FORECASTING

Chapter Eighteen
McGraw-Hill/Irwin Copyright © 2014 by The McGraw-Hill Companies, Inc. All rights reserved.
Learning Objectives
 LO18–1: Understand how forecasting is essential
to supply chain planning.
 LO18–2: Evaluate demand using quantitative
forecasting models.
 LO18–3: Apply qualitative techniques to forecast
demand.
 LO18–4: Apply collaborative techniques to forecast
demand.

18-2
The Role of Forecasting
 Forecasting is a vital function and affects every significant
management decision.
 Finance and accounting use forecasts as the basis for budgeting
and cost control.
 Marketing relies on forecasts to make key decisions such as new
product planning and personnel compensation.
 Production uses forecasts to select suppliers; determine capacity
requirements; and drive decisions about purchasing, staffing, and
inventory.
 Different roles require different forecasting approaches.
 Decisions about overall directions require strategic forecasts.
 Tactical forecasts are used to guide day-to-day decisions.

18-3
Forecasting and Decoupling Point
 Decoupling point: Point at which inventory is
stored, which allows SC to operate independently
 The choice of the decoupling point in a SC is
strategic.
 Forecasting helps determine the level of inventory
needed at the decoupling points.
 The decision will be affected by the error produced
in the forecast and the type of product (easily
inventoried or easily perishable).

18-4
Types of Forecasting
 There are four basic types of forecasts.
1. Qualitative
2. Time series analysis (primary focus of this chapter)
3. Causal relationships
4. Simulation
 Time series analysis is based on the idea that data
relating to past demand can be used to predict
future demand.

18-5
Components of Demand

Average
demand for a Trend
period of time

Seasonal Cyclical
element elements

Random Autocorrelatio
variation n Excel: Components
of Demand

18-6
Trends
 Identification of trend lines is a common starting
point when developing a forecast.
 Common trend types include linear, S-curve,
asymptotic, and exponential.

18-7
Time Series Analysis
 Using the past to predict the future
Short term – forecasting less than 3 months

• Used mainly for tactical decisions

Medium term – forecasting 3 months to 2 years

• Used to develop a strategy that will be implemented over the next 6

to 18 months (e.g., meeting demand)

Long term – forecasting greater than 2 years

• Useful for detecting general trends and identifying major turning

points

18-8
Model Selection
 Choosing an appropriate forecasting model
depends upon
1. Time horizon to be forecast
2. Data availability
3. Accuracy required
4. Size of forecasting budget
5. Availability of qualified personnel

18-9
 Forecasting Method Selection Guide

Forecasting Method Amount of Historical Data Pattern Forecast

Data 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 trend Short
with trend needed to start

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

and seasonality medium

18-10
Simple Moving Average
 Forecast is the average of a fixed number of past
periods.
 Useful when demand is not growing or declining
rapidly and no seasonality is present.
 Removes some of the random fluctuation from the
data.
 Selecting the period length is important.
 Longer periods provide more smoothing.
 Shorter periods react to trends more quickly.

18-11
Simple Moving Average Formula


18-12
Simple Moving Average –
Example

18-13
Weighted Moving Average
 The simple moving average formula implies equal
weighting for all periods.
 A weighted moving average allows unequal
weighting of prior time periods.
 The sum of the weights must be equal to one.
 Often, more recent periods are given higher weights
than periods farther in the past.

18-14
Selecting Weights
 Experience and/or trial-and-error are the simplest
approaches.
 The recent past is often the best indicator of the
future, so weights are generally higher for more
recent data.
 If the data are seasonal, weights should reflect this
appropriately.

18-15
Exponential Smoothing
 A weighted average method that includes all past data in the
forecasting calculation
 More recent results weighted more heavily
 The most used of all forecasting techniques
 An integral part of computerized forecasting
 Well accepted for six reasons
1. Exponential models are surprisingly accurate.
2. Formulating an exponential model is relatively easy.
3. The user can understand how the model works.
4. Little computation is required to use the model.
5. Computer storage requirements are small.
6. Tests for accuracy are easy to compute.
18-16
Exponential Smoothing Model

18-17
 Exponential Smoothing Example
Week Demand Forecast
1 820 820
2 775 820
3 680 811
4 655 785
5 750 759
6 802 757
7 798 766
8 689 772
9 775 756
10 760

18-18
Exponential Smoothing – Effect of Trends

18-19
Example – Exponential Smoothing with Trend
Adjustment

18-20
Choosing Alpha and Delta


18-21
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 that assumes the
relationship between the variables can be explained with
a straight line.
Y = a + bt

18-22
Example 18.2 – Least Squares Method

Quarter Sales Quarter Sales

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

18-23
 Example 18.2 – Calculations

1 600 600 1 360,000 801.3

2 1,550 3,100 4 2,402,500 1,160.9

3 1,500 4,500 9 2,250,000 1,520.5

4 1,500 6,000 16 2,250,000 1,880.1

5 2,400 12,000 25 5,760,000 2,239.7

6 3,100 18,600 36 9,610,000 2,599.4

7 2,600 18,200 49 6,760,000 2,959.0

8 2,900 23,200 64 8,410,000 3,318.6

The forecast is extended to periods 13-16
9 3,800 34,200 81 14,440,000 3,678.2

10 4,500 45,000 100 20,250,000 4,037.8

11 4,000 44,000 121 16,000,000 4,397.4

12 4,900 58,800 144 24,010,000 4,757.1

Sum 78 33,350 268,200 650 112,502,500

18-24
Regression with Excel
 Microsoft
Excel
includes
data analysis
tools, which
can perform
least squares
regression
on a data set.

18-25
Time Series Decomposition
 Chronologically ordered data are 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.

18-26
Seasonal Variation
 Seasonal variation may be either additive or
multiplicative (shown here with a changing trend).

18-27
 Determining Seasonal Factors :
Simple Proportions Example 18.3

 The seasonal factor (or index) is the ratio of the

amount sold during each season divided by the
average for all seasons.
Season Past Sales Average Seasonal
Sales for Factor
Each Season
Spring 200

Summer 350

Fall 300

Winter 150

Total 1000
18-28
Example 18.3 Continued

Expected Average Seasonal Next

Demand Sales for Factor Year’s
for Each Seasonal
Next Year Season Forecast
(1,100y4)
Spring 275 X 0.8 = 220
Summer 275 X 1.4 = 385
Fall 275 X 1.2 = 330
Winter 275 X 0.6 = 165
1100

18-29
Decomposition Using Least Squares Regression

1. Decompose the time series into its components.

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

18-30
Decomposition – Steps 1 and 2

Using the data for periods 1-12, apply time series analysis
(decomposition, linear regression, trend estimate & seasonal
indices) to forecast for periods 13-16

18-31
 Decomposition – Steps 3 and 4
 Develop a least squares regression line for the deseasonalized
data.
 Project the regression line through the period of the forecast.

Regression Results:
Y = 555.0 + 342.2t

Forecast for
periods 13-16

18-32
 Decompostion – Step 5
 Create the final forecast by adjusting the regression
line by the seasonal factor.

Period Quarter Y from Regression Seasonal Factor Forecast (F x

Seasonal Factor
13 I 5,003.5 0.82 4,102.87
14 II 5,345.7 1.10 5,880.27
15 III 5,687.9 0.97 5,517.26
16 IV 6,030.1 1.12 6,753.71

18-33
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
18-34
 Forecast Error Measurements
 Ideally, MAD will be zero  MAPE scales the forecast error to
(no forecasting error). the magnitude of demand.
 Larger values of MAD
indicate a less accurate
model.
 Tracking signal indicates whether
forecast errors are accumulating
over time (either positive or
negative errors).

18-35
Computing Forecast Error

18-36
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.

18-37
Multiple Regression Techniques
 Often, more than one independent variable may be
a valid predictor of future demand.
 In this case, the forecast analyst may utilize
multiple regression.
 Analogous to linear regression analysis, but with
multiple independent variables.
 Multiple regression supported by statistical software
packages.

18-38
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
18-39
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.

18-40
 CPFR Steps

Creation of a
Development Inventory
front-end Joint business Sharing
of demand replenishmen
partnership planning forecasts
forecasts t
agreement

18-41
Principles
 Forecasting is a fundamental step in any planning
process.
 Forecast effort should be proportional to the
magnitude of decisions being made.
 Web-based systems (CPFR) are growing in
importance and effectiveness.
 All forecasts have errors – understanding and
minimizing this error is the key to effective
forecasting processes.

18-42
Forecasting
RULES OF FORECASTING
 The forecast is always wrong.
 The longer the forecast horizon, the worse the
forecast.
 Aggregate forecasts are more accurate.
Utility of Forecasting
 Part of the available tools for a manager
 Despite difficulties with forecasts, it can be used
for a variety of decisions
 Number of techniques allow prudent use of
forecasts as needed
Techniques
 Judgment Methods
 Sales-force composite
 Experts panel
 Delphi method
 Market research/survey
 Time Series
 Moving Averages
 Exponential Smoothing
 Trends
 Regression
 Holt’s method
 Seasonal patterns – Seasonal decomposition
 Trend + Seasonality – Winter’s Method
 Causal Methods
The Most Appropriate
Technique(s)
 Purpose of the forecast
 How will the forecast be used?
 Dynamics of system for which forecast will be
made
 How accurate is the past history in predicting the
future?
SUMMARY
 Matching supply with demand a major challenge
 Forecast demand is always wrong
 Longer the forecast horizon, less accurate the
forecast
 Aggregate demand more accurate than disaggregated
demand
 Need the most appropriate technique
 Need the most appropriate inventory policy