better labor relations

3. Balanced work-load

4. Minimization in the fluctuations of production

Page 2
 5. Better use of production facilities

6. Better material management

7. Better customer service

8. Better utilization of capital and resources

9. Better design of facilities and production system.

The goal of forecasting is not to predict the future but to tell you what you need to know
to take meaningful action in the present.
4.5 Techniques of Forecasting
There are two types of forecasting technique. These are:
A. Qualitative forecasting techniques
B. Quantitative forecasting techniques

A. Qualitative Technique
Qualitative forecasting technique is a technique that is used when there is no historical
data available about past performance. These forecasting techniques are subjective
and judgmental in nature and most of the time they are based on opinion and expertise
judgment.
Qualitative forecasting techniques rely on analysis of subjective inputs obtained from
customers, sales Person, managers and experts.
Forecasts based on judgment, experience or opinions are appropriate when:
Forecasts are prepared quickly in a short period of time,
Available data may be obsolete or up to date information might not be available
because of rapid and continuous changes in the external environment such as
economic and political conditions,
There is no historical data, like demand for a newly introduced product, and
The forecasting period is long range that past events will not repeat themselves in a
similar fashion.
There are four common types of qualitative forecasting techniques. These are:
a. Expert opinion method
b. Sales opinion
Page 3
 c. Consumer surveys
d. Delphi technique
a. Expert Opinion methods
One of the most simple and widely used method of forecasting which consists of
collecting opinions and judgments of individuals who are expected to have the best
knowledge of current activities or future plans. This technique has its own advantages
and disadvantage.
Advantage
 Decision is fast
 Responsibility and accountability is clear
 Brings together the considerable knowledge, experience, skill and talent of
various managers
 Managers (experts) will acquire experience that is obtained in the discussion.
Disadvantage
 Probably poor forecast (due to lack of experience)
 Domination by one or few manger
 Diffusing responsibility for the forecast over the entire group may result in less
pressure to produce a good forecast.
b. Sales force Opinions
In this method, the sales representatives are required to estimate the demand for each
product and the forecast of each sales representative is consolidated to prepare the
overall forecast for the company.
This forecasting technique has also its own advantages and disadvantages
Advantages
 It can reset in quality forecast
 This pools together knowledge
 Can see from different approaches
Disadvantages
 Time taking decision
 Influenced by majority high stares persons
 Avoidance of responsibility
c. Consumer Surveys

Page 4
This forecasting technique is based on the data which is collected from the consumers.
Because it is the consumers who ultimately determine demand, it seems important to
solicit information from them.
Advantage
 Tap information that may not be available else where
 Enhance the quality and accuracy of forecasts
Disadvantage
Experience and knowledge is constructing
Expensive and time consuming

d. Delphi Method
This is a qualitative method of forecasting which involves the development, distribution,
collection and analysis of series of questionnaires to get the views of expertise that are
located at different geographic areas to generate the forecast. A moderator compiles
results and formulates a new questionnaire that is again submitted to the same group
of experts. The goal is to achieve a consensus forecast.
Advantage
The tendency of process loss is avoided/minimized
No influence of the majority
Disadvantage
 It takes time to reach a consensus
 Difficult to coordination and interpretation

B. Quantitative Forecasting Techniques

Quantitative techniques consist of mainly analyzing objective or hard data. This usually
avoids personal biases that sometimes contaminate qualitative methods. It is based on
actual historical statistical data using mathematical and statistical methods to forecast
demand. Thus, it is objective and is also called statistical forecasting.
There are two types of quantitative forecasting techniques:
a. Time Series Analysis
b. Causal Methods
a. Time Series Analysis

Page 5
A time series is a set of some variable (demand) overtime (e.g. hourly, daily, weekly,
quarterly annually). Time series analyses are based on time and do not take specific
account of outside or related factors.
Time series analysis is a time-ordered series of values of some variables. The variables
value in any specific time period is a function of four factors:
a) Trend c) Cycles
b) Seasonality d) Randomness
Trend – is a general pattern of change overtime. It represents a long time secular
movement, characteristic of many economic series.
Seasonality- refers to any regular pattern recurring with in a time period of no more than
one year. These effects are often related to seasons of the year.
Example:
Weather variations – sales of winter and summer
Vacations or holidays – air line travel, greeting card, visitors at tourists and resort
centers.
Theaters demand on weekends
Daily variations: banks may over crowd during the afternoon.
Cycle – are long-term swings about the trend line and are usually associated with a
business cycle (phases of growth and decline in a business cycle).
Randomness – are sporadic effects due to chance and unusual occurrences.

Types of Time Series Analysis

A. Simple Moving average
A simple moving average is obtained by summing and averaging values from a given
number of periods repetitively, each time deleting the oldest value and adding the new
value.

Page 6
 A t 1  A t  2  A t 3  ...  A t  n
SMA = Ft =
n
n

= A
i 1
t i

Where
n
SMA – simple moving average
Ft - Forecast for period t
At-i - Actual demand in period t-i
n - Number of periods (data points) in the moving average

Simple moving average is preferable if the demand for a product is neither growing nor
declining rapidly and also does not have any seasonal characteristics.

Example1:A food processor uses a moving average to forecast next month’s demand.
Past actual demand (in units) is shown in the following table

Month 1 2 3 4 5 6 7 8
Actual demand 105 106 110 110 114 121 130 128

Required
a. Compute a simple 5 month moving average to forecast demand for month 9
b. Find a simple 5 month moving average to forecast the demand for month 10 if the
actual demand for month 9 is 123.

Solution
128  130  121  114  110
a) SMA9 = F9 =
5

= 120.6
Therefore, the forecasted demand for month 9 is 120.6.
123  128  130  121  114
b) SMA10 = F10 = = 116/5 = 123.2
5

Page 7
Therefore, the 5 month moving average forecasted demand for month 10 is 123.2.
Note: In moving average, as each new actual value becomes available, the forecast is
updated by adding the newest value and dropping the oldest value and computing the
average. Consequently the ‘forecast’ moves by reflecting only the most recent values.
B) Weighted Moving average
In weighted moving average, the weight is given in such a way that more weight is given
to the most recent value in the time series. Weights can be percentages or any real
numbers. In weighted moving average, forecasts are calculated by:
Ft = WMA = W1At-1+W2.At-2+… +Wn.At-n
n

= A
i 1
t 1
.W i

Where
Ft =forecast in time t
WMA = weighted moving average
W = weight
A = Actual demand value

Example1: A department store may find that in a four month period the best forecast is
derived by using 40% of the actual demand for the most recent month, 30% two months
ago, 20% of three months ago and 10% of four months ago. The actual demands were
as follows.

Month Month 1 Month 2 Month 3 Month 4

Demand 100 90 105 95
Required:
a. Compute weighted 4-month MA for month 5
WMA = 95x0.4+105x0.3+90x0.2+100x0.10
= 97.5 units
b. Suppose the demand for month 5 actually turned out to be 100. Compute forecast for
month 6.
F6 =WMA = 0.4x100+0.30x95+0.2x105+0.1x90
F6 = 98.5 units.

Page 8
C) Simple Exponential Smoothing
The other type of time series forecasting method is simple exponential smoothing
which weights past data in an exponential manner so that most recent data carry more
weight in the moving average.
With simple exponential smoothing, the forecast is made up of the last period forecast
plus a portion of the difference between the last period actual demand and the last
period forecast.

Mathematically
Ft = Ft-1 + (At-1 - Ft-1)
Where
Ft = Forecast for period t
Ft-1 = Forecast for the previous period
 = Smoothing constant (0<<1)
At-1 = Actual demand for the previous period

The difference between the actual demand and the previous forecast (i.e. At-1 – Ft-1)
represents the forecast error. As we observe from the equation, each forecast is simply
the previous forecast plus some correction for demand in the last period. Thus,
If actual demand is above the last period forecast, the forecast error (correction) will be
positive, and , If the actual demand is below the last period forecast, the correction will
be negative.
The smoothing constant, actually dictates how much corrections will be made. It is a
number between 0 and 1, and it is used to compute the forecast.
Exponential smoothing is the most widely used of all forecasting techniques, because;
Exponential forecasting models provide closer forecasts to actual demand.
Formulating an exponential smoothing model is relatively easy.
The user can easily understand the model
It requires little computation
It requires only three pieces of data
The most recent forecast
The actual demand of the previous period

Page 9
The smoothing constant, 
Example1: The production supervisor at a fiber board plant uses a simple exponential
smoothing technique ( = 0.2) to forecast demand. In April, the forecast was for 20
shipments, and the actual demand was for 20 shipments. The actual demand in May
and June was 25 and 26 shipments. Forecast the value for July.
Solution: First forecast the demand for May and June
F may = F April +  (A April –F April)
= 20+0.2(20-20)
= 20
F June = F May +  (A May –F May)
= 20+0.2(25-20)
= 21
F July = F June + ( A June–F June)
= 21+0.2(26-21)
= 22
Therefore, the forecast for July is 22 shipments.
Trend equation
A linear trend equation has the form
Ft = at + b
Where : Ft = forecast for period t
a = slope of the line
b = value of Ft , at t = 0
t = specified number of time periods from t = 0

The coefficients of the line, a and b can be computed from historical data using these
two equations.

a= n.  ty  t .y
n .t  ( t ) 2
2

b= y  a t
n

Example: Monthly demand for Wonji sugar factory over the past six months for sugar is
given below;

Page 10
 Month (in ‘000 Sept. Oct Nov. Dec. Jan. Feb.
tones)
Actual demand 112 125 120 133 136 140
Required:
Obtain the trend equation?
Forecast the demand for the next two months?

Solution
First let’s find the values of the coefficients a and b.

n .ty  t . y y  a t
a= , b=
n . t  ( t )
2 2
n

T t2 Y ty
1 1 112 112
2 4 125 250
3 9 120 360
4 16 133 532
5 25 136 680
6 36 140 840
=21 91 766 2774

Now let’s compute a, and b

n .ty  t . y 6 x 2774  21 x 766
a= = = 5.314
n . t  ( t ) 6 x 91  ( 21 )
2 2 2

y  a t 766  5 . 31 x 21
b= = = 109
n 6

a) The trend equation

Ft = Y = at + b
= 5.31t + 109
b) Forecast for the next two months (i.e. March and April)
F March = F7 = 5.31(7) + 109

Page 11
 = 146,000 tones
F April = F8 = 5.31 x 8 + 109 = 151,000 tones

b. Casual Forecasting Methods

Casual forecasting techniques rely on identification of related variables that can be
used to predict values of the variable of interest (demand). Casual methods are used
when historical data are available and there is relationship between the factors to be
forecasted.

Example

Page 12
Real estate prices are usually related to, Property location, Square footage, Crop yield
are related to Soil conditions, Amounts and timings of water, Fertilizer application etc…
Types of Casual Methods of Forecasting
a. Regression and Correlation Methods
Regression and correlation techniques are means of describing the association
between two or more variables.
Regression: - It is concerned about the first two issues, i.e. Bringing out the nature of
relationship between any two variables.
Measuring the rate of change in one (the dependent) variable associated with a given
change in the other (independent) variable.
Regression means ‘dependence’ and involves estimating the value of a dependent
variable, Y, from an independent variable X.
Correlation: - is concerned about evaluating the strength of the relationship and
quantifying the closeness of such relationship.
Simple Linear regression and correlation
In simple linear regression, only one independent variable is used and the model takes
the form
Y = a + bx
Where
Y = predicted (dependent) variable, demand
a = value of Y at X = 0
b = slope of the line

Note:
It is convenient to represent the values of the predicted variable on the Y-axis and
values of the predictor variable on the X-axis.
The coefficients a and b of the line are obtained by using the formula
n .xy  x . y
b=
n . x  ( x ) 2
2

y  b x
a= , or y b x
n

n = Number of Period Observations

Page 13
The correlation coefficient r, can be obtained by using the following formula and
coefficient of determination is r2
n .xy  x . y
Coefficient of correlation (r) =
n .x   x   n y   y 
2 2 2
2


Example: The general manager of a building materials production plant feels the
demand for plaster board shipments may be related to the number of constructions
permits issued in the country during the previous quarter. The manager has collected
the data shown in the accompanying table.

Construction Plaster board

Permits shipments
15 6
9 4
40 16
20 6
25 13
25 9
15 10
35 16
Required:
a. Derive a regression forecasting equation?
b. Determine plaster board demand when the number of construction permit is
30
35
40
c. Compute coefficient of determination (r2) and coefficient of correlation (r), and
interpret the numbers

Page 14
Solution
To derive the regression forecasting equation, first let’s find the values of the
Coefficients a and b

X Y XY X2 Y2
15 6 90 225 36
9 4 36 81 16
40 16 640 1600 256
20 6 120 400 36
25 13 325 625 139
25 9 225 625 81
15 10 150 225 100
35 16 560 1225 256
2
x=184 y=80 xy=2146 x =5006 
y2=950
n = 8 pairs of observation
n .xy  x .y
b=
n .x  ( x ) 2
2

8 x 2146184 x 80
=
8 x 5006  (184 )
2

2448
=  0 . 39
6192

y  b x 80  0 . 39 (184 )
a = = = 0.915
n 8

Thus, the regression equation is;

Y = a + bx
 Y = 0.915 + 0.395x

Page 15
B) plaster board demand,
if no of permit = 30
Y = 0.915 + 0.395 (30) = 12.76
= 13 shipments
ii) if not of permit = 35
Y = 0.915 + 0.395 (35)
= 14.74
= 15 shipments
iii) if not of permit = 40
Y = 0.915 + 0.395 (40)
= 16.75
= 17 shipments
Coefficient of correlation and determination
n .xy  x . y
Coefficient of correlation (r) =
n .x   x   n y   y 
2 2 2
2


8 x 2146 184 x 80
r=
 184   80 
2
8 x 5006 2
 x 8 x 950 

2448
r=
7 , 430 , 400

2
r = 0.90  r = 0.81
Interpretation
* r = 0.81 means 81 percent of the total variation in plaster board shipments is explained by
construction permits. What remains is (i.e. 0.19). It indicates that 19% of the total variation,
which remains unexplained, is due to the factors other than the quantity of shipments.

Page 16
Note: Correlation coefficient, r is a number between -1 & 1.
Correlation coefficient can be positive, zero or negative.
r = 1  perfect positive relation.
r = 0 lack of any relationship between the two variables.
r =-1  perfect negative relationship.
Coefficient of correlation, r overstates the degree of relationship. Thus, we use
coefficient of determination, r2.Coefficient of determination, r2 ranges from 0 and 1, and
it is a more objective and definitive measure of the degree of relationship.

Page 17