Quantitative Analysis for Decision Making
Forecasting
Learning Objectives (1 of 2)
After completing this chapter, students will be able to:
5.1 Understand and know when to use various families of
forecasting models.
5.2 Compare moving averages, exponential smoothing,
and other time-series models.
5.3 Calculate measures of forecast accuracy.
5.4 Apply forecast models for random variations.
5.5 Apply forecast models for trends and random
variations.
Topic Outline
5.1 Types of Forecasting Models
5.2 Components of a Time Series
5.3 Measures of Forecast Accuracy
5.4 Forecasting Models – Random Variations Only
5.5 Forecasting Models – Trend and Random Variations
5.6 Adjusting for Seasonal Variations
5.7 Forecasting Models – Trend, Seasonal, and Random
Variations
Introduction
• Main purpose of forecasting
– Reduce uncertainty and make better estimates of what
will happen in the future
• Subjective methods
– Seat-of-the pants methods, intuition, experience
• More formal quantitative and qualitative techniques
Forecasting Models
Figure 5.1 Forecasting Models
Qualitative Models (1 of 3)
• Incorporate judgmental or subjective factors
– Useful when subjective factors are important or
accurate quantitative data is difficult to obtain
• Common qualitative techniques
1. Delphi method
2. Jury of executive opinion
3. Sales force composite
4. Consumer market surveys
Qualitative Models (2 of 3)
• Delphi Method
– Iterative group process
– Respondents provide input to decision makers
– Repeated until consensus is reached
• Jury of Executive Opinion
– Collects opinions of a small group of high-level
managers
– May use statistical models for analysis
Qualitative Models (3 of 3)
• Sales Force Composite
– Allows individual salespersons estimates
– Reviewed for reasonableness
– Data is compiled at a district or national level
• Consumer Market Survey
– Information on purchasing plans solicited from
customers or potential customers
– Used in forecasting, product design, new product
planning
Time-Series Models (1 of 2)
• Predict the future based on the past
• Uses only historical data on one variable
• Extrapolations of past values of a series
• Ignores factors such as
– Economy
– Competition
– Selling price
Components of a Time Series (1 of 3)
• Sequence of values recorded at successive intervals of
time
• Four possible components
– Trend (T)
– Seasonal (S)
– Cyclical (C)
– Random (R)
Components of a Time Series (2 of 3)
Figure 5.2 Scatter Diagram for Four Time Series of
Quarterly Data
Components of a Time Series (3 of 3)
Figure 5.3 Scatter Diagram of a Time Series with Cyclical
and Random Components
Time-Series Models (2 of 2)
• Two basic forms
– Multiplicative
Demand T S C R
– Additive
Demand T S C R
– Combinations are possible
Measures of Forecast Accuracy (1 of 5)
• Compare forecasted values with actual values
– See how well one model works
– To compare models
Forecast error Actual value Forecast value
• Measure of accuracy
– Mean absolute deviation (MA D):
MAD
forecast error
n
Measures of Forecast Accuracy (2 of 5)
Table 5.1 Computing the Mean Absolute Deviation (MA D)
ACTUAL ABSOLUTE VALUE OF
SALES OF WIRELESS ERRORS (DEVIATION),
MONTH SPEAKERS FORECAST SALES (ACTUAL − FORECAST)
1 110 — —
2 100 110 | 100 110 | 10
3 120 100 | 120 100 | 20
4 140 120 | 140 120 | 20
5 170 140 | 170 140 | 30
6 150 170
| 150 170 | 20
| 160 150 | 10
7 160 150
| 190 160 | 30
8 190 160
| 200 190 | 10
9 200 190
| 190 200 | 10
10 190 200
11 — 190 —
Blank Blank Blank
Sum of errors 160
Blank Blank Blank
MAD 160 9 17.8
Measures of Forecast Accuracy (3 of 5)
Table 5.1 Computing the Mean Absolute Deviation (MA D)
MONTH ACTUAL SALES OF FORECAST SALES
WIRELESS
SPEAKERS
1 110 —
2 100 110
3 120 100
4 140 120
5 170 140
6 150 170
7 160 150
8 190 160
9 200 190
10 190 200
11 — 190
Measures of Forecast Accuracy (4 of 5)
MAD
forecast error 160
17.8
n 9
ACTUAL ABSOLUTE VALUE OF
SALES OF WIRELESS ERRORS (DEVIATION),
MONTH SPEAKERS FORECAST SALES (ACTUAL − FORECAST)
1 110 — —
2 100 110 | 100 110 | 10
3 120 100 | 120 100 | 20
4 140 120 | 140 120 | 20
5 170 140 | 170 140 | 30
6 150 170 | 150 170 | 20
| 160 150 | 10
7 160 150
| 190 160 | 30
8 190 160
| 200 190 | 10
9 200 190
| 190 200 | 10
10 190 200
11 — 190 —
Blank Blank Blank
Sum of errors 160
Blank Blank Blank
MAD 160 9 17.8
Measures of Forecast Accuracy (5 of 5)
• Other common measures
– Mean squared error (M S E)
MSE
(error)2
n
– Mean absolute percent error (MAP E)
error
actual
MAPE 100%
n
– Bias is the average error
Forecasting Random Variations
• No other components are present
• Averaging techniques smooth out forecasts
– Moving averages
– Weighted moving averages
– Exponential smoothing
Moving Averages (1 of 2)
• Used when demand is relatively steady over time
– The next forecast is the average of the most recent n
data values from the time series
– Smooths out short-term irregularities in the data series
Sum of demands in previous n periods
Moving average forecast =
n
Moving Averages (2 of 2)
• Mathematically
Yt Yt 1 ... Yt n 1
Ft 1
n
where
Ft 1 forecast for time period t 1
Yt actual value in time period t
n number of periods to average
Wallace Garden Supply (1 of 4)
• Wallace Garden Supply wants to forecast demand for its
Storage Shed
– Collected data for the past year
– Use a three-month moving average (n 3)
Wallace Garden Supply (2 of 4)
Table 5.2 Wallace Garden Supply Shed Sales
Weighted Moving Averages
• Weighted moving averages use weights to put more
emphasis on previous periods
– Often used when a trend or other pattern is emerging
Ft 1
(Weight in period i )(Actual value in period i )
(Weights)
– Mathematically
w1Yt w 2Yt 1 ... w nYt n 1
Ft 1
w1 w 2 ... w n
where
w i weight for the i th observation
Wallace Garden Supply (3 of 4)
• Use a 3-month weighted moving average model to
forecast demand
– Weighting scheme
Wallace Garden Supply (4 of 4)
Table 5.3 Weighted Moving Average Forecast for Wallace
Garden Supply
Exponential Smoothing (1 of 2)
• Exponential smoothing
– A type of moving average
– Easy to use
– Requires little record keeping of data
New forecast = Last period’s forecast
+α (Last period’s actual demand
Last period’s forecast)
α is a weight (or smoothing constant) with a value 0 α 1
Exponential Smoothing (2 of 2)
• Mathematically
Ft 1 Ft (Yt Ft )
where
Ft 1 new forecast (for time period t + 1)
Yt pervious forecast for time period t
α smoothing constant (0 α 1)
Yt pervious period’s actual demand
The idea is simple the new estimate is the old estimate
plus some fraction of the error in the last period
Exponential Smoothing Example (1 of 2)
• In January, February’s demand for a certain car model was
predicted to be 142
• Actual February demand was 153 autos
• Using a smoothing constant of α 0.20, what is the forecast
for March?
New forecast for March demand = 142 + 0.2(153 142)
= 144.2 or 144 autos
• If actual March demand = 136
New forecast for April demand = 144.2 + 0.2 (136 144.2)
= 142.6 or 143 autos
Exponential Smoothing Example (2 of 2)
• Selecting the appropriate value for α is key to obtaining
a good forecast
• The objective is always to generate an accurate forecast
• The general approach is to develop trial forecasts with
different values of α and select the α that results in the
lowest MA D
Port of Baltimore Example (1 of 2)
Table 5.4 Port of Baltimore Exponential Smoothing
Forecasts for α = 0.10 and α = 0.50
ACTUAL FORECAST
TONNAGE FORECAST USING
QUARTER UNLOADED USING α = 0.10
Alpha equals 0.10
α = 0.50
Alpha equals 0.50
1 180 175 175
2 168 175.5 = 175.00 + 0.10(180 175)
177.5
175.5 equals 175.00 plus 0.10, left parenthesis, 180 minus 175, right parenthesis.
174.75 = 175.50 + 0.10(168 175.50)
3 159
173.18 = 174.75 + 0.10 159 174.75 172.75
4 175 173.36 = 173.18 + 0.10(175 173.18)
165.88
173.18 equals 174.75 plus 0.10, left parenthesis, 159 minus 174.75, right parenthesis.
175.02 = 173.36 + 0.10(190 173.36)
5 190
170.44
173.36 equals 173.18 plus 0.10, left parenthesis, 175 minus 173.18, right parenthesis.
178.02 = 175.02 + 0.10(205 175.02)
6 205 178.22 = 178.02 + 0.10(180 178.02) 180.22
175.02 equals 173.36 plus 0.10, left parenthesis, 190 minus 173.36, right parenthesis.
7 180 178.60 = 178.22 + 0.10(182 178.22)
192.61
178.02 equals 175.02 plus 0.10, left parenthesis, 205 minus 175.02, right parenthesis.
8 182
186.30
178.22 equals 178.02 plus 0.10, left parenthesis, 180 minus 178.02, right parenthesis.
Port of Baltimore Example (2 of 2)
Table 5.5 Absolute Deviations and MAD s for the Port of
Baltimore Example
Exponential Smoothing with Trend (1
of 2)
• The equation for the trend correction uses a new
• smoothing constant β
Ft and Tt must be given or estimated
• Three steps in developing FITt
Step 1: Compute smoothed forecast Ft+1
Smoothed forecast = Previous forecast
including trend + a (Last
error)
Ft 1 FITt (Yt FITt )
Exponential Smoothing with Trend
(2 of 2)
Step 2: Update the trend (Tt +1 ) using
Smoothed forecast = Previous forecast including
trend b(Error or excess in
trend)
Tt 1 Tt (Ft 1 FITt )
Step 3: Calculate the trend-adjusted exponential smoothing
forecast (FITt+1 ) using
Forecast including trend (FITt 1 ) Smoothed forecast (Ft 1 )
Smoothed trend (Tt 1 )
FITt 1 Ft 1 Tt 1
Selecting a Smoothing Constant
• A high value of β makes the forecast more responsive
to changes in trend
• A low value of β gives less weight to the recent trend
and tends to smooth out the trend
• Values are often selected using a trial-and-error approach
based on the value of the MA D for different
values of β
Midwestern Manufacturing (1 of 6)
• Demand for electrical generators from 2007 – 2013
– Midwest assumes F1 is perfect, T1 0, α 0.3, β 0.4
Table 5.6 Midwestern Manufacturing’s Demand
YEAR ELECTRICAL GENERATORS SOLD
2007 74
2008 79
2009 80
2010 90
2011 105
2012 142
2013 122
Midwestern Manufacturing (2 of 6)
For 2008 (time period 2)
Step 1: Compute Ft +1
F2 = FIT1 + α (Y1 FIT1 )
=74 + 0.3(74 74) = 74
Step 2: Update the trend
T2 = T1 + (F2 FIT1 )
=0 +.4(74 74) = 0
Midwestern Manufacturing (3 of 6)
Step 3: Calculate the trend-adjusted exponential
smoothing forecast (Ft +1 ) using
FIT2 = F2 + T2
=74 + 0 = 74
Midwestern Manufacturing (4 of 6)
For 2009 (time period 3)
Step 1: F3 = FIT2 + α (Y2 FIT2 )
= 74 + 0.3(79 74) = 75.5
Step 2: T3 = T2 + .4(F3 FIT2 )
= 0 + .4(75.5 74) = 0.6
Step 3: FIT3 = F3 + T3
= 75.5 + 0.6 = 76.1
Midwestern Manufacturing (5 of 6)
Table 5.7 Midwestern Manufacturing Exponential Smoothing
with Trend Forecasts
TIM DEMAN
E (t) D (Yt) Ft +1 = FITt + 0.3(Yt FITt ) Tt+1 Tt + 0.4(Ft +1 FITt ) FITt +1 = Ft +1 + Tt +1
1 74 74 0 74
2 79 74 74 0.3(74 74) 0 0 0.4(74 74) 74 74 0
3 80 75.5 74 0.3(79 74) 0.6 0 0.4(75.5 74) 76.1 75.5 0.6
4 90 77.270 1.068 78.338 77.270 1.068
76.1 0.3(80 76.1) 0.6 0.4(77.27 76.1)
5 105 81.837 2.468 84.305 81.837 2.468
78.338 0.3(90 78.338) 1.068 0.4(81.837 78.338)
6 142 90.514 4.952 95.466 90.514 4.952
84.305 0.3(105 84.305) 2.468 0.4(90.514 84.305)
7 122 109.426 10.536 119.962 109.426
95.446 0.3(142 95.466) 4.952 0.4(109.426 95.466) 10.536
8
Blank
120.573 10.780 131.353 120.573
119.962 0.3(122 10.536 0.4(120.573 10.780
119.962) 119.962)
Trend Projections (1 of 2)
• Fits a trend line to a series of historical data points
• Projected into the future for medium- to long-range
forecasts
• Trend equations can be developed based on exponential
or quadratic models
• Linear model developed using regression analysis is
simplest
Trend Projections (2 of 2)
• Mathematical formula
Yˆ b0 b1 X
where
Yˆ predicted value
b0 intercept
b1 slope of the line
X time period (i.e.,X 1, 2, 3, , n )
Midwestern Manufacturing (1 of 4)
Based on least squares regression, the forecast equation is
Yˆ 56.71 10.54X
• Year 2014 is coded as X = 8
sales in 2014 = 56.71 + 10.54 8
= 141.03, or 141 generators
• For X = 9
sales in 2015 = 56.71 + 10.54 9
= 151.57, or 152 generators
Seasonal Variations
• Recurring variations over time may indicate the need for
seasonal adjustments in the trend line
• A seasonal index indicates how a particular season
compares with an average season
– An index of 1 indicates an average season
– An index > 1 indicates the season is higher than
average
– An index < 1 indicates a season lower than average
Seasonal Indices
• Deseasonalized data is created by dividing each
observation by the appropriate seasonal index
• Once deseasonalized forecasts have been developed,
values are multiplied by the seasonal indices
• Computed in two ways
– Overall average
– Centered-moving-average approach
Seasonal Indices with No Trend (1 of 3)
• Divide average value for each season by the average of
all data
– Telephone answering machines at Eichler Supplies
– Sales data for the past two years for one model
– Create a forecast that includes seasonality
Seasonal Indices with No Trend (2 of 3)
Table 5.8 Answering Machine Sales and Seasonal Indices
Seasonal Indices with No Trend (3 of 3)
Jan July.
Feb. Aug.
Mar. Sep
Apr. Oct.
Nov.
May.
June. Dec
Seasonal Indices with Trend (1 of 2)
• Changes could be due to trend, seasonal, or random
• Centered moving average (CM A) approach prevents
trend being interpreted as seasonal
• Turner Industries sales contain both trend and seasonal
components
Seasonal Indices with Trend (2 of 2)
• Steps in CM A
1. Compute the CM A for each observation (where
possible)
2. Compute the seasonal ratio = Observation ÷ CMA
for that observation
3. Average seasonal ratios to get seasonal indices
4. If seasonal indices do not add to the number of
seasons, multiply each index by (Number of seasons)
÷ Sum of indices
Turner Industries (1 of 7)
Table 5.9 Quarterly Sales $1,000,000s for Turner
Industries
QUARTER YEAR 1 YEAR 2 YEAR 3 AVERAGE
1 115.67
108 116 123
2 133.67
125 134 142
3 159.00
150 159 168
4 152.67
141 152 165
Average 140.25
131.00 140.25 149.50
Turner Industries (2 of 7)
Table 5.9 Quarterly Sales $1,000,000s for Turner
Industries
Turner Industries (3 of 7)
• To calculate the CM A for quarter 3 of year 1, compare the
actual sales with an average quarter centered on that time
period
• Use 1.5 quarters before quarter 3 and 1.5 quarters after
quarter 3
– Take quarters 2, 3, and 4 and one half of quarters 1,
year 1 and quarter 1, year 2
Turner Industries (4 of 7)
• Compare the actual sales in quarter 3 to the CM A to find
the seasonal ratio
Turner Industries (5 of 7)
Table 5.10 Centered Moving Averages and Seasonal Ratios
for Turner Industries
SEASONAL
YEAR QUARTER SALES ($1,000,000s) CM A
RATIO
1 1 108
Blank Blank
2 125
Blank Blank Blank
3 150 132.000 1.136
Blank
4 141 134.125 1.051
Blank
2 1 116 136.375 0.851
2 134 138.875 0.965
Blank
3 159 141.125 1.127
Blank
4 152 143.000 1.063
Blank
3 1 123 145.125 0.848
2 142 147.875 0.960
Blank
3 168
Blank Blank Blank
4 165
Blank Blank Blank
Turner Industries (6 of 7)
• The two seasonal ratios for each quarter are averaged to
get the seasonal index
Index for quarter 1 = I1 = 0.851 + 0.848 ÷ 2 = 0.85
Index for quarter 2 = I2 = 0.965 + 0.960 ÷ 2 = 0.96
Index for quarter 3 = I3 = 1.136 + 1.127 ÷ 2 = 1.13
Index for quarter 4 = I4 = 1.051 + 1.063 ÷ 2 = 1.06
Turner Industries (7 of 7)
• Scatterplot of Turner Industries Sales Data and Centered
Moving Average
