+ All Categories
Home > Documents > Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting...

Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting...

Date post: 23-Dec-2015
Category:
Upload: lynne-byrd
View: 217 times
Download: 0 times
Share this document with a friend
Popular Tags:
79
“Education in Pursuit of Supply Chain Leadership” Chapter 5 Chapter5 dp&c Forecasting in the Supply Chain Environment
Transcript
Page 1: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-1

“Education in Pursuit of Supply Chain Leadership”

Chapter 5

Chapter5dp&c

Forecasting in the Supply Chain Environment

Page 2: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-2

• Define the elements of forecasting

• Describe the levels of forecasting

• Detail the qualitative techniques of forecasting

• Detail the quantitative techniques of forecasting

• Describe the basic quantitative forecasting techniques

• Detail the basics of time-series analysis

• Review decomposition of a time series

• Work with simple associative models

• Understand the coefficient for regression

Learning Objectives

Page 3: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-3

Learning Objectives (cont.)

• Perform multiple variable associate forecasting

• Review alternatives planning method to using forecasting methods

• Measure forecast error

• Describe why forecasts fail

Page 4: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-4

Inventory Management Basics

Chapter 5Forecasting in the Supply Chain

Environment

Forecasting – An Overview

Page 5: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-5

Defining Forecasting

An objective estimate of future demand attained by projecting the pattern found in the events of the past into the future. It is primarily a calculative rather than an

intuitive management process

Page 6: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-6

Why is Forecasting Important?

• Forecasts enable managers to deal with the underlying uncertainties that reside at the core of demand and supply

• Forecasting permits firms to establish performance measurements for customer service, plan the level of total inventory investment, choose between alternative operating strategies, and develop assumptions about the ability of the business to respond to the future needs of the marketplace

• Forecasts can improve enterprise profitability, productivity, and customer service and ensure competitive advantage

• Forecasting helps businesses eliminate waste in of excess inventory, reduce lost sales and expediting and control costs for plant size, labor, equipment, and transportation

Page 7: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-7

• A forecast is useful in creating, anticipating, and managing change within an organization

• Finally, the communication of accurate and timely forecasts enables companies to construct agile and scalable capabilities aligned with the requirements of the customer

Why is Forecasting Important? (cont.)

Page 8: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-8

Universal Principles of Forecasting

• Forecasts are usually wrong

• Forecasts are most useful when accompanied by a method for measuring forecast error

• Forecasts are more accurate for groups of products

• Forecasts are more accurate for near-term planning

Page 9: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-9

Forecasting Design Issues

• Time horizon

• Level of aggregate detail

• Size of the forecastable database

• Forecast control

• Constancy

• Selection of forecasting models

• Alignment with planning procedures

Page 10: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-10

Forecasting Levels

Planning Horizon Focus

STRATEGICFORECASTS

TACTICALFORECASTS

OPERATIONSFORECASTS

IMMEDIATE-RANGEFORECASTS

ANNUAL – 1-10 years

MONTHLY – 3-12 Months

WEEKLY – 1-52 Weeks

DAILY – 1-365 Days

Financial Goals and Objectives

Product Families,Facilities Planning

Finished Goods, Scheduling, Rough-Cut Capacities

Manufacturing/ Purchasing, AR, AP, Shipments

Page 11: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-11

Forecast Development Steps

1 Define the purpose of the forecast

2 Select the appropriate forecasting models

3 Prepare the statistical components

4Ensure the interaction of the firm’s functional area managers

5 Execute the forecast

6 Track and maintain the forecast through timely and accurate feedback

Page 12: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-12

Elements in the Choice of Forecast Models

System Dynamics

Determine the dynamics of the various systems within the organization to be forecasted

Technology Elements

Requirements for computational power, sophistication forecasting functionality, integration with backbone data warehouses, and electronic interoperability with supply chain partners

Time HorizonSelection of the proper time horizon is critical in the de-termination of the length of time to be considered by the forecast

Data

Understand the nature of the data required and the availability and accuracy of that data within the organization and outside in the supply chain

Page 13: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-13

Elements in the Choice of Forecast Models

CostSelection of forecasting models are based on a simple correlation of forecast cost and the value of the forecasted decision

Accuracy

Data must be accurate if forecast output is to be meaningful and forecasters must understand how the data have been obtained, verified, recorded, and transmitted

Ease of Use and Simplicity

Many planners make the mistake of over-complicating their forecasts by trying to use complex mathematical formulas to solve relatively simple forecasting problems

Page 14: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-14

Inventory Management Basics

Forecasting Techniques

Chapter 5Forecasting in the Supply Chain

Environment

Page 15: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-15

Forecasting data sources based on historical patterns of the data itself from the company data

Forecasting data sources based on external patterns from information outside the company

Internal (Intrinsic)

External (Extrinsic)

Forecasting Data Sources

Page 16: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-16

Internal Factors

Product promotion

Planned advertising or marketing efforts

Planned pricing discounts

Product substitution

Product life cycles

Lead time of product replenishment

Management judgment

Intra-company demand

External Factors

State of the economy

Actions of competitors

Economic cycles

Product seasonality

Sales trends

Random fluctuations

Changing customer preferences and demands

Impact of technology breakthroughs

Examples of Internal and External Factors

Page 17: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-17

General Forecasting Techniques

Based on intuitive or judgmental evaluation

Based on computational projection of a numeric relationship

Qualitative Techniques

Quantitative Techniques

Page 18: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-18

Forecasting Categories

Quantitative Techniques

• Simple average• Moving average• Exponential

smoothing• Time series

decomposition

• Regression• Multiple regression• Historical analogy• Leading indicator• Econometric

Qualitative Techniques

• Expert opinion• Sales force estimate• Pyramid forecasting• Panel consensus• Market research• Delphi technique• Visionary forecast• Product life cycle

analysis

Time Series (Intrinsic)

Associative(Extrinsic)

Judgmental

Page 19: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-19

Qualitative Forecast Development

QuantitativeAnalysis

QualitativeForecast

BehavioralData

QualitativeJudgment

ForecastingObjective

Time series and

cause and effect data

Page 20: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-20

Qualitative Forecast Techniques

IndependentJudgment

Executive/Management

Judgment

Market Research

Sales ForceEstimates

HistoricalAnalogy

• Expert opinion• Visionary forecast

• Focus group• Survey

• Sales force composite

• Product life cycle analysis

• Panel consensus• Delphi technique• Pyramid

Page 21: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-21

Qualitative Forecasting Advantages

Initial quantitative data are missing

Demand patterns and relationships are highly unstable

Strong need exists for executive and expert insight

Long-term forecasting needs behavioral insight from market research

Sales forecasts need to be assembled quickly

Qualitative Forecasting Disadvantages

Bias and overconfidence

Incomplete supporting documentation and data

Not practical when organizations have thousands of stockkeeping units

Adverse effect of peer pressure in group decision making

Qualitative Forecasting Advantages and Disadvantages

Page 22: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-22

Basic Quantitative Forecasting Techniques

Simple average

Year-to-date average

Moving average

Weighted moving average

Exponential smoothing

Time series decomposition

Page 23: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-23

Averages

Formula: Ft = (Dt-1 + Dt-2) / 2

Formula: Ft = (Dt-1 + Dt-2 + . . . . + Dt-n) / n

Formula: Ft = (Dt-1 + Dt-2 + Dt-3) / 3

Simple AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 350 395 390 378 0 0 0 0 0

Year-to-Date AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 350 372 370 374 0 0 0 0 0

Simple AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 350 395 390 378 400 0 0 0 0 0

Year-to-Date AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 350 372 370 374 380 0 0 0 0 0

3 Period Moving AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 372 385 390 0 0 0 0 0

3 Period Moving AveragePeriods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 372 385 390 388 0 0 0 0 0

Page 24: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-24

Averages (cont.)

Formula: Ft = w1Dt-1 + w2Dt-2 + . . . . + wnDt-n

3 Period Weighted AverageWeight 20% 30% 50%Periods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 385 382 388 0 0 0 0 0

3 Period Weighted AverageWeight 20% 30% 50%Periods 1 2 3 4 5 6 7 8 9 10 11 12Demand 325 375 415 365 390 410Forecast 385 382 388 395 0 0 0 0 0

Page 25: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-25

Exponential Smoothing

Formula:

• Use actual demand and the forecast from the previous period

• Assign a smoothing constant (a alpha) to the previous period demand

• Calculate the weighted average of the previous period demand and forecast

• Achieve practical balance of forecast accuracy and minimum standard deviation of forecast error

New forecast = (a)(previous period demand) + (1-a)(previous period forecast)

Page 26: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-26

Calculating the Alpha (a) Factor

= 2(𝑛+ 1) Calculating the alpha (a) based on the number of desired periods

Calculating the alpha (a) from the forecast deviation

𝑎𝑙𝑝ℎ𝑎 = 𝑀𝑒𝑎𝑛 𝐸𝑟𝑟𝑜𝑟𝑀𝐴𝐷

Page 27: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-27

Forecast Technique Comparison

Page 28: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-28

Inventory Management Basics

Page 29: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-29

Inventory Management Basics

Time-Series Analysis

Chapter 5Forecasting in the Supply Chain

Environment

Page 30: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-30

Time Series Examples

I. Weekly Demand for Sump Pump #401-325-01Date Jan 7 Jan 14 Jan 21 Jan 28 Feb 4 Feb 11Demand 21 28 30 26 24 33

II. Monthly Sales Forecast of Submersible Sump PumpsMonth Jan Feb March April May JuneDemand 101 118 145 170 200 250

III. Quarterly Forecast of Shipped DollarsQuarter 1st 2nd 3rd 4th 1st 2ndDollars $235,000 $244,000 $310,000 $375,000 $421,000 $503,000

IV. Yearly Sales of ABC Company

Year 1997 1998 1999 2000 2001 2002Sales (M) $145 $148 $151 $156 $162 $165

Page 31: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-31

Types of Time SeriesForecastVariable Mean

Horizontal

ForecastVariable Mean

Quarters Seasonal

ForecastVariable Mean

MonthsRandom

ForecastVariable Mean

MonthsTrend

ForecastVariable Mean

YearsCyclical

Page 32: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-32

Trend Quantity Calculation

Base forecast calculation. The base forecast can use a moving average or exponential smoothing

Trend quantity calculation. The trend calculation uses the current and prior period base forecast and the assigned beta factor (b). The equation is expressed as follows:

Tt = b (FBt – FBt-1) + (1 - b)Tt-1 where

FB is the forecast baseT is the trend

Forecast calculation. Once the trend quantity is determined, it is added to the base forecast to determine the trended forecast. The forecast is extrapolated into the future by adding the trend quantity to each period’s trended forecast

Page 33: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-33

Exercise 5-1 Trend Forecast

Objective: Calculate a trend quantity forecast using three period moving average

Data: b factor = .3 Period DemandJanuary Year1 100February 109March 119April 131May 140June 148July 160August 175

Solution: 1. Base forecast calculation

Fi = (148 + 160 + 175) / 3 = 161

2. Trend quantity calculationTQ = ((161 – 149.33) x .3) + .7 x 15.39)) = 14.27

Page 34: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-34

Exercise 5.1 Trend Forecast (cont.)

Solution: 3. Trend forecast calculation

TF = 161 + 14.27 = 175.27

4. Trend forecast extrapolationFormula:

TF+n = TF+1 +TQ+1 . . . TF+n + TQ+1

TF+1 = 175.27 + 14.27 = 189.54

Page 35: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-35

Exercise 5.2 Trend Forecast

Objective: Calculate a trend quantity forecast using exponential smoothing

Data: a factor = .3 b factor = .3

Period DemandJanuary Year1 100February 109March 119April 131May 140June 148July 160August 175

Solution: 1. Base forecast calculation

Fi = .3 x 175 + (1- .3) x 159.27 = 163.99

2. Trend quantity calculationTQ = ((163.99 –150.58) x .3) + (.7 x 8.69) = 10.11

Page 36: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-36

Exercise 5.2 Trend Forecast (cont.)

Solution: 3. Trend forecast calculation

TF = 163.99 + 10.11 = 174.10

4. Trend forecast extrapolationFormula:

TF+n = TF+1 +TQ+1 . . . TF+n + TQ+1

TF+1 = 1174.10 + 10.11 = 184.21

Page 37: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-37

Trend Projection – Least Squares

Objective: Calculate a trend quantity forecast using trend projection

Data: Trend review

Period DemandJanuary Year1 100February 109March 119April 131May 140June 148July 160August 175

507090

110130150170190

y-Variable

y-Variable

Page 38: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-38

Trend Projection – Least Squares (cont.)

Linear least squares regression:

Calculating the slope of the regression line:

Y = a +bxwhereY = dependent variable computer by the equationy = the actual dependent variable data point (used below)a = the Y-interceptb = slope of the trend linex = time period

wherex0 = average of the value of xy0 = average of the value of yn = number of data points

Page 39: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-39

Trend Projection – Least Squares (cont.)

Calculating the y-intercept a:

a = y̅ - bx ̅

Page 40: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-40

Exercise 5-3 Trend Projection

Objective: Calculate a trend using trend projection

Data:

Period Time Period (x ) Demand (y ) x² xy

January Year 1 1 100 1 100

February 2 109 4 218

March 3 119 9 357

April 4 131 16 524

May 5 140 25 700

June 6 148 36 888

July 7 160 49 1,120

August 8 175 64 1,400

SUM 36.00 1,082.00 204 5,307

∑x/n 4.50 135.25 ∑y/n

Page 41: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-41

Exercise 5-3 Trend Projection (cont.)

1. Calculate b or slope of the regression line:

2. Calculate a or the y-axis intercept:135 – 1043(4.5) = 88.32

3. Forecast for September:88.32 + (10.43*9 or the next period) = 182.18

4. Trend projection results: Period ForecastSeptember 182.18October 192.61November 203.04December 213.46January Year 2 223.89February 234.32

Page 42: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-42

Excel Trend Calculation

Page 43: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-43

Exercise 5-4 Seasonal Forecast

Objective: Perform a seasonal forecast calculation

Data:Past Demand 1 2 3Year 1-1 Qtr 1-2 Qtr 1-3 Qtr 1-4 Qtr 2-1Qtr 2-2 Qtr 2-3Qtr 2-4 Qtr 3-1Qtr 3-2Qtr 3-3 Qtr 3-4 qtrDemand 150 240 370 455 160 255 390 505 170 270 420 560

Summary Total Avg

Yrs 1,2,3 Ist Qtr 480 160Yrs 1,2,3 2nd Qtr 765 255Yrs 1,2,3 3rd Qtr 1180 393Yrs 1,2,3 4th Qtr 1520 507Totals 3,945 329

Season Index

0.4870.7761.1961.541

4

Page 44: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-44

Exercise 5-4 Seasonal Forecast (cont.)

Solution:Past Demand 1Year 1-1 Qtr 1-2 Qtr 1-3 Qtr 1-4 QtrDemand 150 240 370 455

2Year 2-1 Qtr 2-2 Qtr 2-3 Qtr 2-4 QtrDemand 160 255 390 505

3Year 3-1 Qtr 3-2 Qtr 3-3 Qtr 3-4 QtrDemand 170 270 420 560

New Forecast 4Year 1 Qtr 2 Qtr 3 Qtr 4 QtrDemand 183 291 449 578

Summary Total Avg Season Index Forecast (Yr 4)Yrs 1,2,3 Ist Qtr 480 160 0.487 1,500Yrs 1,2,3 2nd Qtr 765 255 0.776 Avg Forecast per QuarterYrs 1,2,3 3rd Qtr 1180 393 1.196 375Yrs 1,2,3 4th Qtr 1520 507 1.541

Totals 3,945 329 4

Page 45: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-45

Exercise 5-5 Seasonal Forecast w/Trend

Objective: Perform a seasonal forecast calculation with trend

Solution:1. Determine the seasonal index

Seasonal IndexPast Demand 1 2 3Year 1-1 Qtr 1-2 Qtr 1-3 Qtr 1-4 Qtr 2-1 Qtr 2-2 Qtr 2-3 Qtr 2-4 Qtr 3-1 Qtr 3-2 Qtr 3-3 Qtr 3-4 QtrDemand 150 240 370 455 160 255 390 505 170 270 420 560

Summary Total Avg Season Index

Yrs 1,2,3 Ist Qtr 480 160 0.487Yrs 1,2,3 2nd Qtr 765 255 0.776Yrs 1,2,3 3rd Qtr 1180 393 1.196Yrs 1,2,3 4th Qtr 1520 507 1.541

Totals 3945 329 4.00

Page 46: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-46

Trend Calculation

Quarters (x ) Quarter Demand (y )Seasonal

IndexDeseasonalized

Demand (yd )x² xyd

1 1-1 Qtr 150 0.487 308.203 1 308.2032 1-2 Qtr 240 0.776 309.412 4 618.8243 1-3 Qtr 370 1.196 309.248 9 927.7444 1-4 Qtr 455 1.541 295.226 16 1,180.9055 2-1 Qtr 160 0.487 328.750 25 1,643.7506 2-2 Qtr 255 0.776 328.750 36 1,972.5007 2-3 Qtr 390 1.196 325.964 49 2,281.7488 2-4 Qtr 505 1.541 327.669 64 2,621.3499 3-1 Qtr 170 0.487 349.297 81 3,143.672

10 3-2 Qtr 270 0.776 348.088 100 3,480.88211 3-3 Qtr 420 1.196 351.038 121 3,861.41912 3-4 Qtr 560 1.541 363.355 144 4,360.263

78.00 Totals 3,945 3,945.000 650 26,401.258

∑x/n 6.5 ∑y/n 328.75 ∑yd/n 328.75

Exercise 5-5 Seasonal Forecast w/Trend

2. Deseasonalize the original demand

3. Combine the quarter (x) and the deseasonalized demand (yd) found in column (xyd).

Page 47: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-47

Exercise 5-5 Seasonal Forecast w/Trend

4. Compute a least squares regression line for the deseasonalized demand (a axis and b slope)

𝑏= 𝛴𝑥𝑦𝑑− 𝑛𝑥�̅��̅�𝛴𝑥² − 𝑛𝑥²̅ a = y̅d - bx̅ b or slope 5.31a or y -axis 294.26

5. Calculate the forecast for year 4

Quarter Period Slope Intercept Seasonal Factor Forecast

1 13 5.306 294.261 0.487 176.792 14 5.306 294.261 0.776 285.873 15 5.306 294.261 1.196 447.294 16 5.306 294.261 1.541 584.35

Page 48: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-48

Inventory Management Basics

Associative (Correlation) Forecasting

Chapter 5Forecasting in the Supply Chain

Environment

Page 49: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-49

Associative (Correlation) Technique

Also known as explanatory or extrinsic forecasting,

these techniques seek to predict the future by

using additional associated data beyond the time

series data recorded for a specific occurrence (for

example, weekly sales of an item). The idea

behind the method is to leverage other patterns of

events occurring in the marketplace beyond

historical data to predict more precisely the course

of future demand.

Page 50: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-50

Exercise 5-6 Simple Associative Model

Objective: Perform a simple associative model forecast for quarterly sales

Data:Correlating housing starts with quarterly pump sales

QuarterNumber of Housing Starts (0,000 units)

(x )

Sales (US$000,000) (y)

1 1 2.02 3 3.03 2 2.44 3 3.15 4 3.76 6 4.57 5 4.08 4 3.5

Page 51: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-51

Exercise 5-6 Simple Associative Model

Solution:1. Perform a last squares regression

QuarterNumber of Housing Starts (0,000 units)

(x )

Sales (US$000,000) (y)

x² xy

1 1 2.0 1 2.0002 3 3.0 9 9.0003 2 2.4 4 4.8004 3 3.1 9 9.3005 4 3.7 16 14.8006 6 4.5 36 27.0007 5 4.0 25 20.0008 4 3.5 16 14.000

∑x 28.00 ∑x² 116

∑y 26.200 ∑xy 100.900

b or slope 0.511a or y -axis 1.486

Page 52: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-52

Exercise 5-6 Simple Associative Model

2. Calculate the next quarter forecast.

3. Formula: y = (b*x) + a where: b = slope

a = y-axisx = number of housing starts

4. Forecast calculation:

Housing Starts OptionsOpt1 Opt 2 Opt 3 Opt 44.5 5.000 5.500 6.000

ForecastOption ForecastOpt 1 3.79Opt 2 4.04Opt 3 4.30Opt 3 4.55

Page 53: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-53

Exercise 5-7 Correlation Coefficient

Objective: Perform a correlation coefficient calculation

Data:

QuarterNumber of Housing Starts

(0,000 units) (x )Sales (US$000,000)

(y)x² xy y ²

1 1.00 2.0 1 2.000 4.0002 3.00 3.0 9 9.000 9.0003 2.00 2.4 4 4.800 5.7604 3.00 3.1 9 9.300 9.6105 4.00 3.7 16 14.800 13.6906 6.00 4.5 36 27.000 20.2507 5.00 4.0 25 20.000 16.0008 4.00 3.5 16 14.000 12.250

∑x 28.00 ∑x² 116 ∑y ² 90.560

∑y 26.20 ∑xy 100.900

Page 54: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-54

Exercise 5-7 Correlation Coefficient (cont.)

Solution: Formula. r = coefficient of correlation

𝑟 = 𝑛∑𝑥𝑦− ∑𝑥∑𝑦ඥሾ𝑛∑𝑥2 − ሺ∑𝑥ሻ2ሿ[𝑛∑𝑦² − (∑𝑦)²]

𝑟 = ሺ8ሻሺ100.9ሻ−ሺ28ሻ(26.2)ඥሾሺ8ሻ(116) − ሺ28ሻ2ሿ[(8)(90.56) − (26.2)²]

= 807.2− 733.6ඥ(144)(38.04) = 73.6ξ5477.76

= 73.674.0118 = .9944

Page 55: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-55

Exercise 5-8 Multiple Variable Associative Forecast

Objective: Perform a multiple variable associative forecast

Data:

Page 56: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-56

Exercise 5-8 Multiple Variable Associative Forecast (cont.)

Solution: Use Excel functions SLOPE and INTERCEPT

CoefficientsSales 3.144444444Interest Rates -0.333333333Housing Starts 0.344444444

1. Generate coefficients

Page 57: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-57

Exercise 5-8 Multiple Variable Associative Forecast (cont.)

2. Generate forecast

If the interest rate is 2.3 and housing starts is 5.0, then the new sales forecast is computed as:

3.144+ (-0.3333 x 2.3) + (0.3444 x 5.0) = 4.10

Page 58: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-58

Inventory Management Basics

Alternative Forecasting

Methods

Chapter 5Forecasting in the Supply Chain

Environment

Page 59: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-59

Alternative Forecasting Methods

Supply demand smoothing

Ignores traditional forecasting and determines demand by focusing on making the supply chain more flexible and agile to capture demand as it is actually occurring, thereby linking fulfillment functions directly with customer requirements as they occur.

Supply chain demand

smoothing

The objective of forecasters employing this model is to actively pursue channel management techniques that smooth current demand, rather than depend on traditional passive forecasting tools that accept demand patterns as given

Customer collaboration

By utilizing interoperable technologies and integrative practices, this model seeks to establish an open, real-time sharing of demand-related information between customers and suppliers

Page 60: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-60

Technology Tools for Dynamic Forecasting

Electronic data interchange (EDI). Perhaps the oldest alternative technique to classical forecasting, EDI enables the company-to-company transmission of demand data

Quick response (QR). Originating in the retail sphere, QR is a technology-driven cooperative effort between customers and suppliers to improve channel inventory management by closely matching consumer buying patterns to merchandise availability

Point of sales (POS) and scanning tools. Capturing sales data as it happens and transmitting it through electronic data interchange (EDI) or the Internet enables forecasters to better manage short- and medium-term demand patterns

Page 61: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-61

Technology Tools for Dynamic Forecasting

Vendor-managed inventories (VMI). In this model, the full responsibility for the management of inventories is turned over to the supplier who directly manages the entire resupply process

ERP-to-ERP integration. By linking ERP systems together, customers can directly load their demand requirements into the ERP systems of their suppliers

Supply chain event management (SCEM). SCEM can be described as a computer application layer that standardizes and transmits demand information as it flows between channel trading partners.

Page 62: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-62

CPFR – Definition

A collaboration process whereby supply chain

trading partners can jointly plan key supply

chain activities from production and delivery of

raw materials to production and delivery of final

products to end customers. Collaboration

encompasses business planning, sales

forecasting, and all operations required to

replenish raw materials and finished goods

APICS Dictionary

Page 63: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-63

CPFR – Implementation Steps

In the first phase, two or more members of a supply chain establish a collaborative partnership with the intent of creating the technical and operations management architectures necessary to address the existing gaps impeding the synchronization of critical supply chain information

In phase 2, the CPFR partners agree to share critical demand information detailing what products are going to be marketed, how they are going to be promoted and merchandized, and when sales cycles are to begin.

In phase 3, each partner agrees to implement techniques that provide for the real-time sharing of channel inventory levels, point of sales (POS) transactions, and internal supply chain constraints. Each trading partner is responsible for ensuring continuous forecast and inventory accuracy as well as database update

Page 64: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-64

CPFR – Implementation Steps (cont.)

In the phase four, the generation of a consensus forecast is created and shared by all participating supply chain trading partners detailing what is to be sold, how it will be merchandized and promoted, in which marketplaces, and during what time period

Page 65: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-65

Inventory Management Basics

Chapter 5Forecasting in the Supply Chain

Environment

Managing Forecast Performance

Page 66: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-66

Forecast Performance Dynamics

Shifting customer demand

Forecasters should have in place monitoring and change management techniques that enable them to respond quickly to shifting marketplace priorities

Impact of demand assumptions

The level of forecast accuracy is shaped by the quality and objectiveness of demand assumptions, selection of the appropriate forecasting models, and the overall management of the forecasting process

Forecasts are naturally wrong

Forecasts, by their very nature, will be inaccurate and need to be continuously monitored and a mechanism is in place to replace suboptimal models.

Impact of forecast bias

The magnitude of the variance between cumulative actual demand and the cumulative forecast demand

Pursuit of continuous

improvement

Identify opportunities for continuous improvements in forecasting techniques, as well as in business processes for reducing bias and demand variation

Page 67: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-67

Forecast Monitoring Tools

Forecast error: the deviation of the actual demand from the proposed forecast

Absolute percent of error (APE): the deviation of the actual demand from the proposed forecast expressed as a percentForecast error: the deviation of the actual demand from the proposed forecast

Mean absolute deviation (MAD): the average of the absolute values of the deviation of the variance between actual and forecast demand

Mean absolute percent error (MAPE): the sum of the APE over n periods, divided by the same n periods

Tracking signal: used to alert the forecaster to a signal (a ratio) that a variation between forecast and demand has occurred for several periods in a row in the same direction indicating a bias in the forecast

Page 68: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-68

Calculating Forecast Error

Definition:

The difference between actual demand and forecast demand, stated as an absolute value or as a percentage.

Formula:

Et = Dt – FtwhereE = the forecast errorF = the forecastD = the demandt = time

Example:

E = 125 units (D) – 110 units (F) = 15 units

Page 69: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-69

Absolute Percent of Error (APE)

Definition:

Calculates how much the forecast deviates from the actual demand for a given period by expressing the period forecast error as a percent

Formula:

WhereF = the forecastD = the demandt = time

Example:

𝐴𝑃𝐸= ȁ+325− 300ȁ+325 = 7.7% Forecast = 300 Demand = 325

𝐴𝑃𝐸= ȁ+𝐷𝑡− 𝐹𝑡ȁ+𝐷𝑡

Page 70: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-70

Forecast Bias (Mean Error)

Definition:

Bias indicates a consistent deviation of demand from the mean in one direction (high or low)

Formula:

WhereF = the forecastD = the demandn = periods

Example:

Periods = 12Forecast = 3,600 Demand = 3,666

𝐵𝑖𝑎𝑠= ∑(𝐷𝑛− 𝐹𝑛)𝑛

3,666− 360012 = 6612 = 5.5

Page 71: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-71

Mean Absolute Deviation (MAD)

Definition:

The average of the absolute values of the deviation between actual and forecast demand

Formula:

WhereF = the forecastD = the demandn = periods

Example:

Periods = 12Forecast = 3,600 Demand = 3,600Total absolute error = 118

𝑀𝐴𝐷= σ = 1ȁ+𝐷𝑖 − 𝐹𝑖ȁ+𝑛𝑖 𝑛

𝑀𝐴𝐷= 11812 = 9.83

Page 72: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-72

Mean Absolute Percent Error (MAPE)

Definition:

Monitors the accuracy of the forecast by calculating how much the average percentage of forecast deviates from the actual demand

Formula:

WhereF = the forecastD = the demandn = periods

Example:

1. Subtract the demand from the forecast to arrive at the forecast error for each period

2. Convert the forecast error to an absolute value

𝑀𝐴𝑃𝐸= ∑ቚ𝐷𝑡− 𝐹𝑡𝐷𝑡 ቚ100𝑛

Page 73: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-73

Mean Absolute Percent Error (MAPE) (cont.)

3. In each period, divide the absolute forecast error by the period’s demand

4. Multiply this value by 100 to arrive at the period APE5. Sum all of the period-level APEs6. Divide the sum APE by n periods to arrive at the MAPE percent

Period Forecast DemandForecast

Error (FE)Absolute

Error

Absolute Percent Error

(APE)1 300 299 -1 1 0.33%2 300 305 5 5 1.64%3 300 312 12 12 3.85%4 300 310 10 10 3.23%5 300 294 -6 6 2.04%6 300 315 15 15 4.76%7 300 306 6 6 1.96%8 300 286 -14 14 4.90%9 300 292 -8 8 2.74%

10 300 285 -15 15 5.26%11 300 311 11 11 3.54%12 300 285 -15 15 5.26%

Totals 3,600 3,600 0 118 39.51%

MAPE 3.29%

Page 74: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-74

Tracking Signals

Definition:

used to measure forecast bias and are calculated by dividing the cumulative sum of the forecast errors (also referred to as the running sum of forecast errors, RSFE) by the MAD

Formula:

WhereF = the forecastD = the demand

Example:

1. Subtract the demand from the forecast to arrive at the forecast error for each period

2. Convert the forecast error to an absolute value

𝑇𝑟𝑎𝑐𝑘𝑖𝑛𝑔 𝑆𝑖𝑔𝑛𝑎𝑙 = ∑(𝐷𝑖 − 𝐹𝑖)𝑀𝐴𝐷

Page 75: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-75

Tracking Signals (cont.)

Example:

Forecast = 300 unitsDemand = 302 units

Period Forecast DemandForecast

Error (FE)

Cum FE Error

Absolute Error

MADTracking Signal

1 300 302 2 2 2 2 1.002 300 310 10 12 10 6 2.003 300 290 -10 2 10 7 0.274 300 325 25 27 25 12 2.305 300 328 28 55 28 15 3.676 300 310 10 65 10 14 4.59

Page 76: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-76

Forecast Error Review

Period Demand ForecastForecast

ErrorAbsolute

ErrorBias MAD APE MAPE TS

1 1,0002 1,1003 1,2004 1,050 1,100 -50 50.00 -50.00 50.00 4.76% 4.76% -1.005 900 1,117 -217 216.67 -133.33 133.33 24.07% 14.42% -2.006 1,200 1,050 150 150.00 33.33 138.89 12.50% 13.78% -0.847 900 1,050 -150 150.00 -50.00 141.67 16.67% 14.50% -1.888 800 1,000 -200 200.00 -50.00 153.33 25.00% 16.60% -3.049 1,250 967 283 283.33 38.89 175.00 22.67% 17.61% -1.05

10 1,100 983 117 116.67 9.52 166.67 10.61% 16.61% -0.40Avg. 1,029 1,038 Total Bias -201.59 Total MAPE 98.28%

Forecast Method:

Three-period moving average

Page 77: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-77

Management involvement

Perhaps the foremost reason why forecasts fail is because of a lack of participation by functional management both in the development and in the execution of the forecast in process

Over-sophistication

and cost

Complex statistical techniques that require sophisticated calculations turn forecasting into a "black box" activity that divorces users from the process

CompatibilityForecasts fail when there is a lack of compatibility between the forecasting system and the capabilities of the using organizations

Data AccuracyAlthough it is obvious that the data used by a forecasting technique must be accurate, errors do arise in the data collection process

Why Forecast Fail

Page 78: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-78

Unnecessary items

Often forecasts are developed for items that should not be forecasted. One example is dependent demand item usage

Lack of management

control

Forecasters must be diligent in monitoring the forecast to ascertain the degree of error, when the forecast should be altered, and what parameters should be used to guide forecast adjustment

Why Forecast Fail (cont.)

Page 79: Chapter5 dp&c 5-1 “Education in Pursuit of Supply Chain Leadership” Chapter 5 dp&c Forecasting in the Supply Chain Environment.

Chapter5dp&c5-79

Chapter 5

End of Session

“Education in Pursuit of Supply Chain Leadership”

Chapter5dp&c


Recommended