3-1 Forecasting

CHAPTER
3
3-2 Forecasting

FORECAST:
 A statement about the future value of a variable of
interest such as demand.
 Forecasts affect decisions and activities throughout an
organization
 Accounting, finance
 Human resources
 Marketing
 MIS
 Operations
 Product / service design
3-3 Forecasting

Why???
 "Those who have knowledge, don't predict.
Those who predict, don't have knowledge. "
--Lao Tzu, 6th Century BC Chinese Poet
 The company that doesn’t see trouble ahead is

headed for real trouble.

 "A good forecaster is not smarter than

everyone else, he merely has his ignorance

better organised. "
--Anonymous
3-4 Forecasting

Uses of Forecasts

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

3-5 Forecasting

Common features of forecasting

 It assumes causal system: past ==> future
 Forecasts rarely perfect because of randomness
 Forecasts more accurate for groups vs. individuals
 Forecast accuracy decreases as time horizon increases
I see that you
In your will get an A
. help this semester
3-6 Forecasting

Elements of a Good Forecast

Timely

Reliable Accurate

e
f ul us
n g to
ni Written y
s
ea Ea
M
3-7 Forecasting

Steps in the Forecasting Process

“The forecast”

Step 6 Monitor the forecast

Step 5 Prepare the forecast
Step 4 Gather and analyze data
Step 3 Select a forecasting technique
Step 2 Establish a time horizon
Step 1 Determine purpose of forecast
3-8 Forecasting

Approaches to Forecasts
A. Qualitative - uses subjective inputs
 Usually based on judgments/opinions about causal factors

that underlie the demand of particular goods or services

 Do not require a demand history for the product or

service, therefore are useful for new goods/services

 Approaches vary in sophistication from scientifically

conducted surveys to intuitive hunches about future

events
 Involves intuition and experience

 Preferred in the absence of recorded previous data

3-9 Forecasting

Qualitative Forecasts models

 Executive opinions
 Sales force opinions
 Consumer/market surveys
 Marketing research
 Outside opinion – experts
 Historical analogy
 Delphi method
 Opinions of managers and staff
 Achieves a consensus forecast
3-10 Forecasting

B. Quantitative Forecasts Models

i. Time series - uses historical data assuming the future
will be like the past
 Trend - long-term movement in data

 Seasonality - short-term regular variations in data

 Cycle – wavelike variations of more than one year’s

duration
 Irregular variations - caused by unusual circumstances

 Random variations - caused by chance

3-11 Forecasting

Forecast Variations

Irregular
variation

Trend

Cycles

90
89
88
Seasonal variations
3-12 Forecasting

Quantitative Forecasting Approaches

 Based on the assumption that the “forces” that generated
the past demand will generate the future demand, i.e.,
history will tend to repeat itself
 Analysis of the past demand pattern provides a good basis
for forecasting future demand
 Majority of quantitative approaches fall in the category of
time series analysis
 Characters:
 Simple to use
 Quick and easy to prepare
 Easily understandable
 Can be a standard for accuracy
3-13 Forecasting

Time Series Analysis

 A time series is based on a sequence (order) of evenly
spaced data points (weekly, monthly, quarterly, and so
on).
 Forecasting time-series data implies that future values are
predicted only from past values and that other variables
are ignored.
 Analysis of the time series identifies patterns
 Once the patterns are identified, they can be used to
develop a forecast
3-14 Forecasting

Techniques for Averaging

 Simple moving average
 Weighted moving average
 Exponential smoothing
3-15 Forecasting

Simple Moving Average

 An averaging period k is given or selected
 The forecast for the next period is the arithmetic average
of the k most recent actual demands

F = (A n + An-1  ....  An-k+1)

1
n+1 k
 The multiple-period-ahead and the one-period-ahead
forecasts are identical
 It is called a “simple” average because each period used
to compute the average is equally weighted
3-16 Forecasting

Simple Moving Average

 It is called “moving” because as new demand data
becomes available, the oldest data is not used
 By increasing the value of k, the forecast is less
responsive to fluctuations in demand (low impulse
response)
 By decreasing the k, the forecast is more responsive to
fluctuations in demand (high impulse response).
 However small k will result in forecasts with greater
variation from period to period (low stability)
 Compute: the demand for tires in a tire store in the past 5 weeks
were as follows. Compute a three-period moving average forecast
for demand in week 6. 83 80 85 90 94
3-17 Forecasting

Simple Moving Average

Actual
MA5
47
45
43
41
39
37
MA3
35
1 2 3 4 5 6 7 8 9 10 11 12
n

 A i
MAn = i=1

n
3-18 Forecasting

Weighted Moving Average

 This is a variation on the simple moving average where
instead of the weights used to compute the average being
equal, they are not equal
 The weights must add to 1.0 and generally decrease in
value with the age of the data
 This allows more recent demand data to have a greater

effect on the moving average

 Example

F4 = 0.5 A 3 + 0.3 A 2  0.2 A1

3-19 Forecasting

Weighted moving average

 Weighted moving average – More recent values in a
series are given more weight in computing the forecast
Step i. n


i=1
W i=1

n
Step ii.  WiAi
MAn =
i=1

 For the previous demand data, compute a weighted average

forecast using a weight of .40 for the most recent period, .30 for
the next most recent, .20 for the next and .10 for the next.
 If the actual demand for week 6 is 91, forecast demand for week
7 using the same weights
3-20 Forecasting

Exponential Smoothing

Ft = Ft-1 + (At-1 - Ft-1)

• Premise: the most recent observations might have the
highest predictive value.
• Therefore, we should give more weight to the more
recent time periods when forecasting.
• The forecast is the weighted average of the old forecast
and the actual/observed value
 Weighted averaging method based on previous forecast
plus a percentage of the forecast error
 A-F is the error term,  is the % feedback
3-21 Forecasting

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
3-22 Forecasting

Picking a Smoothing Constant

Actual
50
.4
.1
45
Demand

40

35
1 2 3 4 5 6 7 8 9 10 11 12
Period
3-23 Forecasting

Common Nonlinear Trends

Figure 3.5

Parabolic

Exponential

Growth
3-24 Forecasting

Linear Trend Equation

Ft

Ft = a + bt

0 1 2 3 4 5 t

 Ft = Forecast for period t

 t = Specified number of time periods
 a = Value of Ft at t = 0
 b = Slope of the line
3-25 Forecasting

Calculating a and b

n  (ty) -  t  y
b =
n  t 2 - (  t) 2

 y - b t
a =
n
3-26 Forecasting

Linear Trend Equation Example

t y
2
W eek t S a le s ty
1 1 150 150
2 4 157 314
3 9 162 486
4 16 166 664
5 25 177 885

 t = 15  t 2 = 5 5  y = 812  ty = 2 4 9 9
(  t ) = 2 2 5
2
3-27 Forecasting

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
3-28 Forecasting

Exercise 1
 National Mixer Inc. sells can openers.
Monthly sales for a seven-month period Month Sales
were as follows: (1000)
 Forecast September sales volume using each Feb 19
of the following:
Mar 18
 A five-month moving average
 Exponential smoothing with a smoothing constant Apr 15
equal to .20, assuming a March forecast of 19. May 20
 The naive approach
Jun 18
 A weighted average using .60 for August, .30 for
July, and .10 for June. Jul 22
Aug 20
3-29 Forecasting

Exercise 2
 A dry cleaner uses simple exponential smoothing to
forecast equipment usage at its main plant. August usage
was forecast to be 88% of capacity. Actual usage was
89.6%. A smoothing constant of 0.1 is used.
 Prepare a forecast for September
 Assuming actual September usage of 92%, prepare

a forecast of October usage

3-30 Forecasting

Exercise 3
An electrical contractor’s records during the last five
weeks indicate the number of job requests:
Week: 1 2 3 4 5
Requests: 20 22 18 21 22
Predict the number of requests for week 6 using each of
these methods:
 Naïve/time series
 A four-period moving average
 Exponential smoothing with a smoothing constant
of .30. Use 20 for week 2 forecast.
3-31 Forecasting

Double exponential smoothing

 Single smoothing does not excel in following the data when
there is a trend
 This method is used when the data shows a trend.
 Exponential smoothing with a trend works much like simple
smoothing except that two components must be updated each
period - level and trend.
 Lt = a Dt+ (1 - a) Ft - for the level and
 Tt = b (Lt - Lt-1 ) + (1 - b) Tt-1 - for the trend.
 Then the forecasting formula in double exponential smoothing
is
 The one period ahead forecast is given by
Fn+1= Ln + Tn
 The ‘k’ period ahead forecast is given by
Fn+k = Ln + k. Tn
3-32 Forecasting

Exponential Smoothing Example

During the past 8 quarters, ECX has unloaded large quantities of
coffee. ( = .10). The first quarter forecast was 175..
Quarter Actual
1 180
2 168
Find the forecast for the 9th
3 159
4 175
quarter.
5 190
6 205
7 180
8 182
9 ?
3-33 Forecasting

Cont’d …
ii. Associative models - uses explanatory variables to
predict the future

Uh, give me a minute....

We sold 250 wheels last
week.... Now, next week we
should sell....

Assumption: the forecast for any period

equals the previous period’s actual value
3-34 Forecasting

Associative Forecasting
 Predictor variables - used to predict values of variable interest
 Regression - technique for fitting a line to a set of points
 Least squares line - minimizes sum of squared deviations around
the line
 Simple linear regression analysis analyzes the linear
relationship that exists between two variables.

F a  bt
where:
F= Value of the dependent variable
t = Value of the independent variable
a = Population’s F-intercept
b = Slope of the population regression line
3-35 Forecasting

Linear Model Seems Reasonable

X Y Computed
7 15 relationship
2 10
6 13 50

4 15 40

14 25 30

15 27 20

16 24
10

0
12 20 0 5 10 15 20 25

14 27
20 44
15 34 A straight line is fitted to a set of sample points.
7 17
3-36 Forecasting

Forecast Accuracy
 Error - difference between actual value and predicted
value
 Accuracy is how well the forecasted values match the
actual values
 Accuracy of a forecasting approach needs to be
monitored to assess the confidence you can have in its
forecasts and changes in the market may require
reevaluation of the approach
3-37 Forecasting

Accuracy can be measured in several ways

 Mean Absolute Deviation (MAD)
 Average absolute error
 Mean Squared Error (MSE)
 Average of squared error
 Mean Absolute Percent Error (MAPE)
 Average absolute percent error
3-38 Forecasting

MAD, MSE, and MAPE

 Actual  forecast
MAD =
n
2
 ( Actual  forecast)
MSE =
n -1

 Actual  forecast / Actual*100)

MAPE =
n
3-39 Forecasting

Example
Period Actual Forecast (A-F) |A-F| (A-F)^2 (|A-F|/Actual)*100
1 217 215 2 2 4 0.92
2 213 216 -3 3 9 1.41
3 216 215 1 1 1 0.46
4 210 214 -4 4 16 1.90
5 213 211 2 2 4 0.94
6 219 214 5 5 25 2.28
7 216 217 -1 1 1 0.46
8 212 216 -4 4 16 1.89
-2 22 76 10.26

MAD= 2.75
MSE= 10.86
MAPE= 1.28
3-40 Forecasting

Controlling the Forecast

 Control chart
 A visual tool for monitoring forecast errors

 Used to detect non-randomness in errors

 Forecasting errors are in control if

 All errors are within the control limits

 No patterns, such as trends or cycles, are present

Sources of error
 Model may be inadequate
 Irregular variations
 Incorrect use of forecasting technique
3-41 Forecasting

Choosing a Forecasting Technique

 No single technique works in every situation
 Two most important factors
 Cost

 Accuracy

 Other factors include the availability of:

 Historical data

 Computers

 Time needed to gather and analyze the data

 Forecast horizon
3-42 Forecasting

Reasons for Ineffective Forecasting

 Not involving a broad cross section of people
 Not recognizing that forecasting is integral to business
planning
 Not recognizing that forecasts will always be wrong
 Not forecasting the right things
 Not selecting an appropriate forecasting method
 Not tracking the accuracy of the forecasting models
3-43 Forecasting