CHAPTER 5 DEMAND FORECASTING Principles of Supply Chain Management: A Balanced Approach Prepared by Daniel A. Glaser-Segura, PhD.

CHAPTER 5DEMAND FORECASTING

Principles of Supply Chain Management:

A Balanced Approach

Prepared by Daniel A. Glaser-Segura, PhD

© 2009 South-Western, a division of Cengage Learning 2

Learning Objectives

You should be able to:– Explain the role of demand forecasting in a

supply chain.– Identify the components of a forecast– Compare & contrast qualitative &

quantitative forecasting techniques– Assess the accuracy of forecasts– Explain collaborative planning, forecasting,

& replenishment

© 2009 South-Western, a division of Cengage Learning 3

Chapter Five Outline• Introduction• Demand Forecasting• Forecasting Techniques

– Qualitative Methods– Quantitative Methods

• Components of Time Series Data• Time Series Forecasting Methods• Forecast Accuracy• Useful Forecasting Websites• Collaborative Planning, Forecasting,

& Replenishment (CPFR)• Software Solutions

© 2009 South-Western, a division of Cengage Learning 4

Introduction• Supply chain members find it important

to manage demand, especially in pull manufacturing environments.

• Suppliers must find ways to better match supply & demand to achieve optimal levels of cost, quality, & customer service to enable them to compete with other supply chains.

• Improved forecasts benefit all trading partners in the supply chain & mitigates supply-demand mismatch problems.

© 2009 South-Western, a division of Cengage Learning 5

Demand Forecasting • A forecast is an estimate of future

demand & provides the basis for planning decisions

• The goal is to minimize forecast error • The factors that influence demand

must be considered when forecasting.• Managing demand requires timely &

accurate forecasts• Good forecasting provides reduced

inventories, costs, & stockouts, & improved production plans & customer service

© 2009 South-Western, a division of Cengage Learning 6

Forecasting Techniques • Qualitative forecasting is based on

opinion & intuition.

• Quantitative forecasting uses mathematical models & historical data to make forecasts.

• Time series models are the most frequently used among all the forecasting models.

© 2009 South-Western, a division of Cengage Learning 7

Forecasting Techniques (Cont.)

Qualitative Forecasting MethodsGenerally used when data are limited, unavailable, or not currently relevant. Forecast depends on skill & experience of forecaster(s) & available information.

Four qualitative models used are:

1. Jury of executive opinion

2. Delphi method

3. Sales force composite

4. Consumer survey

© 2009 South-Western, a division of Cengage Learning 8

Forecasting Techniques (Cont.)

Quantitative Methods• Time series forecasting- based on the

assumption that the future is an extension of the past. Historical data is used to predict future demand.

• Cause & Effect forecasting- assumes that one or more factors (independent variables) predict future demand.

It is generally recommended to use a combination of quantitative & qualitative techniques.

© 2009 South-Western, a division of Cengage Learning 9

Forecasting Techniques (Cont.)

Components of Time SeriesData should be plotted to detect for the following components: – Trend variations: increasing or decreasing– Cyclical variations: wavelike movements that are

longer than a year (e.g., business cycle)– Seasonal variations: show peaks & valleys that

repeat over a consistent interval such as hours, days, weeks, months, years, or seasons

– Random variations: due to unexpected or unpredictable events

© 2009 South-Western, a division of Cengage Learning 10

Forecasting Techniques (Cont.)

Time Series Forecasting Models

Naïve Forecast- the estimate of the next period is equal to the demand in the past period.

Ft+1 = At

Where Ft+1 = forecast for period t+1

At = actual demand for period t

© 2009 South-Western, a division of Cengage Learning 11

Forecasting Techniques (Cont.)

Time Series Forecasting Models (Cont.)Simple Moving Average Forecasting Model. Uses historical data to generate a forecast. Works well when demand is stable over time.

© 2009 South-Western, a division of Cengage Learning 12

Forecasting Techniques (Cont.)

Simple Moving Average (Fig. 5.1)

© 2009 South-Western, a division of Cengage Learning 13

Forecasting Techniques (Cont.)

Time Series Forecasting Models (Cont.)Weighted Moving Average Forecasting Model- based on an n-period weighted moving average, follows:

© 2009 South-Western, a division of Cengage Learning 14

Forecasting Techniques (Cont.)Weighted Moving Average (Fig. 5.2)

© 2009 South-Western, a division of Cengage Learning 15

Forecasting Techniques (Cont.)

Time Series Forecasting Models (Cont.)

Exponential Smoothing Forecasting Model- a type of weighted moving average. Only two data points are needed.

Ft+1 = Ft+(At - Ft) or Ft+1 = At + (1 – ) Ft

Where

Ft+1 = forecast for Period t + 1

Ft = forecast for Period t

At = actual demand for Period t

= a smoothing constant (0 ≤ ≤1).

© 2009 South-Western, a division of Cengage Learning 16

Forecasting Techniques (Cont.)

Exponential Smoothing (Fig. 5.3)

© 2009 South-Western, a division of Cengage Learning 17

Forecasting Techniques (Cont.)

Time Series Forecasting Models (Cont.)

Linear Trend Forecasting Model. The trend can be estimated using simple linear regression to fit a line to a time series.

Ŷ = b0 + b1x

where

Ŷ = forecast or dependent variable

x = time variable

b0 = intercept of the line

b1 = slope of the line

© 2009 South-Western, a division of Cengage Learning 18

Forecasting Techniques (Cont.)Regression (Fig. 5.4)

© 2009 South-Western, a division of Cengage Learning 19

Forecasting Techniques (Cont.)

Cause & Effect ModelsOne or several external variables are identified that are related to demand Simple regression. Only one explanatory variable is used & is similar to the previous trend model. The difference is that the x variable is no longer time but an explanatory variable.

Ŷ = b0 + b1xwhere

Ŷ = forecast or dependent variablex = explanatory or independent variable

b0 = intercept of the line

b1 = slope of the line

© 2009 South-Western, a division of Cengage Learning 20

Forecasting Techniques (Cont.)

Cause & Effect Models (Cont.)Multiple regression. Several explanatory

variables are used to make the forecast.

Ŷ = b0 + b1x1 + b2x2 + . . . bkxk

where

Ŷ = forecast or dependent variable

xk = kth explanatory or independent variable

b0 = intercept of the line

bk = regression coefficient of the independent variable xk

© 2009 South-Western, a division of Cengage Learning 21

Forecast Accuracy

The formula for forecast error, defined as the difference between actual quantity & the forecast, follows:

Forecast error, et = At - Ft

whereet = forecast error for Period t

At = actual demand for Period t

Ft = forecast for Period t

© 2009 South-Western, a division of Cengage Learning 22

Forecast Accuracy (Cont.)

Several measures of forecasting accuracy follow:

– Mean absolute deviation (MAD)- a MAD of 0 indicates the forecast exactly predicted demand.

– Mean absolute percentage error (MAPE)- provides a perspective of the true magnitude of the forecast error.

– Mean squared error (MSE)- analogous to variance, large forecast errors are heavily penalized

© 2009 South-Western, a division of Cengage Learning 23

Forecast Accuracy (Cont.)

Mean absolute deviation (MAD)- a MAD of 0 indicates the forecast exactly predicted demand.

where et = forecast error for period tAt = actual demand for period t;n = number of periods of evaluation

© 2009 South-Western, a division of Cengage Learning 24

Forecast Accuracy (Cont.)

Mean absolute percentage error (MAPE)- provides a perspective of the true magnitude of the forecast error.

where et = forecast error for period tAt = actual demand for period t;n = number of periods of evaluation

© 2009 South-Western, a division of Cengage Learning 25

Forecast Accuracy (Cont.)

Mean squared error (MSE)- analogous to variance, large forecast errors are heavily penalized

Whereet = forecast error for period tn = number of periods of evaluation

© 2009 South-Western, a division of Cengage Learning 26

Forecast Accuracy (Cont.)

Running Sum of Forecast Errors (RSFE) indicates bias in the forecasts or the tendency of a forecast to be consistently higher or lower than actual demand.

n

n

tte

1

2

Running Sum of Forecast Errors, RSFE =

Whereet = forecast error for period t

n

tte

1

© 2009 South-Western, a division of Cengage Learning 27

Forecast Accuracy (Cont.)

Tracking signal determines if forecast is within acceptable control limits. If the tracking signal falls outside the pre-set control limits, there is a bias problem with the forecasting method and an evaluation of the way forecasts are generated is warranted.

Tracking Signal =

MAD

RSFE

MAD

RSFE

© 2009 South-Western, a division of Cengage Learning 28

Useful Forecasting Websites

• Institute for Forecasting Education

www.forecastingeducation.com• International Institute of Forecasters

www.forecasters.org• Forecasting Principles

www.forecastingprinciples.com• Stata (Data analysis & statistical software)

www.stata.com/links/stat_software.html

© 2009 South-Western, a division of Cengage Learning 29

Collaborative Planning, Forecasting, & Replenishment (CPFR)A business practice that combines the intelligence of multiple trading partners in the planning & fulfillment of customer demands.

Links sales & marketing best practices, such as category management, to supply chain planning processes to increase availability while reducing inventory, transportation & logistics costs.

© 2009 South-Western, a division of Cengage Learning 30

CPFR (Cont.) Real value of CPFR comes from sharing of forecasts among firms rather than sophisticated algorithms from only one firm.

Does away with the shifting of inventories among trading partners that suboptimizes the supply chain.

CPFR provides the supply chain with a plethora of benefits but requires a fundamental change in the way that buyers & sellers work together.

© 2009 South-Western, a division of Cengage Learning 31

CPFR (Cont.)

VICS’s CPFR Model with Retailer & Manufacturer tasks (Fig. 5.5)

© 2009 South-Western, a division of Cengage Learning 32

CPFR (Cont.)

CPFR ModelStep 1: Collaboration Arrangement

Step 2: Joint Business Plan

Step 3: Sales Forecasting

Step 4: Order Planning/Forecasting

Step 5: Order Generation

Step 6: Order Fulfillment

Step 7: Exception Management

Step 8: Performance Assessment

© 2009 South-Western, a division of Cengage Learning 33

Software Solutions

Forecasting Software• John Galt

– www.johngalt.com• SAS

– www.sas.com• New Energy Associates

– www.newenergyassociates.com• Business Forecast Systems, Inc.

– www.forecastpro.com

© 2009 South-Western, a division of Cengage Learning 34

Software Solutions (Cont.)

CPFR SoftwareJDA Software Group

www.jda.com

i2 Technologieswww.i2.com

Oraclewww.oracle.com