OPRATIN management assignment ( uploaded all related chapters ) &

Operations Management in the Supply Chain Decisions and Cases Seventh Edition

Chapter 10

Forecasting

© McGraw-Hill Education. All rights reserved. Authorized only for instructor use in the classroom. No reproduction or further distribution permitted without the prior written consent of McGraw-Hill Education.

10-2© McGraw-Hill Education.

Learning Objectives

10.1 Describe why forecasting is important.

10.2 Describe the four common methods of qualitative forecasting.

10.3 Calculate a moving average and exponential smoothing forecast, and explain when they should be used.

10.4 Evaluate forecast accuracy using a variety of methods.

10.5 Carry out causal forecasting.

10.6 Evaluate factors that impact forecasting method selection.

10.7 Explain the benefits and costs of CPFR

10-3© McGraw-Hill Education.

Forecasting for Decision Making

• Forecasting demand for operations output

– Forecasting: what we think demand will be

– Planning: what we think demand should be

– Demand: may differ from sales

• Forecasts are used in all functional areas: marketing, finance, human resources, etc.

• Forecasts are necessary for operations decision areas: process design, capacity planning, inventory management, scheduling

10-4© McGraw-Hill Education.

Use of Forecasting: Operations Decisions

Time Horizon

Accuracy Required

Number of Forecasts

Management Level

Forecasting Method

Process design Long Medium Single or few

Top Qualitative or causal

Capacity planning, facilities

Long Medium Single or few

Top Qualitative or causal

Aggregate planning

Medium High Few Middle Causal and time series

Scheduling Short Highest Many Lower Time series

Inventory management

Short Highest Many Lower Time series

10-5© McGraw-Hill Education.

Use of Forecasting: Marketing, Finance, & HR

Time Horizon

Accuracy Required

Number of Forecasts

Management Level

Forecasting Method

Long-range marketing programs

Long Medium Single or few

Top Qualitative

Pricing decisions Short High Many Middle Time series

New product introduction

Medium Medium Single Top Qualitative and causal

Cost estimating Short High Many Lower Time series

Capital budgeting Medium Highest Few Top Causal and time series

Labor planning Medium Medium Few Lower Qualitative and time series

10-6© McGraw-Hill Education.

‘Qualitative’ Forecasting Methods

• Based on managerial judgment when there is a lack of data available. No specific model.

• Major methods:

– Delphi technique

– Market surveys

– Life-cycles analogy

– Informed judgment (naïve models)

10-7© McGraw-Hill Education.

Time-Series Forecasting

• Components of data:

– Level - average

– Trend - general direction (increasing/descreasing)

– Seasonality - short term recurring cycles

– Cycle - long term business cycle

– Error - random or irregular component

10-8© McGraw-Hill Education.

Moving Average

• Assumes no trend, seasonality, or cycle

• Simple moving average:

tt

Nttt t

AF

N

DDD A

=

+++ =

+

+−−

1

11 ......

• Weighted moving average:

11211 ...... +−−+ ++== NtNtttt DWDWDWAF

10-9© McGraw-Hill Education.

Moving Average Example

Period Actual Demand Forecast

1 10

2 18

3 29

4 - 19

• Compute three period moving average for Period 4 (number of periods is forecaster’s decision)

( ) ( ) 318294F

193101829AF

5

34

++=

=++==

periodfordemandactual

10-10© McGraw-Hill Education.

Time-Series Data (Figure 10.2)

Note: The forecast is smoother as the number of periods in the moving average increases.

10-11© McGraw-Hill Education.

Exponential Smoothing

• The new average is computed from the old average:

( ) 11 -tt AA ααDt −+=

• The value of the smoothing constant (α) is a choice. It determines the extent to which the new forecast weights recent demand (smooths random variation).

( ) 0.2.-0.1usuallyisand1,and0betweenrangesAlpha α

10-12© McGraw-Hill Education.

Simple Exponential Smoothing

• Forecast:

( )ttt FDαF −+=+1tF

F = forecast of demand

D = actual demand

t = time period

• Assumes no trend, seasonality, or cycle

• Note: we are adjusting 1tt FgettoF +

10-13© McGraw-Hill Education.

Exponential Smoothing Example

• The Sept. forecast was 15, but Sept. actual sales were 13. ( ) 0.2.ofalphaUse α – What is the October forecast?

( ) ( )

14.6

forecastOctober

=−=

−+=

−α

+ =

4.015

15132.015

forecastSept.actualSept.

forecastSept.

10-14© McGraw-Hill Education.

Time-Series Data (Figure 10.3)

( ) reduced.isalphaofvalue theassmootherisforecastThe

:Note α

10-15© McGraw-Hill Education.

Forecast Accuracy

In addition to the forecast, firms should estimate forecast accuracy:

• To monitor erratic demand observations or “outliers”

• To determine when the forecasting method is no longer tracking actual demand

• To determine the parameter values that provide the forecast with the least error

10-16© McGraw-Hill Education.

Measures of Forecast Accuracy

• Cumulative sum of forecast errors (CFE)

• Mean square error (MSE)

• Mean absolute deviation (MAD)

• Mean absolute percentage errors (MAPE)

• Tracking Signal (TS)

10-17© McGraw-Hill Education.

Forecast Accuracy: Formulas

t

n

t t

t

n

t t

n

t t

n

t t

n

D

e

n

e

n

e

e

MAD

CFE TSsignalTracking

100

MAPEerrorspercentageabsoluteMean

MADdeviationabsoluteMean

MSEerrorsquareMean

CFEerrorsforecastofsumCumulative

1

1

1

2

1

=

=

=

=

=

∑

∑

∑

∑

=

=

=

=

10-18© McGraw-Hill Education.

Advanced Time-Series Forecasting

• Adaptive exponential smoothing

– ( ) variedistcoefficienSmoothing α

• Mathematical models

– Linear or nonlinear

• Box-Jenkins method

– Requires about 60 periods of past data

10-19© McGraw-Hill Education.

Causal Forecasting Methods

• Cause-and-effect model, using a data set of other variables to predict demand (forecast).

• Examples:

– Use population and location characteristics to forecast restaurant sales.

– Use supply chain data on inventory levels to forecast sales of new generation products such as cell phones.

10-20© McGraw-Hill Education.

Causal Forecasting Models

• The general regression model:

bxay += ∧

• Other forms of causal model:

– Econometric

– Input-output

– Simulation models

10-21© McGraw-Hill Education.

Example of Causal Model ( )

( ) ( )

( ) ( ) 7Periodforforecast128.3472.39738.23F

s000'incomefamilymedianI

salesforecastedF

tyearinsalesactualD

2.3965142bSlope

38.230937aIntercept

128.3437.6

125.46413036.4

123.78612535.7

122.58812435.2

125.22411936.3

123.78612435.7

121.1512034.6

tbaY

7

t

t

t

t

==+=

=

=

=

+=

ttt FDI

10-22© McGraw-Hill Education.

Selecting a Forecasting Method

• Use or decision characteristics

– Scheduling decision? Facility expansion?

– Short range? Long range?

• Data availability

– Quantity and quality

• Data pattern

– Level? Unstable?

10-23© McGraw-Hill Education.

Collaborative Planning, Forecasting, and Replenishment (CPFR)

• Aim is to achieve more accurate forecasts

• Share information across supply chain with customers and suppliers

• Compare forecasts

– If discrepancy, look for reason

– Reach a consensus forecast

• Works best in B2B with few customers (e.g., a small number of large retailers)

10-24© McGraw-Hill Education.

Summary

10.1 Describe why forecasting is important.

10.2 Describe the four common methods of qualitative forecasting.

10.3 Calculate a moving average and exponential smoothing forecast, and explain when they should be used.

10.4 Evaluate forecast accuracy using a variety of methods.

10.5 Carry out causal forecasting.

10.6 Evaluate factors that impact forecasting method selection.

10.7 Explain the benefits and costs of CPFR

© McGraw-Hill Education. All rights reserved. Authorized only for instructor use in the classroom. No reproduction or further distribution permitted without the prior written consent of McGraw-Hill Education.

10-25

End of Presentation