Chapter 3

Forecasting

McGraw-Hill/Irwin
Copyright © 2012 by The McGraw-Hill Companies, Inc. All rights reserved.
Chapter 3: Learning Objectives
You should be able to:
1. List the elements of a good forecast
2. Outline the steps in the forecasting process
3. Describe at least three qualitative forecasting techniques and the
advantages and disadvantages of each
4. Compare and contrast qualitative and quantitative approaches to
forecasting
5. Describe averaging techniques, trend and seasonal techniques,
and regression analysis, and solve typical problems
6. Explain three measures of forecast accuracy
7. Compare two ways of evaluating and controlling forecasts
8. Assess the major factors and trade-offs to consider when
choosing a forecasting technique

#profevad2015

2
Forecast
Forecast – a statement about the future value of a
variable of interest
We make forecasts about such things as weather,
demand, and resource availability
Forecasts are an important element in making informed
decisions

#profevad2015
3
Forecasts affect decisions and activities throughout an organization

Accounting Cost/profit estimates

Finance Cash flow and funding

Human Resources Hiring/recruiting/training

Marketing Pricing, promotion, strategy

MIS IT/IS systems, services

Operations Schedules, MRP, workloads

Product/service design New products and services

Two Important Aspects of Forecasts
Expected level of demand
The level of demand may be a function of some
structural variation such as trend or seasonal variation
Accuracy
Related to the potential size of forecast error

#profevad2015
5
Features Common to All Forecasts
1. Techniques assume some underlying causal system that
existed in the past will persist into the future
2. Forecasts are not perfect
3. Forecasts for groups of items are more accurate than
those for individual items
4. Forecast accuracy decreases as the forecasting horizon
increases

#profevad2015
6
Elements of a Good Forecast
The forecast
 should be timely
 should be accurate
 should be reliable
 should be expressed in meaningful units
 should be in writing
 technique should be simple to understand and use
 should be cost effective

#profevad2015
7
 Steps in the Forecasting Process
1. Determine the purpose of the forecast
2. Establish a time horizon
3. Obtain, clean, and analyze appropriate data
4. Select a forecasting technique
5. Make the forecast
6. Monitor the forecast

#profevad2015
8
Forecasting Approaches
 Qualitative Forecasting
 Qualitative techniques permit the inclusion of soft information such as:
 Human factors
 Personal opinions
 Hunches

 These factors are difficult, or impossible, to quantify

 Quantitative Forecasting
 Quantitative techniques involve either the projection of historical data or
the development of associative methods that attempt to use causal
variables to make a forecast
 These techniques rely on hard data

#profevad2015 9
Judgmental Forecasts
Forecasts that use subjective inputs such as opinions
from consumer surveys, sales staff, managers,
executives, and experts
Executive opinions
Salesforce opinions
Consumer surveys
Delphi method

#profevad2015 10
Time-Series Forecasts
Forecasts that project patterns identified in recent
time-series observations
Time-series - a time-ordered sequence of observations
taken at regular time intervals
Assume that future values of the time-series can be
estimated from past values of the time-series

#profevad2015 3-11
11
 Time-Series Behaviors
Trend
Seasonality
Cycles
Irregular variations
Random variation

#profevad2015 3-12
12
Trends and Seasonality
Trend
 A long-term upward or downward movement in data
 Population shifts
 Changing income
Seasonality
 Short-term, fairly regular variations related to the calendar or time
of day
 Restaurants, service call centers, and theaters all experience
seasonal demand

#profevad2015 13
 Cycles and Variations
Cycle
 Wavelike variations lasting more than one year
 These are often related to a variety of economic, political, or even
agricultural conditions
Random Variation
 Residual variation that remains after all other behaviors have been
accounted for
Irregular variation
 Due to unusual circumstances that do not reflect typical behavior
 Labor strike
 Weather event

#profevad2015 14
 Time-Series Behaviors

#profevad2015
15
 Time-Series Forecasting - Naïve Forecast
Naïve Forecast
Uses a single previous value of a time series as the basis
for a forecast
 The forecast for a time period is equal to the previous

time period’s value

Can be used with
 a stable time series
 seasonal variations
 trend

#profevad2015
16
 Naïve Forecasts
Forecast for any period = previous period’s actual
value

Ft = At-1
F: forecast A: Actual t: time period

#profevad2015
Naïve Forecast Example

Week Sales (actual) Sales (forecast) Error

t A F A-F
1 20 -
2 25 20 5
3 15 25 -10
4 30 15 15
5 27 30 -3

#profevad2015
 Naïve Forecasts
Simple to use
Virtually no cost
Quick and easy to prepare
Data analysis is nonexistent
Easily understandable
Cannot provide high accuracy
Can be a standard for accuracy

#profevad2015
Uses for Naïve Forecasts

#profevad2015
Time-Series Forecasting - Averaging

These Techniques work best when a series tends to

vary about an average
Averaging techniques smooth variations in the data
They can handle step changes or gradual changes in the
level of a series
Techniques
1. Moving average
2. Weighted moving average
3. Exponential smoothing

#profevad2015
21
Moving Average
Technique that averages a number of the most recent
actual values in generating a forecast
n

A t i
Ft  MA n  i 1
n
where
Ft  Forecast for time period t
MA n  n period moving average
At 1  Actual value in period t  1
n  Number of periods in the moving average
#profevad2015
22
Moving Average
As new data become available, the forecast is
updated by adding the newest value and dropping the
oldest and then re-computing the average
The number of data points included in the average
determines the model’s sensitivity
Fewer data points used-- more responsive
More data points used-- less responsive

#profevad2015
23
Moving Average Example

Week Sales (actual) Sales (forecast) Error

t A F = MA3 A-F
20
1 -  

2 25 -  

3 15 -  

4 30 20 10

5 27 23.3333 3.66667

6   24  

#profevad2015
24
Simple Moving Average
Figure 3-4 Revised Forecast Forecast
Actual (MA3) (MA5)
47
45
43
41
39
37
35
1 2 3 4 5 6 7 8 9 10 11 12

Questions:
• Why is MA3 longer than MA5?
• Which curve fluctuate the most?
• Which curve is the smoothest?

#profevad2015
Simple Moving Average
Responsiveness vs. Stability
• Smaller m, responsiveness , stability 
• Larger m, responsiveness , stability 
• Must maintain stability when fluctuations are high.
Figure 3-4 Revised Forecast Forecast
Actual (MA3) (MA5)
47
45
43
41
39
37
35
1 2 3 4 5 6 7 8 9 10 11 12

#profevad2015
Weighted Moving Average
The most recent values in a time series are given more
weight in computing a forecast
The choice of weights, w, is somewhat arbitrary and
involves some trial and error

Ft  wt ( At )  wt 1 ( At 1 )  ...  wt  n ( At  n )
where
wt  weight for period t , wt 1  weight for period t  1, etc.
At  the actual value for period t , At 1  the actual value for period t  1, etc.

#profevad2015
27
Weighted Moving Average Example
Week Sales (actual) Sales (forecast) Error
t A F = MA3 A-F
1 20 -
2 25 -
3 15 -
4 30 19 11
5 27 24.5 2.5
6 25.5

#profevad2015
Exponential Smoothing
A weighted averaging method that is based on the
previous forecast plus a percentage of the forecast
Ft  Ft 1   ( At 1  Ft 1 )
error
where
Ft  Forecast for period t
Ft 1  Forecast for the previous period
 = Smoothing constant
At 1  Actual demand or sales from the previous period

3-29
#profevad2015 29
Exponential Smoothing

Ft = Ft-1 + (At-1 - Ft-1)

Weighted averaging method based on previous forecast
plus a percentage of the forecast error
A-F is the error term,  is the % feedback

#profevad2015
 Example 3 - Exponential Smoothing

Period Actual Alpha = 0.1 Error Alpha = 0.4 Error

1 42
2 40 42 -2.00 42 -2
3 43 41.8 1.20 41.2 1.8
4 40 41.92 -1.92 41.92 -1.92
5 41 41.73 -0.73 41.15 -0.15
6 39 41.66 -2.66 41.09 -2.09
7 46 41.39 4.61 40.25 5.75
8 44 41.85 2.15 42.55 1.45
9 45 42.07 2.93 43.13 1.87
10 38 42.36 -4.36 43.88 -5.88
11 40 41.92 -1.92 41.53 -1.53
12 41.73 40.92

#profevad2015
Picking a Smoothing Constant

Actu
50
al
.4
45  .1
Demand

40

35
1 2 3 4 5 6 7 8 9 10 11 12
Period

#profevad2015
Other Forecasting Methods - Focus
Focus Forecasting
Some companies use forecasts based on a “best current
performance” basis
 Apply several forecasting methods to the last several periods
of historical data
 The method with the highest accuracy is used to make the

forecast for the following period

 This process is repeated each month

#profevad2015
33
Other Forecasting Methods - Diffusion
Diffusion Models
Historical data on which to base a forecast are not
available for new products
 Predictions are based on rates of product adoption and usage
spread from other established products
 Take into account facts such as
 Market potential

 Attention from mass media

 Word-of-mouth

#profevad2015
34
 Techniques for Trend
Linear trend equation
Non-linear trends

#profevad2015
35
Linear Trend
A simple data plot can reveal the existence and nature
of a trend
Linear trend equation

Ft  a  bt
where
Ft  Forecast for period t
a  Value of Ft at t  0
b  Slope of the line
t  Specified number of time periods from t  0

#profevad2015
36
 Linear Trend Equation

Ft

Ft = a + bt

0 1 2 3 4 5
Ft = Forecast for period t
t
t = Specified number of time periods
a = Value of Ft at t = 0
b = Slope of the line

#profevad2015
Linear Trend Equation
180
160
140
120
100 Actual
80 Trend
60
Underlying
Demand
a 40
20 180
160
0 140
120
1 2 3 4 5 6 7 8 9 10 11 12 13 10014 15 Actual
80 Trend
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 b (slope)

#profevad2015
Linear Trend Equation
Easy to use: a built-in function
in spreadsheet software
Based on Least Squares method Yt Deviation
used in linear regression, which 180
160

Minimizes the sum of the 140

120
100 Actual

squares of the deviations 80

60
Trend

40

Uses equal weight for all time 20

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

periods
Both a and b must be At
recalculated on regular basis to
include new data.

#profevad2015
 Estimating slope and intercept
Slope and intercept can be estimated from historical
data
n  ty   t  y
b
n t   t
2
2

a
 y  b t
or y  bt
n
where
n  Number of periods
y  Value of the time series


#profevad2015 3-40
40
Linear Trend Equation Example
Period (t) Sales (Yt) t*Yt t^2
1 67 67 1 Exercise: Calculate b and a by
2 74 148 4 using the sums on the left.
3 102 306 9
4 87 348 16
n  tYt    t  Yt
5 106 530 25
b
n t    t 
6 86 516 36 2 2
7 117 819 49
8 113 904 64
10 5968  55 998
9 130 1170 81
  5.81
10 385   55
10 116 1160 100 2
55 998 5968 385

2 a
 Y  b t
t

t  Yt  tYt t n
998  5.81 55
  67.84
10
#profevad2015
Linear Trend Equation Example
t y
2
Week t Sales ty
1 1 150 150
2 4 157 314
3 9 162 486
4 16 166 664
5 25 177 885

2
 t = 15 t = 55  y = 812  ty = 2499
2
(t) = 225

#profevad2015
Linear Trend Calculation
5 (2499) - 15(812) 12495 -12180
b = = = 6.3
5(55) - 225 275 - 225

812 - 6.3(15)
a = = 143.5
5

y = 143.5 + 6.3t
#profevad2015
 Techniques for Seasonality
Seasonality – regularly repeating movements in series
values that can be tied to recurring events
 Expressed in terms of the amount that actual values deviate
from the average value of a series
 Models of seasonality
 Additive
 Seasonality is expressed as a quantity that gets added to or
subtracted from the time-series average in order to incorporate
seasonality
 Multiplicative
 Seasonality is expressed as a percentage of the average (or trend)
amount which is then used to multiply the value of a series in order
to incorporate seasonality

#profevad2015
44
Models of Seasonality

#profevad2015
45
Seasonality
Volume

120
100
80
60
40
20
0

Volume

#profevad2015
Deseasonalized Data
Deseasonalize

90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00

Deseasonalize

#profevad2015
Seasonal Relatives
Seasonal relatives
 The seasonal percentage used in the multiplicative seasonally
adjusted forecasting model
Using seasonal relatives
 To deseasonalize data
 Done in order to get a clearer picture of the nonseasonal (e.g.,
trend) components of the data series
 Divide each data point by its seasonal relative

 To incorporate seasonality in a forecast

1. Obtain trend estimates for desired periods using a trend
equation
2. Add seasonality by multiplying these trend estimates by the
corresponding seasonal relative

#profevad2015 48

Techniques for Cycles
Cycles are similar to seasonal variations but are of longer
duration
Explanatory approach
 Search for another variable that relates to, and leads, the variable
of interest
 Housing starts precede demand for products and services
directly related to construction of new homes
 If a high correlation can be established with a leading variable,
an equation can be developed that describes the relationship,
enabling forecasts to be made

#profevad2015
49
 Associative Forecasting Techniques
Associative techniques are based on the
development of an equation that summarizes
the effects of predictor variables
Predictor variables - variables that can be used to
predict values of the variable of interest
 Home values may be related to such factors as home and
property size, location, number of bedrooms, and number of
bathrooms

#profevad2015
50
 Simple Linear Regression
Regression - a technique for fitting a line to a set of
data points
Simple linear regression - the simplest form of
regression that involves a linear relationship between
two variables
 The object of simple linear regression is to obtain an
equation of a straight line that minimizes the sum of
squared vertical deviations from the line (i.e., the least
squares criterion)

#profevad2015
51
Least Squares Line
yc  a  bx
where
yc  Predicted (dependent) variable
x  Predictor (independe nt) variable
b  Slope of the line
a  Value of yc when x  0 (i.e., the height of the line at the y intercept)
and
n  xy     x   y 
b
 
n  x2   x
2

a
 y  b x or y  b x
n
where
n  Number of paired observatio ns

#profevad2015 52
Standard Error
Standard error of estimate
A measure of the scatter of points around a regression
line
If the standard error is relatively small, the predictions
using the linear equation will tend to be more accurate
than if the standard error is larger
 y  y 
2
c
Se 
n2
where
S e  standard error of estimate
y  y value of each data point
n  number of data points
#profevad2015 53
Correlation Coefficient
Correlation, r
 A measure of the strength and direction of relationship between
two variables
 Ranges between -1.00 and +1.00

n  xy     x   y 
r
 
n  x2   x
2
 
n  y2   y
2

r2, square of the correlation coefficient

 A measure of the percentage of variability in the values of y that is
“explained” by the independent variable
 Ranges between 0 and 1.00

#profevad2015 54
 Simple Linear Regression Assumptions
1. Variations around the line are random
2. Deviations around the average value (the line)
should be normally distributed
3. Predictions are made only within the range of
observed values

#profevad2015
55
Forecast Accuracy and Control
Forecasters want to minimize forecast errors
It is nearly impossible to correctly forecast real-world
variable values on a regular basis
So, it is important to provide an indication of the extent
to which the forecast might deviate from the value of the
variable that actually occurs
Forecast accuracy should be an important forecasting
technique selection criterion
Error = Actual – Forecast
If errors fall beyond acceptable bounds, corrective
action may be necessary

#profevad2015
56
Forecast Accuracy Metrics
MAD 
 Actual t  Forecast t
MAD weights all errors
n evenly

  Actual t  Forecast t 
2
MSE weights errors according
MSE  to their squared values
n 1

Actual t  Forecast t
 Actual t
100
MAPE weights errors
MAPE 
n according to relative error

#profevad2015 57
Forecast Error Calculation
Actua Forecas
l t (A-F)
Period [|
Error
(A) (F) |Error| Error 2
Error|/Actual]x100
1 107 110 -3 3 9 2.80%
2 125 121 4 4 16 3.20%
3 115 112 3 3 9 2.61%
4 118 120 -2 2 4 1.69%
5 108 109 1 1 1 0.93%
Sum 13 39 11.23%
n=5 n-1 = 4 n=5
MAD MSE MAPE
= 2.6 = 9.75 = 2.25%

#profevad2015 58
 Issues to consider:
Always plot the line to verify that a linear
relationship is appropriate
The data may be time-dependent.
If they are
 use analysis of time series
 use time as an independent variable in a multiple
regression analysis
A small correlation may indicate that other
variables are important

#profevad2015
59
 Monitoring the Forecast
 Tracking forecast errors and analyzing them can provide useful
insight into whether forecasts are performing satisfactorily
 Sources of forecast errors
 The model may be inadequate
 Irregular variations may have occurred
 The forecasting technique has been incorrectly applied
 Random variation
 Control charts are useful for identifying the presence of non-
random error in forecasts
 Tracking signals can be used to detect forecast bias

#profevad2015
60
Choosing a Forecasting Technique
Factors to consider
Cost
Accuracy
Availability of historical data
Availability of forecasting software
Time needed to gather and analyze data and prepare a
forecast
Forecast horizon

#profevad2015
61
 Using Forecast Information
Reactive approach
 View forecasts as probable future demand
 React to meet that demand
Proactive approach
 Seeks to actively influence demand
Advertising

 Pricing
 Product/service modifications
 Generally requires either an explanatory model or a subjective
assessment of the influence on demand

#profevad2015
62
Operations Strategy
The better forecasts are, the more able organizations will
be to take advantage of future opportunities and reduce
potential risks
 A worthwhile strategy is to work to improve short-term forecasts
 Accurate up-to-date information can have a significant effect on
forecast accuracy:
 Prices
 Demand

 Other important variables

 Reduce the time horizon forecasts have to cover

 Sharing forecasts or demand data through the
supply chain can improve forecast quality

#profevad2015 63