OPM 301, Chapter II: Forecasting

Dr. M. Ashraf Hossain

Learning Objectives
Identify or Define:
– Forecasting
– Types of forecasts
– Time horizons
– Approaches to forecasts
Describe or Explain:
– Moving averages
– Exponential smoothing
– Trend projections
– Regression and correlation analysis
– Measures of forecast accuracy
 What is Forecasting?
• Process of predicting a future event
• Underlying basis of all business decisions
– Production Sales will
be $200
– Inventory Million!
– Personnel
– Facilities
• Stages of introduction and growth require longer
forecasts than maturity and decline
• Forecasts useful in projecting -- staffing levels,
inventory levels, and factory capacity,
as product passes through life cycle stages
 Product Demand Charted over 4 Years
with Trend and Seasonality

Seasonal peaks Trend component

Demand for product or service

Actual demand
line

Average demand
over four years
Random
variation
Year Year Year Year
1 2 3 4
 Forecasting
Elements of good forecast
• Accurate
• Timely
• Reliable
• In writing
• Technique should be simple to understand and use
Steps in forecasting process
• Determine the purpose of the forecast
• Establish a time horizon
• Select a forecasting technique
• Gather and analyse relevant data
• Prepare the forecast
• Monitor the forecast
 Forecasting
Approaches to Forecasting
1. Based on Judgement and Opinion:

• Executive opinion
• Sales-force opinion
• Consumer surveys
• Delphi method
2. Based on Time Series Data:

• Trend
• Seasonality
 Forecasting
Averaging Techniques Weighted moving average
Simple moving average n

n
Aw  i i

 Ai Ft  i 1
n
Ft  i 1

n
w
i 1
i

t 1    At 1  Ft 1 
Exponential Smoothing F  F
t

For smoother demand,  is less, the forecasting trend will be

smoother. For fluctuating demand,  is more, the trend will be
less smooth
Simple Linear Regression
n xy   x  y a  y  bx
y  a  bx b
n x    x 
2 2
 Forecasting
Accuracy of Forecasts

et  At  Ft
Two commonly used measures:

Mean absolute
MAD 
 Actual  Forecast
deviation (MAD)
n
Mean squared error (MSE)

  Actual  Forecast  2

MSE 
n 1
 Forecasting at Tupperware

• Each of 50 profit centers around the

world is responsible for computerized
monthly, quarterly, and 12-month sales
projections
• These projections are aggregated by
region, then globally, at Tupperware’s
World Headquarters
• Tupperware uses all techniques discussed
in text
Three Key Factors for Tupperware
• The number of registered “consultants” or sales
representatives
• The percentage of currently “active” dealers (this
number changes each week and month)
• Sales per active dealer, on a weekly basis

Tupperware - Forecast by Consensus

• Although inputs come from sales, marketing,
finance, and production, final forecasts are the
consensus of all participating managers.
• The final step is Tupperware’s version of the “jury
of executive opinion”
 Types of Forecasts by Time Horizon

• Short-range forecast
– Up to 1 year; usually less than 3 months
– Job scheduling, worker assignments
• Medium-range forecast
– 3 months to 3 years
– Sales & production planning, budgeting
• Long-range forecast
– 3+ years
– New product planning, facility location
 Short-term vs. Longer-term Forecasting

• Medium/long range forecasts deal with more

comprehensive issues and support
management decisions regarding planning
and products, plants and processes.
• Short-term forecasting usually employs
different methodologies than longer-term
forecasting
• Short-term forecasts tend to be more
accurate than longer-term forecasts.
 Types of Forecasts

• Economic forecasts
– Address business cycle, e.g., inflation rate,
money supply etc.
• Technological forecasts
– Predict rate of technological progress
– Predict acceptance of new product
• Demand forecasts
– Predict sales of existing product
 Forecasting Approaches
Qualitative Methods Quantitative Methods
• Used when situation • Used when situation
is vague & little data is ‘stable’ & historical
exist data exist
– New products – Existing products
– New technology – Current technology
• Involves intuition, • Involves
experience mathematical
– e.g., forecasting sales techniques
on Internet – e.g., forecasting sales
of color televisions
 Overview of Qualitative Methods
• Jury of executive opinion
– Pool opinions of high-level executives, sometimes
augment by statistical models
• Delphi method
– Panel of experts, queried iteratively
• Sales force composite
– Estimates from individual salespersons are
reviewed for reasonableness, then aggregated
• Consumer Market Survey
– Ask the customer
 Jury of Executive Opinion
• Involves small group of high-level
managers
– Group estimates demand by working
together
• Combines managerial experience with
statistical models
• Relatively quick
• ‘Group-think’
disadvantage
 Sales Force Composite
• Each salesperson
Sales
projects his or her
sales
• Combined at district &
national levels
• Sales reps know
customers’ wants
• Tends to be overly
optimistic
 Delphi Method
• Iterative group
process Decision Makers
(Sales?)
• 3 types of people (Sales will be 50!)
Staff
– Decision makers (What will
– Staff sales be?
– Respondents survey)

• Reduces ‘group-
Respondents
think’ (Sales will be 45, 50, 55)
 Consumer Market Survey
How many hours will
you use the Internet
• Ask customers
next week?
about purchasing
plans
• What consumers
say, and what
they actually do
are often different
• Sometimes
difficult to answer
 Overview of Quantitative Approaches

• Naïve approach
• Moving averages
• Exponential smoothing Time-series
Models
• Trend projection

• Linear regression Associative

models
Quantitative Forecasting Methods
(Non-Naive)
Quantitative
Forecasting

Time Series Associative

Models Models

Moving Exponential Trend Linear

Average Smoothing Projection Regression
 What is a Time Series?
• Set of evenly spaced numerical data
– Obtained by observing response variable at
regular time periods
• Forecast based only on past values
– Assumes that factors influencing past and
present will continue influence in future
• Example
Year: 19981999200020012002
Sales: 78.763.589.793.292.1
 Time Series Components

Trend Cyclical

Seasonal Random
Trend Component
• Persistent, overall
upward or downward
pattern Response
• Due to population,
technology etc. Mo., Qtr., Yr. © 1984-1994 T/Maker Co.

• Several years duration

 Seasonal Component
• Regular pattern of up &
down fluctuations Summer
• Due to weather, customs
Response
etc.
• Occurs within 1 year
Mo., Qtr.

Common Seasonal Patterns

Period of Pattern “Season” Length No. of “Seasons” in Pattern
Week Day 7
Month Week 4–4½
Month Day 28 – 31
Year Quarter 4
Year Month 12
Year Week 52
Cyclical Component
• Repeating up & down movements
• Due to interactions of factors influencing economy
• Usually 2-10 years duration Cycle
Response

Random Component 
Mo., Qtr., Yr.
• Erratic, unsystematic, ‘residual’ fluctuations
• Due to random variation or unforeseen events
– Union strike
– Tornado
• Short duration & non-repeating
 General Time Series Models
• Any observed value in a time series is the
product (or sum) of time series components

• Multiplicative model
Yi = Ti · Si · Ci · Ri (if quarterly or mo. data)

• Additive model
Yi = Ti + Si + Ci + Ri (if quarterly or mo.
data)
 Naive Approach
• Assumes demand in next period is the
same as demand in most recent period
– e.g., If May sales were 48, then June
sales will be 48
• Sometimes cost effective & efficient

Moving Average Method

• MA is a series of arithmetic means
• Used if little or no trend
• Used often for smoothing
– Provides overall impression of data over time
• Equation
MA   Demand in Previous n Periods
n
 Moving Average Example
You’re manager of a museum store that
sells historical replicas. You want to
forecast sales (000) for 2016 using a
3-period moving average.
2011 4
2012 6
2013 5
2014 3
2015 7
.
 Moving Average Solution
Time Response Moving Moving Time Response Moving Moving
Yi Total Average Yi Total Average
(n=3) (n=3) (n=3) (n=3)
2011 4 NA NA 2011 4 NA NA
2012 6 NA NA 2012 6 NA NA
2013 5 NA NA
2013 5 NA NA
2014 3 4+6+5=15 15/3 = 5
2015 7
2014 3 4+6+5=15 15/3 = 5
2016 NA 2015 7 6+5+3=14 14/3=4 2/3
2016 NA

Time Response Moving Moving Sales

Yi Total Average
(n=3) (n=3)
8 Actual
2011 4 NA NA
2012 6 NA NA 6 Forecast
2013
2014
5
3
NA
4+6+5=15
NA
15/3=5.0
4
2015 7 6+5+3=14 14/3=4.7 2 Moving Average Graph
2016 NA 5+3+7=15 15/3=5.0
95 96 97 98 99 00
Year
 Weighted Moving Average Method

• Used when trend is present

– Older data usually less important
• Weights based on intuition
– Often lay between 0 & 1, & sum to 1.0
• Equation

Σ(Weight for period n) (Demand in period n)

WMA =
ΣWeights
 Actual Demand, Moving Average, Weighted
Moving Average
35
Weighted moving average
30
25 Actual sales
Sales Demand

20
15

10 Moving average
5

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month
 Disadvantages of
Moving Average Methods
• Increasing n makes forecast less sensitive
to changes
• Do not forecast trend well
• Require much historical
data

© 1984-1994 T/Maker Co.

Exponential Smoothing Method

• Form of weighted moving average

– Weights decline exponentially
– Most recent data weighted most
• Requires smoothing constant ()
– Ranges from 0 to 1
– Subjectively chosen
• Involves little record keeping of past data
 Exponential Smoothing Equations

• Ft = At - 1 + (1-)At - 2 + (1- )2·At - 3

+ (1- )3At - 4 + ... + (1- )t-1·A0
– Ft = Forecast value
– At = Actual value
 = Smoothing constant
• Ft = Ft-1 + (At-1 - Ft-1)
– Use for computing forecast
 Exponential Smoothing Example
During the past 8 quarters, the Port of Baltimore has unloaded
large quantities of grain. ( = .10). The first quarter forecast
was 175..
Quarter Actual
1 180
2 168
3 159
4 175
Find the forecast
5 190
for the 9th quarter.
6 205
7 180
8 182
9 ?
 Forecast Effects of
Smoothing Constant 
Ft =  At - 1 + (1- ) At - 2 + (1- )2At - 3 + ...
Weights
= Prior Period 2 periods ago 3 periods ago
 (1 - ) (1 - )2

= 0.10
10% 9% 8.1%
= 0.90
90% 9% 0.9%
 Choosing 
Seek to minimize the Mean Absolute Deviation (MAD)

If: Forecast error = demand - forecast

 forecast errors
Then: MAD 
n
 Linear Trend Projection
• Used for forecasting linear trend line
• Assumes relationship between
response variable, Y, and time, X, is a
linear function
Yi  a  bX i
• Estimated by least squares method
– Minimizes sum of squared errors
Linear Regression Equations
or Least Squares Equations
Equation: Ŷi  a  bx i
n
 x i y i  nx y
Slope: b  i 1
n
 x i2  nx 2
i 1

Y-Intercept: a  y  bx
 Computation Table

2 2
Xi Yi Xi Yi X iY i
2 2
X1 Y1 X1 Y1 X 1Y 1
2 2
X2 Y2 X2 Y2 X 2Y 2
: : : : :
2 2
Xn Yn Xn Yn X nY n
2 2
ΣX i ΣYi ΣXi ΣY i Σ X iY i
 Using a Trend Line
Year Demand
The demand for electrical
2009 74
power at a Company over
2010 79
the years 2009 – 2015 is
2011 80
given at the left. Find the
2012 90
related values of a, b, x¨ ÿ
2013 105
of the equation:
2014
2015
142
122
Ŷi  a  bx i
and forecast demand for
next 2 years and also
show overall trend.
 Finding a Trend Line
Year Time Power
Period Demand
x2 xy
(x) (Mega Watts) :y

2009 1 74 1 74
2010 2 79 4 158
2011 3 80 9 240
2012 4 90 16 360
2013 5 105 25 525
2014 6 142 36 852
2015 7 122 49 854
x=28 y=692 x2=140 xy = 3,063
 The Trend Line Equation
Σx 28 Σy 692
x  4 ÿ   98.86
n 7 n 7

Σxy - nxy 3,063  (7)(4)(98.86) 295

b 2    10.54
Σx  nx 2
140  (7)(4) 2
28

a  y - bx  98.86 - 10.54(4)  56.70

Demand in 2016  56.70  10.54(8)  141.02 megawatts

Demand in 2017  56.70  10.54(9)  151.56 megawatts

 Forecast Error Equations
• Mean Square Error (MSE)
n

 (y i  ŷ i ) 2

 forecast errors
2

MSE  i 1

n n

• Mean Absolute Deviation (MAD)

n

| y i  yˆ i |
 | forecast errors |
MAD  i 1

n n
 Selecting Forecasting Model Example

You’re a marketing analyst for a Toys Co. You’ve

forecast sales with a linear model & exponential
smoothing. Find the value of MAD and MSE for both
methods.
Linear Exponential
ActualModel Smoothing
Year___Sales Forecast Forecast__
2011 1 0.6 1.0
2012 1 1.3 1.0
2013 2 2.0 1.9
2014 2 2.7 2.0
2015 4 3.4 3.8
 Linear Model Evaluation
|Error|
Year Yi Y^ i Error Error 2
|Error|
Actual
2011 1 0.6 0.4 0.16 0.4 0.40
2012 1 1.3 -0.3 0.09 0.3 0.30
2013 2 2.0 0.0 0.00 0.0 0.00
2014 2 2.7 -0.7 0.49 0.7 0.35
2015 4 3.4 0.6 0.36 0.6 0.15
Total 0.0 1.10 2.0 1.20
MSE = Σ Error2 / n = 1.10 / 5 = 0.220
MAD = Σ |Error| / n = 2.0 / 5 = 0.400
MAPE = 100 Σ|absolute percent errors|/n= 1.20/5 = 0.240
 Exponential Smoothing Model Evaluation
^ |Error|
Year Y Y Error Error 2
|Error|
i i Actual
2011 1 1.0 0.0 0.00 0.0 0.00
2012 1 1.0 0.0 0.00 0.0 0.00
2013 2 1.9 0.1 0.01 0.1 0.05
2014 2 2.0 0.0 0.00 0.0 0.00
2015 4 3.8 0.2 0.04 0.2 0.05
Total 0.3 0.05 0.3 0.10
MSE = Σ Error2 / n = 0.05 / 5 = 0.01
MAD = Σ |Error| / n = 0.3 / 5 = 0.06
MAPE = 100 Σ |Absolute percent errors|/n = 0.10/5 = 0.02
 Tracking Signal (TS)
RSFE
• Measures how well the TS 
forecast is predicting MAD
actual values
• Ratio of running sum of n
forecast errors (RSFE) to   yi  ŷ i 
mean absolute deviation  i 
(MAD) MAD
– Good tracking signal
has low values
• Should be within upper  forecast error

and lower control limits MAD
 Plot of a Tracking Signal
Signal exceeded limit

Tracking signal
Upper control limit
+

0
MAD

Acceptable range

-
Lower control limit

Time
 Exponential Smoothing Example
During the past 8 quarters, the Port of Baltimore has unloaded
large quantities of grain. ( = .10). The first quarter forecast
was 175..
Quarter Actual
1 180
2 168
3 159
4 175
Find the forecast
5 190
for the 9th quarter.
6 205
7 180
8 182
9 ?
 Exponential Smoothing Solution

Ft = Ft-1 + 0.1(At-1 - Ft-1)

Forecast, F t
Quarter Actual
( α = .10)
1 180 175.00 (Given)
2 168 175.00 +
3 159
4 175
5 190
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Quarter Actual
Actua
( α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(
3 159
4 175
5 190
6 205
 Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, Ft
Quarter Actual
( α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 -
3 159
4 175
5 190
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, Ft
Quarter Actual
(α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00)
3 159
4 175
5 190
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, Ft
Quarter Actual
(α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00) = 175.50
3 159
4 175
5 190
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Quarter Actual
( α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00) = 175.50
3 159 175.50 + .10(168 - 175.50) = 174.75
4 175
5 190
6 205
 Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Quarter Actual
(α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00) = 175.50
3 159 175.50 + .10(168 - 175.50) = 174.75
4 175 174.75 + .10(159 - 174.75)= 173.18
5 190
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Quarter Actual
(α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00) = 175.50
3 159 175.50 + .10(168 - 175.50) = 174.75
4 175 174.75 + .10(159 - 174.75) = 173.18
5 190 173.18 + .10(175 - 173.18) = 173.36
6 205
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Quarter Actual
(α = .10)
1 180 175.00 (Given)
2 168 175.00 + .10(180 - 175.00) = 175.50
3 159 175.50 + .10(168 - 175.50) = 174.75
4 175 174.75 + .10(159 - 174.75) = 173.18
5 190 173.18 + .10(175 - 173.18) = 173.36
6 205 173.36 + .10(190 - 173.36) = 175.02
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Time Actual
(α = .10)
4 175 174.75 + .10(159 - 174.75) = 173.18
5 190 173.18 + .10(175 - 173.18) = 173.36
6 205 173.36 + .10(190 - 173.36) = 175.02
7 180 175.02 + .10(205 - 175.02) = 178.02
8
9
Exponential Smoothing Solution
Ft = Ft-1 + 0.1(At-1 - Ft-1)
Forecast, F t
Time Actual
(α = .10)
4 175 174.75 + .10(159 - 174.75) = 173.18
5 190 173.18 + .10(175 - 173.18) = 173.36
6 205 173.36 + .10(190 - 173.36) = 175.02
7 180 175.02 + .10(205 - 175.02) = 178.02
8 182 178.02 + .10(180 - 178.02) = 178.22
9 ? 178.22 + .10(182 - 178.22) = 178.58