2023-10-22

McGraw-Hill Education / Jacobs Copyright ©2020 McGraw-Hill Education. All rights reserved. No reproduction or distribution
without the prior written consent of McGraw-Hill Education.

OPERATIONS MANAGEMENT
Lecture - 7

Chapter 3: Forecasting

Forecasting in Starbucks

• Starbucks, the largest coffee chain in the world with over 30000 stores in
80 countries.
o Variety of products beyond coffee, coffee beans, salads, sandwiches, mugs, etc.
o Product offerings varies by season, and some are store location-specific.
o Many are perishable, some runs the risk of becoming obsolete.
o Starbucks branded coffee and ice cream are sold in grocery stores.
o Need for forecasting regional and global demand as well as store-specific or demand.

Supply chain at Starbucks.

• The company brings raw beans from Latin America, Africa and Asia in giant shipping
containers to USA and European 6 storage locations near roasting facilities .
• Roasted coffee is packaged and shipped to distribution centers (DCs) 4 in USA, 2 in
Europe and 2 in Asia).
• DCs also store many more items that are sold through Starbucks besides coffee.

1
 2023-10-22

Forecasting in Operations and Supply Chain Management

• Forecasting is vital to every business organization and impacts every

significant management decision
o Forecasting is the basis of corporate planning and control
o Finance and accounting use forecasts as the basis for budgeting and cost control
o Marketing relies on forecasts to make key decisions such as new product planning
and personnel compensation
o Production uses forecasts to select suppliers, determine capacity requirements, and
to drive decisions about purchasing, staffing, and inventory

• Different activities require different forecasting approaches

o Decisions about overall directions or aggregated demand requires strategic
forecasts
o Tactical forecasts are used to guide day-to-day decisions and the goal is to
estimate demand in short term

Decoupling Points

• Decoupling points occur when inventory is positioned in the supply

chain to allow one operation to act independently of another
• For example, inventory at a retail store separates (buffers) the manufacturer
from the actions of individual consumers
• Forecasts of demand at these decoupling points allows inventory to
be set to the proper level

2
 2023-10-22

Forecasting Methods

 Basic types of forecasts:

 Qualitative
 Quantitative
• Time series analysis (primary focus of this chapter)
• Causal relationships
• Simulation
 Time series analysis is based on the idea that data relating to past
demand can be used to predict future demand

Time Series Model VS. Causal relationship Model

Time Series Model Causal relationship Model

Past predicts future Examines potential cause-effect relationship

Use time series data Use cross sectional or time series data

Key variable is time (t) Key variables are denoted as 𝑋 , 𝑋 , 𝑋 , … , 𝑋

Easy to apple More difficult and time-consuming

Less accurate It provides insight to the system under study

Examples: Moving averages, exponential smoothing Regression models

3
 2023-10-22

Components of Demand

Average demand for the period

Trend: A long-term movement in the historical demand data

Seasonal element: short-term regular and repetitive variations in data

Cyclical elements: long(er) term, occasionally caused by unusual

circumstances, (war, economic downturn, etc.)

Random variation: caused by chance

Autocorrelation: denotes persistence of occurrence (momentum driven)

Components of Demand

Exhibit 3.1

4
 2023-10-22

Components of Demand

• Identification of trend
lines is a common
starting point when
developing a forecast
• Common trend types
include linear, S-curve,
asymptotic, and
exponential : we focus
on linear

10

Time Series Analysis

• Using the past to predict the future

Short term – forecasting less than three months

• Used mainly for tactical decisions (e.g. replenishing inventory)

Medium term – forecasting three months to two years

• Used to develop a strategy which will be implemented over the next
six to eighteen months (e.g. meeting demand)
Long term – forecasting greater than two years
• Useful for detecting general trends and identifying major turning
points

11

5
 2023-10-22

Model Selection

• Choosing an appropriate forecasting model depends upon:

1. Time horizon to forecast
2. Data availability
3. Accuracy required
4. Size of forecasting budget
5. Availability of qualified personnel
• Other factors may also be considered
1. Degree of flexibility (can the firm react quickly if the forecast is inaccurate?)
2. Consequence of a bad forecast (important or costly decisions require a good
forecast)

12

Forecasting Method Selection Guide

Forecasting Method Amount of Historical Data Data Pattern Forecast

Horizon
Simple moving average 6 to 12 months; weekly Stationary (i.e. no Short
data are often used trend or
seasonality)
Weighted moving 5 to 10 observations Stationary Short
average and simple needed to start
exponential smoothing
Exponential smoothing 5 to 10 observations Stationary and Short
with trend needed to start trend

Linear regression 10 to 20 observations Stationary, trend, Short to

and seasonality Medium
Trend and seasonal 2 to 3 observations per Stationary, trend Short to Exhibit 3.3
models season and seasonality Medium

13

6
 2023-10-22

Time Series Models

• No Trend; No Seasonality
• Moving Average Techniques
• Simple Moving Average
• Weighted Moving Average
• Exponential Smoothing Technique
• With Trend but Not Seasonality
• Exponential Smoothing With Trend
• Simple Linear Regression (least Squares method)
• With Trend and Seasonality
• Additive vs Multiplicative Demand Models

14

Smoothing methods

• Now we discuss three forecasting methods that are appropriate for a

time series with a horizontal (stationary) pattern:
• Moving Averages
• Weighted Moving Averages
• Exponential Smoothing
• They are called smoothing methods because their objective is to
smooth out the random fluctuations in the time series
• They are most appropriate for short-range forecasts

15

7
 2023-10-22

Moving Averages

• Forecast is based on average demand over the most recent periods

• Useful when demand is not growing or declining rapidly and no seasonality
is present
• Removes some of the random fluctuation from the data
• The moving average method consists of computing an average of the most
recent n data values for the series and using this average for forecasting the value
of the time series for the next period

∑(𝑚𝑜𝑠𝑡 𝑟𝑒𝑐𝑒𝑛𝑡 ′𝑛 𝑝𝑒𝑟𝑖𝑜𝑑 𝑑𝑎𝑡𝑎) 𝐴 +𝐴 +𝐴 + +𝐴

𝐹 = =
𝑛 𝑛
where 𝐹 = forecast of the time series for period t

• Note that, each observation in the moving average calculation receives the same
weight, 1/n

16

Simple Moving Averages

• To use moving averages to forecast, we must first select the order n,

or number of time series values, to be included in the moving average
• A smaller value of n will track shifts in a time series more quickly than a larger
value of n
• If more past observations are considered relevant, then a larger value of n is
better

17

8
 2023-10-22

Simple Moving Average Example

Week Demand
1 650
2 678
3 720 • Question: What are the 3-week and 6-week
4
5
785
859
moving average forecasts for demand?
6
7
920
850
• Assume you only have 3 weeks and 6 weeks
8 758 of actual demand data for the respective
9 892
10 920 forecasts
11 789
12 844
𝑨𝒕 𝟏 + 𝑨𝒕 𝟐 + 𝑨𝒕 𝟑 + +𝑨𝒕 𝒏
𝑭𝒕 =
𝒏

18

Simple Moving Average Example

Plotting the moving averages and comparing them

shows how the lines smooth out to reveal the overall
Week Demand 3-Week 6-Week
Note that no forecast upward trend in this example
is possible until “n”
1 650
periods have passed
2 678 Simple Moving Average
3 720 950
4 785 682.67 F4=(650+678+720)/3 =682.67 900
5 859 727.67 850
6 920 788.00 800
7 850 854.67 768.67 F7=(650+678+720+785+859+920)/6
750

8 758 876.33 802.00 =768.67 700

9 892 842.67 815.33 650

10 920 833.33 844.00 600

1 2 3 4 5 6 7 8 9 10 11 12
11 789 856.67 866.50
Demand 3-Week 6-Week
12 844 867.00 854.83
Note how the 3-Week is smoother than the
Demand, and 6-Week is even smoother

19

9
 2023-10-22

Weighted Moving Averages

While the moving average formula implies an equal weight being

placed on each value that is being averaged, the weighted moving
average permits an unequal weighting on prior time periods

𝐹 =𝑤 𝐴 +𝑤 𝐴 +𝑤 𝐴 + ⋯+ 𝑤 𝐴
𝑤𝒏 = weight given to time period “t” occurrence (weights must add to one)
𝒘𝒊 = 𝟏
𝒊 𝟏

20

Weighted Moving Averages Example

Question: Given the weekly demand and weights, what is the forecast for the
4th period or Week 4?

Week Demand
Period Weights
1 650
t-1 0.5
2 678
t-2 0.3
3 720
t-3 0.2
4

Note that the weights place more emphasis on the most recent data,
that is time period “t-1”

21

10
 2023-10-22

Weighted Moving Averages Example

• To use this method, we must first select the

number of data values, n, to be included in
the average
• Next, we must choose the weight for each of
the data values
• The more recent observations are typically
given more weight than older observations
• For convenience, the weights should sum to 1

22

Moving Average Methods vs Exponential Smoothing Methods

• If you are making forecasts for 20 thousand items in a Wal-Mart and

using a 60-day moving average method, there is a great
computational burden? You need to keep too much data in storage
and need to do too many calculations.
• Exponential Smoothing technique (which is very easy to use and
understand), resolves this issue (the issue of large number of
calculations and storage requirements) with surprising accuracy.

23

11
 2023-10-22

Exponential Smoothing Method

• This method is a special case of a

weighted moving averages method; 𝑭𝒕 = 𝑭𝒕 𝟏 + 𝜶(𝑨𝒕 𝟏 − 𝑭𝒕 𝟏)
we select only the weight for the most
recent observation Or,
• The weights for the other data values 𝑭𝒕 = 𝜶𝑨𝒕 𝟏 + 𝟏 − 𝜶 𝑭𝒕 𝟏
are computed automatically and Where,
become smaller as the observations 𝐹 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑣𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
grow older 𝐹 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑣𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑡𝑖𝑚𝑒
• The exponential smoothing forecast is 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡−1
a weighted average of all the 𝐴 =𝐴𝑐𝑡𝑢𝑎𝑙 𝑜𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡−1
observations in the time series 𝛼=𝐴𝑙𝑝ℎ𝑎, 𝑡ℎ𝑒 𝑠𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 : Determines
how reactive your forecasts are to actual demand
• The term exponential smoothing changes
comes from the exponential nature of Many time assumes 𝑭𝟏 = 𝑨𝟏 to initiate the
the weighting scheme for the computations
historical values

24

Exponential Smoothing Method Example

Week Demand
1 820 Question: Given the weekly demand data, what are the
2 775 exponential smoothing forecasts for periods 2-10 using
3 680 a=0.2?
4 655 Assume F1=D1
5 750
𝑭𝒕 = 𝑭𝒕 𝟏 + 𝜶(𝑨𝒕 𝟏 − 𝑭𝒕 𝟏)
6 802 Or,
7 798 𝑭𝒕 = 𝜶𝑨𝒕 𝟏 + 𝟏 − 𝜶 𝑭𝒕 𝟏
8 689 Where,
9 775 𝐹 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑣𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
𝐹 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑣𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡−1
10 𝐴 =𝐴𝑐𝑡𝑢𝑎𝑙 𝑜𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡−1
𝛼=𝐴𝑙𝑝ℎ𝑎, 𝑡ℎ𝑒 𝑠𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 : Determines how reactive your
forecasts are to actual demand changes

25

12
 2023-10-22

Exponential Smoothing Method Example

Week Demand Forecast

1 820 820
2 775 820 𝐹 = 𝐹 +𝛼 𝐴 −𝐹 =820 + 0.2(820-820)
3 680 811
4 655 785 𝐹 = 𝐹 +𝛼 𝐴 −𝐹 =811+ 0.2(680-811)
5 750 759
6 802 757 𝐹 = 𝐹 +𝛼 𝐴 −𝐹 =759+ 0.2(750-759)
7 798 766
8 689 772 𝐹 = 𝐹 +𝛼 𝐴 −𝐹 =766+ 0.2(798-766)
9 775 756
10 760 𝐹 = 𝐹 +𝛼 𝐴 −𝐹 =756+ 0.2(775-756)

26

Exponential Smoothing Method Example

Week Demand Alpha=0.1 Alpha=0.2 Alpha=0.6

Exponential Smoothing
1 820 820 820 820
850
2 775 820 820 820
800
3 680 816 811 793
4 655 802 785 725 750
5 750 787 759 683 700
6 802 784 757 723
650
7 798 785 766 770
600
8 689 787 772 787
1 2 3 4 5 6 7 8 9 10
9 775 777 756 728
Demand Alpha=0.1 Alpha=0.2 Alpha=0.6
10 777 760 756

Note how that the smaller alpha results in a smoother line in this example. Forecasts
are less reactive to demand changes in that case.

27

13
 2023-10-22

Classification of Time Series Models

• No Trend; No Seasonality
• Moving Average Techniques
• Simple Moving Average
• Weighted Moving Average
• Exponential Smoothing Technique
• With Trend but Not Seasonality
• Exponential Smoothing With Trend
• Simple Linear Regression (least Squares method)
• With Trend and Seasonality
• Additive vs Multiplicative Demand Models

28

Exponential Smoothing with Trends

𝐹 = 𝐹𝐼𝑇 + 𝛼(𝐴 − 𝐹𝐼𝑇 )

• The presence of a trend in the
data causes the exponential
smoothing forecast to always lag
behind the actual data
• This can be corrected by adding
a trend adjustment
• The trend smoothing constant is
delta (𝛿)
𝑇 = 𝑇 + 𝛿(𝐹 − 𝐹𝐼𝑇 )
𝐹𝐼𝑇 = 𝐹 + 𝑇
Where,
𝐹 =𝐸𝑥𝑝𝑜𝑛𝑒𝑛𝑡𝑖𝑎𝑙𝑙𝑦 𝑠𝑚𝑜𝑜𝑡ℎ𝑒𝑑 𝑓𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
𝑇 =𝐸𝑥𝑝𝑜𝑛𝑒𝑛𝑡𝑖𝑎𝑙𝑙𝑦 𝑠𝑚𝑜𝑜𝑡ℎ𝑒𝑑 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
𝐹𝐼𝑇 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑖𝑛𝑐𝑙𝑢𝑑𝑖𝑛𝑔 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
𝐹𝐼𝑇 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑖𝑛𝑐𝑙𝑢𝑑𝑖𝑛𝑔 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑝𝑒𝑟𝑖𝑜𝑑
𝐴 =𝐴𝑐𝑡𝑢𝑎𝑙 𝑜𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑝𝑒𝑟𝑖𝑜𝑑
𝛼=𝑆𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (𝑎𝑙𝑝ℎ𝑎)
𝛿=𝑆𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (𝑑𝑒𝑙𝑡𝑎)

29

14
 2023-10-22

Exponential Smoothing with Trends Example

MONTH ACTUAL DEMAND

Calculate the exponential smoothing with trend forecast for these data
1 31
2 34
using an α of 0.30, a δ of 0.30, an initial trend forecast (T1) of 1, and an
3 33
initial exponentially smoothed forecast (F1) of 30.
4 35
𝐹 = 𝐹𝐼𝑇 + 𝛼(𝐴 − 𝐹𝐼𝑇 )
5 37
𝑇 = 𝑇 + 𝛿(𝐹 − 𝐹𝐼𝑇 )
6 36
𝐹𝐼𝑇 = 𝐹 + 𝑇
7 38
8 40
Where,
𝐹 =𝐸𝑥𝑝𝑜𝑛𝑒𝑛𝑡𝑖𝑎𝑙𝑙𝑦 𝑠𝑚𝑜𝑜𝑡ℎ𝑒𝑑 𝑓𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
9 40 𝑇 =𝐸𝑥𝑝𝑜𝑛𝑒𝑛𝑡𝑖𝑎𝑙𝑙𝑦 𝑠𝑚𝑜𝑜𝑡ℎ𝑒𝑑 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
𝐹𝐼𝑇 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑖𝑛𝑐𝑙𝑢𝑑𝑖𝑛𝑔 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑒𝑟𝑖𝑜𝑑 𝑡
10 41
𝐹𝐼𝑇 =𝐹𝑜𝑟𝑒𝑐𝑎𝑠𝑡 𝑖𝑛𝑐𝑙𝑢𝑑𝑖𝑛𝑔 𝑡𝑟𝑒𝑛𝑑 𝑓𝑜𝑟 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑝𝑒𝑟𝑖𝑜𝑑
𝐴 =𝐴𝑐𝑡𝑢𝑎𝑙 𝑜𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑝𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑝𝑒𝑟𝑖𝑜𝑑
𝛼=𝑆𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (𝑎𝑙𝑝ℎ𝑎)
𝛿=𝑆𝑚𝑜𝑜𝑡ℎ𝑖𝑛𝑔 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 (𝑑𝑒𝑙𝑡𝑎)

30

Exponential Smoothing with Trends Example

ACTUAL
MONTH F T FIT
DEMAND
1 31 30.00 1.00 31.00
2 34 31.00 1.00 32.00
3 33 32.60 1.18 33.78
4 35 33.55 1.11 34.66 𝐹 = 𝐹𝐼𝑇 + 𝛼 𝐴 − 𝐹𝐼𝑇 = 32 + 0.3 × 34 − 31 = 32.60
5 37 34.76 1.14 35.90 𝑇 = 𝑇 + 𝛿 𝐹 − 𝐹𝐼𝑇 = 1 + 0.3 × 32.60 − 32 = 1.18
6 36 36.23 1.24 37.47 𝐹𝐼𝑇 = 𝐹 + 𝑇 = 32.6 + 1.18 = 33.78
7 38 37.03 1.11 38.14
8 40 38.10 1.10 39.19
9 40 39.43 1.17 40.60
10 41 40.42 1.11 41.54

31

15
 2023-10-22

Choosing Alpha and Delta

• Relatively small values for 𝛼 and 𝛿 are common

• Usually in the range 0.1 to 0.3
• 𝛼 depends upon how much random variation is present
• 𝛿 depends upon how steady the trend is
• Measurements of forecast error can be used to select values of 𝛼 and
𝛿 to minimize overall forecast error

32

Linear Regression Analysis

• Regression is used to identify the functional relationship between two or

more correlated variables, usually from observed data
• One variable (the dependent variable) is predicted for given values of the
other variable (the independent variable)
• Linear regression is a special case which assumes the relationship between
the variables can be explained with a straight line

𝑌 = 𝑎 + 𝑏𝑡
Where,
𝑌 −𝑡ℎ𝑒 𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒, (Sales)
𝑎 −𝑡ℎ𝑒 𝑦−𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑙𝑖𝑛𝑒
𝑏 −𝑡ℎ𝑒 𝑠𝑙𝑜𝑝𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑙𝑖𝑛𝑒
𝑡 −𝑖𝑛𝑑𝑒𝑥 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑

33

16
 2023-10-22

Linear Regression Analysis Example 3.2 – Least Squares Method

Quarter Sales Quarter Sales

1 600 7 2,600
2 1,550 8 2,900
3 1,500 9 3,800
4 1,500 10 4,500
The least squares method determines
5 2,400 11 4,000
the parameters a and b such that the
6 3,100 12 4,900
sum of the squared errors is minimized –
the “least squares”
𝑆𝑢𝑚 𝑜𝑓 𝑠𝑞𝑢𝑎𝑟𝑒𝑑 𝑒𝑟𝑟𝑜𝑟𝑠
𝑦 −𝑌 + 𝑦 −𝑌 + ⋯+ 𝑦 −𝑌

34

Linear Regression Analysis Example 3.2 – Least Squares Method

n t y ty t2
∑ 𝒕𝒚 − 𝒏𝒕̅ 𝒚 𝟐𝟔𝟖, 𝟐𝟎𝟎 − 𝟏𝟐 ∗ 𝟔. 𝟓 ∗ 𝟐, 𝟕𝟕𝟗. 𝟐
1 1 600 600 1 𝒃= = = 𝟑𝟓𝟗. 𝟔
∑ 𝒕𝟐 − 𝒏𝒕̅𝟐 𝟔𝟓𝟎 − 𝟏𝟐 ∗ 𝟔. 𝟓𝟐
2 2 1,550 3,100 4 𝐚 = 𝒚 − 𝒃𝒕̅ = 𝟐, 𝟕𝟕𝟗. 𝟐 − 𝟑𝟓𝟗. 𝟔 ∗ 𝟔. 𝟓 = 𝟒𝟒𝟏. 𝟔𝟕
3 3 1,500 4,500 9
4 4 1,500 6,000 16 𝒀𝟏 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏 = 𝟖𝟎𝟏. 𝟑
5 5 2,400 12,000 25 𝒀𝟐 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟐 = 𝟏, 𝟏𝟔𝟎. 𝟗
6 6 3,100 18,600 36 ⋮
7 7 2,600 18,200 49 𝒀𝟏𝟐 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏𝟐 = 𝟒, 𝟕𝟓𝟕. 𝟏
8 8 2,900 23,200 64
The forecast is then extended to periods 13-16
9 9 3,800 34,200 81
10 10 4,500 45,000 100 𝒀𝟏𝟑 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏𝟑 = 𝟓, 𝟏𝟏𝟔. 𝟒
11 11 4,000 44,000 121 𝒀𝟏𝟒 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏𝟒 = 𝟓, 𝟒𝟕𝟔. 𝟎
12 12 4,900 58,800 144 𝒀𝟏𝟓 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏𝟓 = 𝟓, 𝟖𝟑𝟓. 𝟔
SUM 78 33,350 268,200 650 𝒀𝟏𝟔 = 𝒂 + 𝒃𝒕 = 𝟒𝟒𝟏. 𝟔𝟕 + 𝟑𝟓𝟗. 𝟔 ∗ 𝟏𝟔 = 𝟔, 𝟏𝟗𝟓. 𝟐
Average 6.5 2779.2
𝒕̅ 𝒚

35

17
 2023-10-22

Linear Regression Analysis Example 3.2 – Least Squares Method

Microsoft Excel includes data analysis tools, which

can perform least squares regression on a data set

36

Classification of Time Series Models

• No Trend; No Seasonality
• Moving Average Techniques
• Simple Moving Average
• Weighted Moving Average
• Exponential Smoothing Technique
• With Trend but Not Seasonality
• Exponential Smoothing With Trend
• Simple Linear Regression (least Squares method)
• With Trend and Seasonality
• Additive vs Multiplicative Demand Models

37

18
 2023-10-22

Decomposition of a Time Series

• Chronologically ordered data is referred to as a time series

• A time series may contain one or many elements
• Trend, seasonal, cyclical, autocorrelation, and random
• Identifying these elements and separating the time series data into
these components is known as decomposition
• Trend
• Seasonality

38

Seasonal Variation

• Seasonal variation may be either additive or multiplicative (shown here

with a changing trend)

Trend + Seasonality Trend × Seasonality

39

19
 2023-10-22

Forecasting with Trend and Seasonal Components

• Additive Seasonal Time Series Model

• Forecast including trend and Seasonality = Trend + Seasonality
• Multiplicative Seasonal Time Series Model
• Forecast including trend and Seasonality = Trend × Seasonality
• Mostly we experience multiplicative models.
• Given Trend line example: If we are given the actual demand data and a
Trend line as in exhibit 3.10, see example 3.4 and Obj Question 28 (solution
provided) to see how to find the seasonality indices and making forecasts.
• Trend line is Unknow: if we are given the actual demand data and are asked
to decompose the data into its multiplicative trend and seasonality
components and then make forecasts, follow the next procedure.

40

Decomposition Using Least Squares Regression

1. Decompose the time series into its components

a) Find seasonal component
b) De-seasonalize the demand
c) Find the trend component
2. Forecast future values of each component
a) Project trend component into the future
b) Multiply the trend component by the seasonal component

41

20
 2023-10-22

Multiplicative Time Series Model

Decomposition Steps of Multiplicative Time Series Model

step 1 = avg. demand for the entire data

step 2 = compute avg. demand per season

step 3 = divide avg. seasonal dem. by avg. dem. to get seasonal indices

step 4 = de-seasonalize the data: divide the actuals by its seasonal index

step 5 = apply regression to the de-seasonalized data

step 6 = re-seasonalize the regr. Predictions: Y(regr)*Seasonal Index

42

Multiplicative Time Series Model Example

Period Actual Demand Average of Same Quarters De-seasonalized

Quarter seasonal indices(Step 3)
(t) (y) of Each Year (Step 2) Demand (Step 4)
600 + 2,400 + 3,800 600
1 I 600 3
= 2,267 = 0.8156 = 0.82 .82
= 735.7
.
1,550 + 3,100 + 4,500 1,550
2 II 1,550 3
= 3,050 =1.0975 = 1.10 1.10
= 1,412.4
.
1,500 + 2,600 + 4,000 1,500
3 III 1,500 3
= 2,700 =0.9715 = 0.97 0.97
= 1,544.0
.
1,500 + 2,900 + 4,900 1,500
4 IV 1,500 3
= 3,100 =1.1154= 1.12 1.12
= 1,344.8
.
5 I 2,400 0.82 2942.6
6 II 3,100 Step 1: Average demand 1.10 2824.7
7 III 2,600 2779.17 0.97 2676.2
8 IV 2,900 Step1:Average Demand = (600+1550+1500+…+4900)/12=2779.17 1.12 2599.9
9 I 3,800 0.82 4659.2
10 II 4,500 1.10 4100.4
11 III 4,000 0.97 4117.3
12 IV 4,900 1.12 4392.9

43

21
 2023-10-22

Multiplicative Time Series Model Example

Period (t) De-seasonalized Demand

1 735.7 ∑ 𝒕𝒚 − 𝒏𝒕̅ 𝒚
2 1412.4 𝒃=
∑ 𝒕𝟐 − 𝒏𝒕̅𝟐
3 1544.0
4 1344.8 𝐚 = 𝒚 − 𝒃𝒕̅
5 2942.6 Regression Results:
6 2824.7 Y = 555.0 + 342.2t
7 2676.2
8 2599.9 Step 5 Forecast for
9 4659.2
10 4100.4
periods 13-16
11 4117.3
12 4392.9
13 5003.3
14 5345.5
15 5687.7
16 6029.9

Step 6 Create the final forecast by adjusting the Forecast (F x Seasonal

Period Quarter Y from Regression Seasonal Factor
Factor
regression line by the seasonal factor 13 I 5,003.5 0.82 4,102.87
Y = 555.0 + 342.2t 14 II 5,345.7 1.10 5,880.27
Y = 555 + 342.2(13)= 5003.5 x 0.82 = 4102.87 15 III 5,687.9 0.97 5,517.26
Y = 555 + 342.2(14)= 5345.8 x 1.10 = 5880.27 16 IV 6,030.1 1.12 6,753.71
Y = 555 + 342.2(15)= 5687.9 x 0.97 =5517.26
Y = 555 + 342.2(16)= 6030.1 x 1.12 = 6753.71

44

Forecast Errors

• Forecast error is the difference between the forecast value and what
actually occurred
• All forecasts contain some level of error
• Sources of error
• Bias – when a consistent mistake is made
• Random – errors that are not explained by the model being used
• Measures of error
• Mean absolute deviation (MAD)
• Mean absolute percent error (MAPE)
• Tracking signal

45

22
 2023-10-22

Forecast Error Measurements

 Ideally, MAD will be zero  Mean Absolute Percent Error

(no forecasting error) (MAPE) scales the forecast error to
 Larger values of MAD
the magnitude of demand
MAD
indicate a less accurate MAPE =
model Average Demand

MAD =
∑ 𝐴 −𝐹  Tracking signal indicates whether
𝑛
𝑤ℎ𝑒𝑟𝑒 forecast errors are accumulating
𝑡 = period number over time or not (either positive or
𝐴 = actual demand during period 𝑡 negative errors)
𝐹 = forecast demand during period 𝑡
𝑛 = total number of periods Running sum of forecast errors
𝑇𝑆 =
Mean absolute deviation

46

Computing Forecast Error

Month Forecast Actual Deviation RSFE Abs. Sum of MAD TS = RSFE

(error) Dev. Abs. Dev. (MAE) /MAD
1 1,000 950 -50 -50 50 50 50 -1
2 1,000 1,070 +70 +20 70 120 60 0.33
3 1,000 1,100 +100 +120 100 220 73.3 1.64
4 1,000 960 -40 +80 40 260 65 1.2
5 1,000 1,090 +90 +170 90 350 70 2.4
6 1,000 1,050 +50 +220 50 400 66.7 3.3
Overall
400 66.7 220
MAD = = 66.7 MAPE = = 6.43% TS = = 3.3
6 1036.7 66.7

∑ 𝐴 −𝐹 MAD Running sum of forecast errors

MAD = MAPE = 𝑇𝑆 =
𝑛 Average Demand Mean absolute deviation

47

23
 2023-10-22

Causal Relationship Forecasting

• Causal relationship forecasting uses independent variables other than

time to predict future demand
• This independent variable must be a leading indicator
• Many apparently causal relationships are actually just correlated
events – care must be taken when selecting causal variables

48

Multiple Regression Techniques

• Identify factors (independent variables) that can be used to predict the values for
the forecast variable (e.g., sales).
• A little more involved data collection than the time series cases.
• Use Excel (Tools/Data analysis) to obtain the statistics.
• Check each independent variable and the intercept for statistical significance (p-
values ~ ≤ 0.05)
• Drop insignificant variable(s), one at a time and re-run the model as many times
as needed
• If the “clean” model has a good adjusted R2 (subjective measure) the final model
can be used to make decisions.
See the solution for Problem 29 in the Solutions to Suggested Problems from
Forecasting Chapter.pdf.

49

24
 2023-10-22

Qualitative Forecasting Techniques

• Generally used to take advantage of expert knowledge

• Useful when judgment is required, when products are new, or if the
firm has little experience in a new market
• Examples
• Market research
• Panel consensus
• Historical analogy
• Delphi method

50

Collaborative Planning, Forecasting, and Replenishment (CPFR)

• A web-based process used to coordinate the efforts of a supply chain

• Demand forecasting
• Production and purchasing
• Inventory replenishment
• Integrates all members of a supply chain – manufacturers,
distributors, and retailers
• Depends upon the exchange of internal information to provide a
more reliable view of demand

51

25