Faculty Bios
Global Executive MBA in Shanghai
Dr. Tyrone W. Callahan - Associate Professor of Clinical Finance and
Business Economics Finance and Business
Economics
Tyrone Callahan received his PhD in Finance in 1999 from the
University of California, Los Angeles
Tyrone has published papers in journals that include The
Review of Futures Markets and the Journal of Physical
Chemistry.
He is an ad-hoc referee at the Journal of Finance, the Journal
of Financial Markets, and the International Review of Finance.
Tyrone has consulted for Hotchkis and Wiley.
Contact Information
Phone: (213) 740-6498
Fax: (213) 740-6650
Email: tyrone.callahan@marshall.usc.edu
1
Global Executive MBA in Shanghai
Dr. Joseph C. Nunes -- Associate Professor of Marketing
Joseph C. Nunes received his Ph.D. in Marketing and
Behavioral Decision Research in 1998 from the University of
Chicago.
He is widely known for his research on loyalty programs,
status and luxury goods, pricing and consumer and
managerial decision-making.
Prof. Nunes has published numerous papers in the top
marketing journals which include the Journal of Marketing
Research, the Journal of Consumer Research, Marketing Science and the
Journal of Marketing. He has alos published in HBR and other publications
targeted to practitioners.
He currently serves on the Editorial Boards of the Journal of Marketing and
Marketing Science.
He was the recipient of the Marshall School of Business Dean's Award for
Excellence in Research in 2006.
Prof. Nunes has consulted for a variety of companies including Southwest
Airlines, Kampgrounds of America, Nestlè, Abbott Laboratories, BBDO
Advertising, Tribune/Knight Ridder Services
Contact Information
Phone: (213) 740-5044
Fax: (213) 740-7828
Email: jnunes@marshall.usc.edu
2
Overall Schedule
Global Executive MBA in Shanghai
GEMBA VIII
Theme 3A – Sept.29- Oct. 3 2011, Shanghai
JN – Professor Joseph Nunes
TC – Professor Ty Callehan
Time
Thursday
September 29
Breakfast
Friday
September 30
Breakfast
Saturday
October 1
Breakfast
Sunday
October 2
Breakfast
Monday
October 3
Breakfast
8:30-9:45
Mktg 1
JN
Mktg 4
JN
FIN5
TC
FIN 8
TC
Mktg 11
JN
9:45-10:00
Break
Break
Break
Break
Break
10:00-11:15
Mktg 2
JN
Mktg 5
JN
FIN 6
TC
FIN 9
TC
Joint Class
TC and JN
11:15-11:30
11:30-12:45
Break
Mktg 3
JN
Break
Mktg 6
JN
Break
FIN 7
TC
Break
Mktg 9
JN
Break
FIN 10
TC
12:45-1:45
Lunch
Lunch
Lunch
Lunch
Lunch
1:45-3:00
FIN 1
TC
Break
FIN 2
TC
FIN 3
TC
Break
FIN4
TC
Mktg 7
JN
Break
Mktg 8
JN
Mktg 10
JN
Break
G7& G8
Joint Session
FIN 11
TC
Break
Summary and
Evaluations
TC and JN
7:00-8:30
3:00-3:15
3:15-4:30
4:30-4:45
4:45-6:00
In-Class
PUB Session
Grid for Theme 3A, 29 Sept – 3 Oct
3
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Thursday, 29 September
08:30
Marketing Session 1 – Professor Nunes
Topic: Course Introduction & Review
You should have downloaded and installed the Markstrat Online software
before class. This session will include time to ensure software is up and
running on all laptops as well as time to form teams for the simulation.
Required Reading: Johnson, Richard R. and Ron Mentus: “Marketing
Economics”, Darden UV0404
09:45
Break
10:00
Marketing Session 2 – Professor Nunes
Topic: Introduction to Markstrat Online
Required Reading: You should have read the Markstrat Participant Handbook
before arriving on campus. Please review it before class.
10:45
Break
11:30
Marketing Session 3 – Professor Nunes
Topic: Using Perceptual Maps
Required Reading: Wilcox, Ronald T.: “Methods for Producing Perceptual Maps
from Data”, UV0405
Assignment: Exercise 1 – Markstrat Familiarity
12:45
Lunch
13:45
Finance Session 1 – Professor Callahan
Topics: Introduction, Growth and Financial Planning
We will look at the relation between a company’s growth rate and need
for capital. We introduce the concepts of a firm’s internal growth rate and
sustainable growth rate and how each is used to anticipate a firm’s need
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
4
Page: 1 / 10
Global Executive MBA in Shanghai
for capital. Detailed forecasting of (external) funds needed is based on a
pro forma analysis of cash flows.
Required Reading: RWJ Chapters 3 and 26
Discussion Questions:
1. What is a company’s internal growth rate? What can a company do to
improve its internal growth rate?
2. What is a company’s sustainable growth rate? What can a company do
to improve its sustainable growth rate?
3. What are the implications for a company that grows at a rate above its
internal growth rate?
4. What are the implications for a company that grows at a rate above its
sustainable growth rate?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
Due: Turn in the Pre-Theme Questions
15:00
Break
15:15
Finance Session 2 – Professor Callahan
Topic: Growth and Financial Planning – applied
We will use the Clarkson Lumber Company case to illustrate how growth
implies important changes in the type and amount of financing needed by
a firm. This case will also serve as a precursor to our examination of
capital structure choices.
Required Reading: Clarkson Lumber Company case
Discussion Questions:
1. Why has Clarkson Lumber borrowed increasing amounts despite its
consistent profitability?
2. Over recent years, how have the financing needs of the company been
met? Has the company’s financial strength improved or deteriorated?
3. Assuming projected 1996 sales of $5.5M, do you agree with Mr.
Clarkson’s estimate of the company’s loan requirements?
4. Would you advise Mr. Dodge to make the requested loan?
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
5
Page: 2 / 10
Global Executive MBA in Shanghai
5. Would you advise Mr. Clarkson to proceed as planned?
To do before this session: Prepare the case for discussion
Friday, 30 September
08:30
Marketing Session 4 – Professor Nunes
Topic: Conjoint Analysis and Product Design
Required Reading: Wilcox, Ronald T.: “A Practical Guide to Conjoint
Analysis”, UV0406
09:45
Break
10:00
Marketing Session 5 – Professor Nunes
Topic: Breakout Sessions
This session will include time to complete earlier lectures and ensure
comprehension, as well as time to meet as teams with the instructor.
11:15
Break
11:30
Marketing Session 6 – Professor Nunes
Topic: Pricing, Cost Curves and Market Price Evolution
Required Readings:
-
Dhebar, Anirudh: “Price-Quantity Determination”, HBS 9-191-093
-
Steenburgh, Thomas and Jill Avery: “Marketing Analysis Toolkit: Pricing
and Profitability Analysis”, HBS 9-511-028
Due: Exercise 1 – Markstrat Familiarity
Assignment: Exercise 2 – Reading Perceptual Maps
12:45
Lunch
13:45
Finance Session 3 – Professor Callahan
Topic: Introduction to Capital Structure Choices
This session is the first of three sessions in which we examine the
concepts, theories, and evidence concerning corporate capital structure
(i.e., the debt/equity mix, or amount of leverage, used to finance a firm).
We will start by understanding how changes in a firm’s leverage change
the allocation of operating cash flows and operating risk among the firm’s
debt holders and equity holders. Next we will look at how the tax
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
6
Page: 3 / 10
Global Executive MBA in Shanghai
deductibility of interest payments to debt holders can increase firm value
and decrease a firm’s cost of capital.
Required Reading: RWJ Chapters 15 and 16
Discussion Questions:
1. What are the differences in the cash flow rights, control rights, and risk
profile of debt versus equity?
2. How does an increase in corporate leverage change the allocation of
operating risk between debt and equity?
3. Does US (and most other countries’) corporate tax law favor equity or
debt?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
15:00
Break
15:15
Finance Session 4 – Professor Callahan
Topic: Capital Structure – Additional Considerations
In this session we discuss a host of factors that can play a significant role
in firms’ capital structure choices. Potential downsides of debt include the
risk of bankruptcy and a negative impact on the firm’s relations with
customers, employees, and suppliers. Debt can also restrict a firm’s
operating flexibility. Excessive debt can create incentives for
stockholders to under or overinvest relative to the firm’s optimal
investment policy. Potential upsides of debt include its disciplinary effect
on managers and the confidence it can project to financial markets.
Upsides of equity include maintaining financial flexibility and its use as a
form of compensation. Downsides of equity are the potentially negative
signal its use sends to financial markets.
Required Reading: RWJ Chapters 17 and 30
Discussion Questions:
1. Why are financial distress costs generally considered to be much greater
than bankruptcy costs?
2. What are two ways in which equity holders might gain at the expense of
debt holders and what mechanisms can debt holders use to try and
protect themselves?
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
7
Page: 4 / 10
Global Executive MBA in Shanghai
3. How can debt and equity encourage managers to be thrifty and
hardworking?
4. What inferences do markets make about firm value when a firm issues
debt? When a firm issues equity?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
16:30
Break
16:45
In-Class Pub session
Saturday, 1 October
08:30
Finance Session 5 – Professor Callahan
Topic: Capital Structure Choices – Putting It All Together
In this session we will use a case discussion to review and summarize
the important tax, risk, information, and agency consequences of a major
change in a firm’s capital structure policy.
Required Reading: Blaine Kitchenware, Inc.: Capital Structure case
Case Discussion Questions:
1. Do you believe Blaine’s capital structure and payout policies are
appropriate? Why or why not?
2. Should Dubinski recommend a large share repurchase to Blaine’s board?
What are the primary advantages and disadvantages of such a move?
3. Consider the following share repurchase proposal: Blaine will use
$209M of cash from its balance sheet and $50M in new interest-bearing
debt at a rate of 6.75% to repurchase 14.0M shares at a price of $18.50
per share. How would such a buyback affect Blaine? Consider the impact
on, among other things, Blaine’s earnings per share and ROE.
4. As a member of Blaine’s controlling family, would you be in favor of this
proposal? Would you be in favor of it as a non-family shareholder?
To do before this session: Prepare the case for discussion
09:45
Break
10:00
Finance Session 6 – Professor Callahan
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
8
Page: 5 / 10
Global Executive MBA in Shanghai
Topic: Leverage and the Cost of Capital for a Firm or Project/Division
In this session we will learn how to calculate a firm’s or project’s cost of
capital. (They may or may not be the same!) We will discuss the
appropriate cost of debt financing, preferred stock financing, and
common equity financing and how each is commonly derived. We will
also look at the interaction between capital structure choices and the
firm’s cost of capital.
Required Readings:
-
RWJ Chapter 13 (review, especially 13.9)
-
RWJ Chapter 18 (with an emphasis on 18.3, 18.5, and 18.7)
Discussion Questions:
1. Intuitively, what does a stock beta represent?
2. What is the difference between a stock beta, a debt beta, and an asset
beta?
3. What are three primary determinants of a company’s stock beta (i.e.,
what determines the relative exposure of a company stock to economy
wide fluctuations)?
4. Why might a project beta or division beta differ from a company’s stock
beta? Give a real world example of a project with the same beta as the
parent company and a real world example of a project with a different
beta than the parent company.
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
11:15
Break
11:30
Finance Session 7 – Professor Callahan
Topic: Estimating the Cost of Capital – An Example
In this session we will apply what we have learned in the previous
session to the Midland Energy Resources, Inc. case and further refine
our financial analysis skills.
Required Reading: Midland Energy Resources, Inc.: Cost of Capital
Discussion Questions:
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
9
Page: 6 / 10
Global Executive MBA in Shanghai
1. How are Mortensen’s estimates of Midland’s cost of capital used? How, if
at all, should these anticipated uses affect the calculations?
2. Calculate Midland’s corporate WACC. Be prepared to defend your
specific assumptions about the various inputs to the calculations. Is
Midland’s choice of EMRP appropriate? If not, what recommendations
would you make and why?
3. Should Midland use a single corporate hurdle rate for evaluating
investment opportunities in all of its divisions? Why or why not?
4. Compute a separate cost of capital for the E&P and Marketing & Refining
divisions. What causes them to differ from one another?
5. How would you compute a cost of capital for the Petrochemical division?
To do before this session: Prepare the case for discussion
12:45
Lunch
13:45
Marketing Session 7 – Professor Nunes
Topic: Saurer: The China Challenge
Required Readings: Saurer: The China Challenge (A) IMD399
Discussion Questions:
1. Should Saurer enter the market for lower functionality twisting machines
in China? What are the major advantages and disadvantages to doing
this?
2. If Saurer were to introduce the proposed machine, what marketing
strategy would you recommend for the product in China?
3. If Saurer were to not introduce the machine, what changes would you
recommend for the company’s current strategy in China?
Optional Reading: Gadiesh, Orit, Philip Leung and Till Vestring: “The Battle for
China’s Good Enough Market”, HBR Reprint R0709E
Due: Exercise 2 – Reading Perceptual Maps
15:00
Break
15:15
Marketing Session 8 – Professor Nunes
Topic: Breakout Sessions
This session will include time to review Exercise 2 before allowing teams
to work on their first practice decision.
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
10
Page: 7 / 10
Global Executive MBA in Shanghai
Assignment: Make Markstrat Practice Decision
Sunday, 2 October
08:30
Finance Session 8 – Professor Callahan
Topic: Quiz (covering material through Finance session 6)
Required Readings and Discussion Questions: None – study for the quiz.
To do before this session: Study for the quiz.
09:45
Break
10:00
Finance Session 9 – Professor Callahan
Topic: Raising Long-Term Capital
This session provides an opportunity to discuss how firms raise external
capital, what types of capital are acquired and why, and the comparative
costs and benefits of strategic capital acquisition. Topics will include
concepts of market efficiency and the process and costs of underwriting
common stock and debt.
Required Readings: RWJ Chapters 14 and 20
Discussion Questions:
1. Do you believe markets are efficient? Why or why not?
2. What do your views on market efficiency imply about firms’ financial
policies?
3. Think about whatever questions you have regarding the institutional
details or mechanics of stock or bond issuance.
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
11:15
Break
11:30
Marketing Session 9 – Professor Nunes
Please review the results of your practice Markstrat Decision. We will
discuss and I will answer questions.
Topic: Brand Identity & Brand Image
Required Readings: Ward, Scott, Larry Light and Jonathan Goldstine: “What
High-Tech Managers Need to Know about Brands”, HBR Reprint 994II
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
11
Page: 8 / 10
Global Executive MBA in Shanghai
12:45
Lunch
13:45
Marketing Session 10 – Professor Nunes
Topic: Pepsi Case
Required Readings: Handed out in Class
Assignment: Make Decision 1
15:00
Break
15:15
G7 and G8 Joint Session
Monday, 3 October
08:30
Marketing Session 11 – Professor Nunes
Topic: Mattel Case
Required Readings:
-
Mattel and the Toy Recalls (B) Ivey 908M11
-
Donaldson, Thomas “Values in Tension: Ethics Away from Home” HBR
Reprint 96502
Discussion Questions:
1. What went wrong with Mattel’s recall strategy?
2. Who are Mattel’s stakeholders? Who did Mattel cater to in the recall?
3. What values did Mattel exhibit during the recall? How did they affect
Mattel?
4. What should Mattel do right now (at the point in the case) and in the
future?
09:45
Break
10:00
Integrative Session with Professors Nunes and Callahan
Topic: TBD
11:15
Break
11:30
Finance Session 10 – Professor Callahan
Topic: Returning Value to Shareholders (aka Payout Policy)
In recent years, about 40% of firms’ income has been paid out to
shareholders via dividends. But it is also true that many firms pay no
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
12
Page: 9 / 10
Global Executive MBA in Shanghai
dividends at all. What do we know about dividend policy? This session
will explore various methods of returning value to shareholders including
cash dividends, stock repurchases, and extraordinary dividends. Topics
will include dividend policy and personal taxes, factors favoring a high
dividend policy, and some empirical regularities about US corporate
payout policy.
Required Readings: RWJ Chapter 19
Discussion Questions:
1. What are the typical methods used by firms to return cash to
shareholders?
2. What role do taxes play in corporate payout policy?
3. What other factors influence corporate payout policy?
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
12:45
Lunch
13:45
Finance Session 11 – Professor Callahan
Topic: Returning Value to Shareholders – Applied
In this session we will use the General Motors Corporation (D) case to
review payout policy theory and look at various tradeoffs inherent in
choosing a policy. We examine the interaction between payout policy,
investment policy, and capital structure policy.
Required Reading: The General Motors Corporation (D) case
Discussion Questions:
1. What options are Mr. Finnegan and his staff evaluating?
2. What are the pros and cons of each option?
3. What do you recommend General Motors to do?
To do before this session: Prepare the case for discussion
15:00
Break
15:15
Summary and Evaluations – Professors Callahan and Nunes
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
13
Page: 10 / 10
Thursday, September 29
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Thursday, 29 September
08:30
Marketing Session 1 – Professor Nunes
Topic: Course Introduction & Review
You should have downloaded and installed the Markstrat Online software
before class. This session will include time to ensure software is up and
running on all laptops as well as time to form teams for the simulation.
Required Reading: Johnson, Richard R. and Ron Mentus: “Marketing
Economics”, Darden UV0404
09:45
Break
10:00
Marketing Session 2 – Professor Nunes
Topic: Introduction to Markstrat Online
Required Reading: You should have read the Markstrat Participant Handbook
before arriving on campus. Please review it before class.
10:45
Break
11:30
Marketing Session 3 – Professor Nunes
Topic: Using Perceptual Maps
Required Reading: Wilcox, Ronald T.: “Methods for Producing Perceptual Maps
from Data”, UV0405
Assignment: Exercise 1 – Markstrat Familiarity
12:45
Lunch
13:45
Finance Session 1 – Professor Callahan
Topics: Introduction, Growth and Financial Planning
We will look at the relation between a company’s growth rate and need
for capital. We introduce the concepts of a firm’s internal growth rate and
sustainable growth rate and how each is used to anticipate a firm’s need
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
14
Page: 1 / 3
Global Executive MBA in Shanghai
for capital. Detailed forecasting of (external) funds needed is based on a
pro forma analysis of cash flows.
Required Reading: RWJ Chapters 3 and 26
Discussion Questions:
1. What is a company’s internal growth rate? What can a company do to
improve its internal growth rate?
2. What is a company’s sustainable growth rate? What can a company do
to improve its sustainable growth rate?
3. What are the implications for a company that grows at a rate above its
internal growth rate?
4. What are the implications for a company that grows at a rate above its
sustainable growth rate?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
Due: Turn in the Pre-Theme Questions
15:00
Break
15:15
Finance Session 2 – Professor Callahan
Topic: Growth and Financial Planning – applied
We will use the Clarkson Lumber Company case to illustrate how growth
implies important changes in the type and amount of financing needed by
a firm. This case will also serve as a precursor to our examination of
capital structure choices.
Required Reading: Clarkson Lumber Company case
Discussion Questions:
1. Why has Clarkson Lumber borrowed increasing amounts despite its
consistent profitability?
2. Over recent years, how have the financing needs of the company been
met? Has the company’s financial strength improved or deteriorated?
3. Assuming projected 1996 sales of $5.5M, do you agree with Mr.
Clarkson’s estimate of the company’s loan requirements?
4. Would you advise Mr. Dodge to make the requested loan?
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
15
Page: 2 / 3
Global Executive MBA in Shanghai
5. Would you advise Mr. Clarkson to proceed as planned?
To do before this session: Prepare the case for discussion
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
16
Page: 3 / 3
U
UV0404
MARKETING ECONOMICS
generally associated with the finance and operations
functions. Marketing-related decisions often made in
conjunction with these calculations include 1) how
many units of a product do I have to sell before I
begin making a profit 2) how much more do I need to
sell to make a particular change in my marketing plan
profitable.
The purpose of this note is to prepare you
for a series of exercises in making Marketing
Economics calculations. While the note will give you
the basic information on the purpose and structure of
various calculations, the note, by itself, will not help
you develop a facility for doing these calculations.
The series of Exercises in Marketing Economics will
help you to become more proficient in performing
what will eventually seem to be “simple”
calculations. These calculations are not replacements
for spreadsheets or more detailed economic
calculations such as net present value or return on
investment. Rather, they are “back of the envelope”
estimates that can easily be communicated to others
to buttress your arguments about what a company
should or should not do in a decision situation.
There are many other situations where these
calculations are common. Venture capitalists must
determine whether a new business is likely to be
profitable. Part of this evaluation necessarily involves
calculating costs and the sales volume necessary to
cover those costs, a break-even analysis. A company
may want to decide whether a new technology
designed to streamline the production process and
reduce costs is worth the price of acquisition. Again,
in integral part of such analysis would be a breakeven.
The Note and the Exercises will prepare you
for the task of learning how to “see” which
calculations are relevant to a particular case. That is
a task that needs practicing. The concepts of fixed
and variable costs, for example, are simple in theory
compared to practice. Deciding what should be
treated (assumed) as variable or fixed says a lot about
how you intend to manage the business. This note
will not make a cost accountant out of you. The goal
is to start you on a process of becoming fluent in the
practice of marketing economic calculations.1
In the next sections we define different types
of costs and provide examples of break-even analysis
and contribution margin calculations. We also briefly
discuss the impact that selling through channel
partners has on such analysis.
Costs
Costs are often divided into fixed and
variable costs. Whereas total fixed costs remain
constant despite changes in sales or production
volume, the sum total of variable costs changes per
unit manufactured or sold.
These types of marketing calculations are
used both for what we normally think of as
marketing-related decisions as well as decisions more
Total Costs = Total Variable Costs + Fixed Costs
1
Portions of this note were drawn from Lawrence
J. Ring, Derek A. Newton, Neil H. Borden, Jr., and
Paul W. Farris, Decisions in Marketing, 2nd Edition,
(Homewood, IL: BPI/Irwin, 1989), pp. 941-51.
This note was prepared by Richard R. Johnson and Ron Mentus under the supervision of Paul W. Farris, Landmark
Communications Professor of Business Administration, Marian C. Moore, Visiting Associate Professor of Business
Administration, and Ronald T. Wilcox, Associate Professor of Business Administration. Copyright © 2001 by the
University of Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send
an e-mail to sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval
system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying,
recording, or otherwise—without the permission of the Darden School Foundation. Rev. 7/02.
17
-2-
UV0404
Figure 1: Total Contribution.
Variable Costs
Variable costs (VC) are costs that vary with
volume. VC can be expressed on a per unit basis, as
VC per unit; for example, as $3.50 labor and material
cost per unit. As more units are manufactured and
sold, total VC equals VC per unit times units sold.
(Note that selling price, SP, is revenue per unit and is
analogous to variable cost per unit. Therefore, as
additional units are sold, the relevant effect in terms
of a net income of funds into a business is shown by:
SP per unit less VC per unit. This residual amount is
called contribution to fixed costs and profit, and this
concept will be further explained in the following
section.)
As more units are manufactured and sold,
total VC (and revenue) behave as shown in Figure 1.
Fixed Costs
Note: The vertical distance between the revenue and variable
cost lines represents total contribution at a particular sales
volume. Contribution per unit (SP – VC per unit) can also be
found by dividing total contribution at a particular sales volume
by the number of units sold.
Fixed costs (FC) do not vary with volume.
As more units are manufactured and sold, FC (such
as rent and depreciation) remain the same, as shown
in Figure 2.
Figure 2: Fixed Costs
Total Costs
Total costs equal FC + total VC (VC per
unit times units sold). In Figure 3, if the number of
units manufactured and sold were 40, then total
costs would be FC = $100 + total VC = $140 (40
times $3.50), for a total of $240.
The fixed costs and variable costs
categories are usually oversimplifications of most
business situations. For example, some costs may
be semi-variable, such as certain employee costs.
That is, employees may be like a fixed cost in that
you do not want to fire them if you do not have 40
hours a week of work for them at all times.
However, things may become so slack that
employees have to be furloughed, or alternatively,
sales may be so good that they will be used overtime.
Other costs may seem relatively fixed for existing
volume but may go up in large chunks when volume
takes a sharp turn. Certain cost situations can be
confusing and ambiguous.
Notice that our simple graphical
interpretation of costs has assumed that the unit cost
of production is constant. In other words, for the
sake of clear explanation, we have assumed that the
first unit a company produces costs them exactly the
same as the 1000th unit. There are many instances in
which variable costs will decrease, or even increase,
as the number of units produced rises. Common
reasons for a decrease in variable costs include 1)
obtaining better prices on raw materials as the
quantity purchased increases (quantity discounts) 2)
becoming more efficient at the production process as
18
-3-
UV0404
Figure 3: Total Costs
By way of introduction, break-even analysis
is a general term that refers to an easy way to study,
under a number of situations, the interrelationship
that exists between: (1) selling price, variable costs,
and fixed costs; and (2) volume of sales. The general
formula for break-even (BE) is as follows:
Break-even point:
Total Costs = Total Revenues
500
Total costs
($)
400
300
200
Break - even volume (in units) =
100
0
0
40
80
Units
Total Fixed Costs
Total Costs
Total Variable Costs
Total Revenue
experience increases. Also, increases can occur
because of very practical consideration such as 1) the
need to pay employees overtime wages to support an
increase in production output and 2) the need to hold
additional levels of inventory to serve the ramped-up
production process. These are just a few of many
possible reasons. Clearly, if variable costs are
anticipated to change with different levels of
production this needs to factored into your analysis.
Breakeven and Target Volume Analysis
Break-even analysis is a useful analytical
technique to use: (1) when a new product is planned
or (2) when a change in the marketing program is
considered. Break-even analysis is most useful when
it is converted to a percentage of market share,
because then the manager will have one more gauge
of whether the assumptions required to make a
project profitable are reasonable.
New product
When a new product is planned, the
marketer needs to calculate how many units (usually
on a per year basis) will need to be sold to break even
on all costs. That is, how many units will have to be
sold to reach a point where the total revenue just
equals the total costs of the product?
Fixed costs
Unit contribution
Unit contribution = Selling price per unit less
variable costs per unit
Explanations of break-even analysis begin
with a consideration of the concept of “contribution,”
because break-even analysis is taught to train
marketing students to think in terms of contribution –
not in terms of sales. Contribution per unit, which is
the selling price per unit less variable costs per unit,
measures a net income of funds into a business as
additional units are sold. That is, as sales of a
product are made, revenue flows into the business at
a rate that equals selling price (SP) times the unit
volume sold. At the same time, other funds –
variable costs (VC) – are flowing out of the business
with each unit sold. Variable costs (VC) include
such variable marketing costs as salesmen’s
commissions and shipping, and such variable
manufacturing costs as raw materials, direct labor,
and the variable portion of factory overhead. Selling
price (SP) and variable costs (VC) thus are closely
related. Marketers want to sell an additional unit for
a certain price, but they also want to know how much
it costs to sell that additional unit. “Unit contribution
to fixed costs and profit” (or simply “unit
contribution”) is calculated as selling price per unit
(SP) less variable costs per unit (V). Unit
contribution, then, is money that each unit sold
“contributes” to paying fixed costs and providing
profit.
Fixed costs may fall into several categories
and usually include such fixed marketing costs as
advertising, sales office rent, etc., and such fixed
manufacturing costs as property taxes, insurance, and
factory depreciation. (For the latter cost, remember
that it is yearly depreciation and not total investment
that becomes a cost to be subtracted from sales to
compute profits for any given year.) When a new
19
-4-
product (or other business venture) is planned, there
will always be some new fixed costs that will occur
regardless of sales volume. Profits are the dollars left
over after fixed costs and variable costs have been
covered.
objective of $6000, then the calculation would be as
follows: ($10,000 + $6,000)/($40) = 400 units.
Change in the Marketing Program
Break-even analysis (or its extension, target
volume analysis) can be used for a number of
purposes, one of which (used in connection with a
new product) has already been demonstrated. Similar
analysis can be used to evaluate the economic effect
of a change in the marketing program. When a
marketing program is changed, it will usually involve
a change in one of three things (or combination
thereof): (1) a change in selling price (e.g., a “price
special”), (2) a change in variable costs (e.g., a more
expensive package for the product), and (3) a change
in fixed costs (e.g., increased advertising
expenditures).
Break-even Volume
If the analyst’s objective is to make a breakeven calculation that will show how many units will
have to be sold to cover all costs, then the formula
would be:
BE (units) =
Fixed Costs
Selling price − Var. costs per unit
BE =
(1)
FC
SP − VC
We have been using an example about a
product whose SP = $100, VC = $60, FC = $10,000,
and the desired profit is $6,000. Assume that this
product comes under a new product manager who
wants to cut the price to $90. He also wants to
advertise the special price, and this will cost an extra
$5,000 for TV advertising. Also, a new package
promoting the special price will cost $5 more than the
regular package. Yet the manager will still wish,
under the new marketing program, to retain the same
dollar amount of profit that is currently being made –
usually, he would not want to make any change that
would reduce his profit, or there would be no point in
changing.
For example, if the following information applied: SP
= $100, VC per unit = $60, and FC = $10,000, then
unit contribution would be $100-$60, or $40 per unit.
Break-even volume then would be $10,000/$40, or
250 units. (Note: In some cases, such division will
yield a fraction. All such fractions in break-even
calculations should be rounded-up since the final
results are expressed in units, and one cannot sell less
than a whole unit.)
Target volume
More than likely, a marketer wants to do
better than simply break even on his costs because at
this volume his profits are zero. If the objective is to
do a calculation that will show how many units have
to be sold to reach break-even on costs and to
produce a profit as well, the break-even formula can
be expanded into a “target volume” (TV) formula:
TV (units) =
Fixed costs + Profit
Selling price − Var. costs per unit
Target volume =
UV0404
In this situation there will be a whole new
set of numbers to substitute into the target volume
formula:
TV (units) =
( 2)
TV ( units ) =
FC + P
SP − VC
Old FC + New FC + Profit
Unit contribution
($10,000 + $5,000) + ($6,000)
($100 − $10) − ($60 + $5)
Target volume (in units) = 840 units
For example, if the information in the previous
example was used in connection with a profit
The example above included the current
figures as well as the new figures for the changed
20
UV0405
METHODS FOR PRODUCING PERCEPTUAL MAPS FROM DATA
Introduction
Graphics are instruments for reasoning about quantitative information. If carefully done,
graphics can describe and explore complex data in a way that suggests the useful information.
We often find that a picture of the data allows greater understanding and helps commit the
important features of the data to memory.1 One very common graphic in marketing is a
perceptual map. These maps are used to help managers visualize how their product(s) relate to
other competitive offerings in the marketplace. Common terms in marketing such as “targeting”
and “positioning” implicitly reference the idea that managers are or should be visualizing the
perceptual differences between their product(s) and others in the marketplace. Perceptual maps
are ubiquitous in marketing presentations to help the audience with visualizations of the
competitive landscape. A simple example of a perceptual map is on the next page. It describes
the market for sport utility vehicles.
1
E.R. Tufte, The Visual Display of Quantitative Information (Graphics Press, 1983).
This note was prepared by Ronald T. Wilcox, Associate Professor of Business Administration, University of
Virginia. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an
administrative situation. Copyright © 2003 by the University of Virginia Darden School Foundation,
Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenpublishing.com. No part
of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any
form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of
the Darden School Foundation.
21
UV0405
-2Figure 1: The Market for Sport Utility Vehicles
Prestigious
•
Cadillac
Escalade
•
•
Chevy
Tahoe
• Nissan
Jeep
Pathfinder
Grand Cherokee
•
•
Not Rugged
•
Hummer
H2
Toyota
4Runner
Ford
Explorer
Rugged
•
Jeep
Cherokee
Not Prestigious
In the above map we can see that, among other things, Nissan Pathfinder, Toyota
4Runner, and Jeep Grand Cherokee are perceived to be fairly similar. Hummer H2 is perceived
to be the most rugged SUV in this group and likewise Cadillac Escalade the most luxurious.
Overall, this map allows us to compare each vehicle to all others on the map along the
perceptual dimensions listed on the axis. This kind of analysis requires several assumptions.
Among the most important of these assumptions are:
1. Consumers evaluate all products in this category using the attributes listed along the axis
of the perceptual map.
2. These attributes are the most important of all possible attributes a consumer might use to
compare products in this category.
3. The points depicted on the map, representing products, are accurate representations of
consumers’ view of the marketplace.
We will now discuss how marketing researchers collect and use data on consumers’
perceptions to develop these maps. We will first consider the situation in which the analyst can
make an initial hypothesis or guess as to what perceptual attributes consumers may be using to
evaluate a product and then move on to a situation in which no such prior hypothesis is possible.
22
UV0405
-3Attribute Ratings Method
Constructing the XY plot
There are situations in which product managers may have some idea about which
perceptual attributes their product(s) and those of their competitors are considered prior to
purchase. A good example is the SUV market depicted in Figure 1. Automobiles are very high
involvement purchases, usually requiring consumers to make mental trade-offs between price
and a number of other attributes of competing products before a purchase decision is made.
Over time, automobile manufacturers have developed a reasonably sophisticated understanding,
often through marketing research, of the attributes consumers consider in this process. In order
to develop a perceptual map, a manager must determine which two of the possible attributes are
the most important in the decision process and how consumers’ evaluate each product in the
competitive set with respect to these two attributes.
The analytics of this method begin by collecting consumer ratings for each product, both
the product of interest and other products in the competitive set, for each possible attribute that
management believes consumers might reasonably consider. Perceptual attributes are not the
same as engineering attributes. For example, “number of miles per gallon” is an engineering
attribute. It has a definitive answer. “Fuel efficiency,” while probably highly related to this
engineering attribute, is a perceptual attribute. Consumers are not being asked to report the
quantitative answer to this question, but are being asked about their overall impression of the
vehicle along this dimension. Likewise, “interior space” is an engineering attribute while
“roominess” is a perceptual attribute.
Measuring consumers’ perceptions is generally accomplished by using a Likert scale
survey instrument.2 The general form of this question is:
[Insert vehicle name] is a [insert attribute] sport utility vehicle.
Strongly
Agree
______
Agree
______
Neither Agree
nor Disagree
Disagree
________
_______
Strongly
Disagree
_______
For example, for one question we would insert the name “Toyota 4Runner” and the
attribute “Rugged.” In general, if the researcher wanted to product a map with N competitive
products and was testing M perceptual attributes, he/she would have to ask NxM Likert scale
questions. Once a sufficient sample of consumers have provided this information, the researcher
can convert this ordinal scale to a cardinal scale by assigning the value “5” to “Strongly Agree,”
“4” to “Agree” etc. and compute the arithmetic mean of each attribute score for each vehicle.
The output of this process is a matrix that takes the general form:
2
For a more detailed discussion of using a Likert Scale to measure consumers’ perceptions of product
attributes, see G. Urban and J. Hauser, Design and Marketing of New Products (Prentice-Hall, 1993), 196–199.
23
-4Table 1: Average Attribute Rating for Each Vehicle
Vehicle 1
Vehicle 2
………..
Attribute 1
Attribute 2
………….
Attribute M
UV0405
Vehicle N
[average score]
In addition to this information, the researcher should also ask a single overall preference
question, using the same Likert scale, for each vehicle. The preference question takes the
general form: “I believe that [insert vehicle name] is an excellent [insert product category].” So,
continuing our example, one question would be “I believe that Toyota 4Runner is an excellent
sport utility vehicle.” Thus, the total data set collected by the research includes NxM vehiclesspecific average attribute scores and N overall preference scores.
Recall that the researcher now needs to accomplish two tasks. First he/she needs to
determine which two of the M possible perceptual dimensions are the most important in the
overall evaluation process. Once that is complete, the researcher then needs to place each
product in the appropriate position on the map.
To determine which attributes should define the axis of the perceptual map, the
researcher can estimate a preference regression using a simple linear model. Specifically, the
researcher should estimate:
Overall Preferencein = α + β1 Attribute1in + β2 Attribute2in + …+βM AttributeMin + εin
where Overall Preferencein is the preference of respondent i for product n, Attibute1in is
respondent i’s rating on Attribute 1 for product n and so forth. Greek letters represent parameters
to be estimated by the regression. To keep things simple, we will assume that εin follows the
standard assumptions of the classical linear model so that Ordinary Least Squares, using a
software package such as Excel, can be used to estimate this model.
Estimating this regression will yield estimates of the parameters (coefficients) with
additional information that will allow the researcher to determine whether the parameter
estimates are statistically significant.3
From this point, determining the attributes to use to define the axes is straightforward.
The researcher should choose the two attributes that have estimated coefficients that are the
greatest in absolute value subject to the restriction that are statistically significant. Practically
speaking, if either of the two estimated coefficients that are the greatest in absolute value are not
3
A good refresher on the basic linear model and problems that can arise with this model can be found in Darden
Technical Notes UVA-QA-0500, Introduction to Least Squares Modeling, and UVA-QA-0416, Problems in
Regression, both authored by Phillip E. Pfeifer.
24
-5-
UV0405
statistically significant this probably means that: (1) the sample size is not large enough or; (2)
the set of attributes tested by the researcher is not capturing the important determinants of
product evaluation, and serious consideration should be given to rethinking the data collection
from the beginning.
The two coefficients that are the greatest in absolute value will indicate which attributes
are most important in determining overall preference and hence the most appropriate attributes to
define the axes of the perceptual map.4 The use of the absolute value is important because it is
possible that some attributes may negatively impact product evaluation. For example, in our
SUV example, if one perceptual attribute being tested was “luxurious,” it is entirely possible that
the estimated coefficient of this attribute would be negative as some individuals would equate
luxury with an SUV that is not sufficiently capable of off-road use. It is also possible that the
researcher could define a perceptual attribute as a negative rather than a positive. The perceptual
attribute “gas guzzling” would likely show up as a negative influence in the preference
regression.
Once these two attributes have been determined, the researcher need only to use the
average attribute ratings for these two attributes, found in Table 1, to plot each vehicle on an XY
axis. It is customary to use the most important attribute as the X axis and the second-most
important as the Y axis although it actually makes little difference. Much like Figure 1, positive
attributes define the positive portions of each axis, and their negative counterparts define the
negative portions.
When constructing the actual map itself, it is useful to define the origin, usually the point
(0,0) as (3,3) instead. The reason is straightforward. We would like to use the entire space to
plot the products. Because perceptual attributes are measured on a one to five scale, defining the
origin as (0,0) would force us to plot all products in the first quadrant. This produces a map that
is more difficult to interpret than a map that uses the entire space. Thus, centering the map at
(3,3) makes the graphic clearer and hence more useful as a decision aid.
Here is Figure 1 reproduced with hypothetical average perceptual attributes used as
coordinates.
4
It is not generally the case that one can interpret coefficients in a linear regression in this way. However,
because all of the independent variables use the exact same scale, we can interpret the magnitude of the estimated
coefficient as an indicator of relative importance in determining the dependent variable.
25
UV0405
-6Figure 2: The Market for Sport Utility Vehicles: Attribute Scores Included
Prestigious
•
Cadillac
Escalade (1.2, 4.7)
•
Not Rugged
•
• Nissan
Chevy
• Jeep
Pathfinder (3.6, 4)
Tahoe (1.9, 3.8) Grand Cherokee (2.5,
3 8)
• Toyota
4Runner (3.5, 3.3)
•
Ford
Explorer (2.1, 2)
•
Hummer
H2 (4.5, 4.6)
Rugged
Jeep
Cherokee (4, 1.7)
Not Prestigious
Constructing the ideal vector
We have now constructed a perceptual map using perceptual attribute ratings. There is
one more potentially valuable piece of information that can be gleaned from the data that has
been collected to develop this map. Up to this point, this map contains no information about
which product would be preferred by consumers. Some preference information does appear
obvious. For example, it appears that, all else equal, Nissan Pathfinder would be preferred by
most people to Ford Explorer because it is generally viewed as more rugged and more
prestigious. But, it is less obvious whether consumers generally prefer the Jeep Cherokee or the
Toyota 4Runner. The Toyota is more prestigious but the Jeep is more rugged. What are the
trade-offs consumers are willing to make? This kind of issue becomes even more interesting
when different perceptual maps are developed for different potential target audiences. For
example, the trade-offs men are willing to make among perceptual attributes may be quite
different from those women are willing to make. This kind of information is extremely valuable
for product design decisions as well as target selection and marketing communication strategies.
Fortunately, the procedure described above already contains enough information to be
able to develop stronger statements about consumers’ preferences. This is generally
accomplished through the construction of an “ideal vector,” a vector depicted on the perceptual
map, which illustrates these trade-offs. To construct the ideal vector, the researcher needs to use
the estimated preference regression model. In particular, he/she needs to use the estimated
coefficients from the perceptual attributes that define the axes of the map. The ratio of these two
26
UV0405
-7-
coefficients will determine the slope of the ideal vector, a vector which originates at the origin of
the map. The best way to explain the details of this procedure is through a simple example.
Suppose we estimated the preference regression for SUVs, and it yielded the following results
for the two most important attributes:
Overall Preference = -2.7 + 1.25*Prestige + 2.5*Ruggedness +………..
where Prestige and Ruggedness have been previously determined to be the two most important
attributes in the decision process. Following earlier arguments, the estimated coefficients
indicate that Ruggedness is more important than Prestige in determining overall preference. In
particular, because 2.5 = 2*1.25, we can conclude that Ruggedness is twice as important as
Prestige. The ideal vector is constructed as follows:
1. Take the ratio of the coefficient of the second-most important perceptual attribute to the
most important. In this case, we would have 1.25/2.5 = ½.
2. Plot a vector with slope defined by this ratio and whose beginning point is at the origin of
the graph.5
Figure 3: Constructing the Ideal Vector
Prestigious
•
Cadillac
Escalade
•
•
Chevy
Tahoe
• Nissan
Jeep
Pathfinder
Grand Cherokee
•
•
Not Rugged
•
Hummer
H2
Toyota
4Runner
Ford
Explorer
Rugged
•
Jeep
Cherokee
Not Prestigious
5
The one exception to this is if the researcher places the most important attribute on the Y-axis and the secondmost important on the X-axis. In that case, one would want to use the ratio of the most important attribute to the
second-most important attribute in determining the slope of the vector.
27
UV0405
-8-
Notice that the ideal vector tilts towards the axis defined by the most important attribute.
Preference increases as you move out along the vector. Points of equal preference, or what
economists would call utility, can be determined by constructing lines that are perpendicular to
the ideal vector. Two of these utility lines are depicted in Figure 4.
Figure 4: Constructing Linear Utility Functions
Prestigious
•
Cadillac
Escalade
•
•
Chevy
Tahoe
• Nissan
Jeep
Pathfinder
Grand Cherokee
•
•
Not Rugged
•
Hummer
H2
Toyota
4Runner
Ford
Explorer
Rugged
•
Jeep
Cherokee
Not Prestigious
The answer to the Toyota 4Runner vs. Jeep Cherokee now becomes clear. Because
utility lines that intersect the ideal vector further from the origin represent higher levels of
preference or utility relative to those that intersect closer to the origin, we can definitely state
that, on average, people prefer the Jeep Cherokee to the Toyota 4Runner. We could make
similar preference comparisons between any two vehicles in this map and hence order all
vehicles in terms of their overall preference. If we constructed separate maps for different
segments of the population, we could use the above analysis not only to see how perceptions
vary across different groups of people but also to observe how differences in overall preferences
relate to these perceptions.
Overall Similarity Method
For some product categories, it is virtually impossible to determine a priori what
attributes consumers might be using to evaluate a product. Consider a movie production
company like New Line Cinema, or Hollywood Pictures that is trying to understand how
consumers evaluate movies and make decisions on which ones to see at the theatre. What
perceptual attributes do consumers use to evaluate a movie? First guesses would include things
like amount of action, romance, presence of a particular film star etc. But, movies by their
28
UV0405
-9-
nature are not easily described by a precise group of attributes that can be easily specified. For
example, the movie About Schmidt is by most accounts an excellent movie. However, unlike the
attribute ratings method discussed earlier, it would be very difficult to adequately capture what
makes this movie excellent via a list of perceptual attributes. It is simply a more intangible
evaluation process than, for example, the evaluation process for sport utility vehicles.
For these types of products, the attribute ratings method is a poor way to generate a
perceptual map. We must turn to another method, overall similarity, to try to generate a picture
of the competitive landscape.
Overall Similarity methods work by asking consumers a series of questions designed to
determine how similar or dissimilar different pairs of products are to each other. Once the
researcher has determined the perceived similarity among all possible products in the
competitive set, she can generate a two-dimensional map that graphically reproduces the
perceptual distance between and among the products in the set. To see this clearly lets begin
with a concrete example focused on current movies.
A company like New Line Cinema might begin this process with a list of movies whose
preferences they are interested in probing. For the purposes of our example let’s consider the
movies About Schmidt, Lord of the Rings: Two Towers, Gangs of New York, Maid in Manhattan,
A Guy Thing, and Bowling for Columbine. The researcher would ask an appropriately specified
sample of consumers to evaluate how similar each movie is to each other movie. In our example
there are six movies and fifteen distinct movie pairs. For each movie pair, say About Schmidt
and Bowling for Columbine, there would be a question that reads:
About Schmidt and
Bowling for Columbine
Very
Different
1
2
[ ]
[ ]
3
[ ]
4
[ ]
Very
Similar
5
[ ]
and a corresponding question for every single product pair. Once this judgment data has been
collected, the researcher can simply take the means of the similarity scores and use these means
to fill in the perceived similarity matrix. A perceived similarity matrix might look like the
following:
29
UV0405
-10Table 2: Movie Similarity Matrix
About Schmidt
Lord of Rings
Gangs of NY
Maid in Manhattan
A Guy Thing
Bowling for Columbine
About
Lord
Schmidt of
Rings
5.0
4.2
5.0
3.0
4.0
2.0
1.5
1.0
2.0
3.5
2.5
Gangs Maid in
A Guy
of NY Manhattan Thing
Bowling for
Columbine
5.0
1.7
3.4
2.2
5.0
5.0
2.1
1.9
5.0
1.2
These data represent perceptual distances, really the inverse of distances, between the
movies. A good perceptual map will reproduce these distances, as closely as possible, in twodimensional space.
Reproducing these distances in two-dimensional space is not a simple task. To see this
clearly consider the seemingly unrelated analogy of a three-legged stool and a four-legged chair.
Which one is more likely to wobble back and forth when you are sitting on it? The answer is
obvious, three-legged stools never rocks back and forth but a four-legged chair might. The
reason can be tied directly to geometry. Just as it takes two points to define a line, it takes three
points (legs) to define a two-dimensional plane (the floor). If you try to match four points (legs)
to a two-dimensional plane, there is a chance that they will not all sit perfectly on the surface of
any single plane. In mathematics this is called an “overidentified” space. For our purposes, this
means that, whenever we have more than three products whose perceptual distances we want to
reproduce in two-dimensional space, we are unlikely to be able to do it perfectly. The best we
can hope for is to find a solution that reproduces these distances as closely as possible. The
process of uncovering this solution is called multidimensional scaling, and because the
perceptual data we collect is cardinal rather than ordinal the specific technique presented here is
termed metric multidimensional scaling.
Much like linear regression, the way metric multidimensional scaling (MMDS)
reproduces these perceptual distances is in a way that minimizes the sum of squared errors
between the actual distances (given in Table 2) and the distances reproduced on the map. The
function that the procedure seeks to minimize is commonly called a “stress function” and is
given by the equation:
Stress =
∑
i< j
(dˆij − d ij ) 2
where d̂ ij are the distances specified on the perceptual map and dij are the actual perceptual
distances found in Table 2. The subscripts i and j indicate that the given distance is between
product i and j. There are a number of algorithms that MMDS can use to minimize this function,
30
UV0405
-11-
but these algorithms are included as easy-to-execute procedures in common statistical software.
A discussion of these algorithms is beyond the scope of this technical note.
Inserting the data in Table 2 into a MMDS procedure produces the following twodimensional map.
MDS for Movie Data
Final Configuration, dimension 1 vs. dimension 2
1.0
0.8
MAID
BOWL
0.6
Dimension 2
0.4
0.2
SCHMIDT
0.0
-0.2
-0.4
RINGS
-0.6
-0.8
-1.0
AGUY
NYGANGS
-0.6
-0.2
0.2
0.6
1.0
1.4
Dimension 1
Two things stand out from this procedure. First, there are no axes. This is because the
procedure only finds distances among the products, not their relationships to any axis. In fact,
you can take the above map and turn it upside down, sideways, or rotate it in any way you wish,
and it is still the solution to the minimization problem. Second, and as a direct consequence of
the first point, this procedure does not identify the perceptual dimensions of the map. So, unlike
the attribute-rating method, the researcher must now decide where to draw the axes and what the
axes should be named.
The most common way to specify the names and position of the axes is for the researcher
to use her/his knowledge of the category to define them. This may seem arbitrary, and in some
respects it is, but often the position of the products in the space suggests some perceptual
dimensions that weren’t obvious to the researcher before the data collection began. We leave it
to the reader of this note to speculate about the dimensions of this particular map.
31
-12-
UV0405
A Common Theme
These two methods for producing perceptual maps, attribute-rating and overall-similarity,
both have the ability to depict important relationships among competitive products in twodimensional space. If the researcher believes he/she knows the relevant attributes in consumers’
decision processes, then the attribute-rating method is generally preferred as it leads to a more
easily interpretable perceptual map. If the researcher is not willing to specify a group of
potentially important perceptual attributes, then the overall-similarity method can be used. The
rational behind making a choice between these methods is similar to the trite expression, “pay
me now or pay me later.” The attribute-rating method requires more managerial insight and
more data collection up-front but has the benefit of an easily interpretably map as an output. The
overall-similarity method requires less initial insight and judgment and uses a significantly
simpler data collection procedure. The cost of this ease is a map, which management must
exercise considerable judgment to interpret.
32
Harvard Business School
9-297-028
Rev. October 29, 1996
Clarkson Lumber Company
After a rapid growth in its business during recent years, the Clarkson Lumber Company, in
the spring of 1996, anticipated a further substantial increase in sales. Despite good profits, the
company had experienced a shortage of cash and had found it necessary to increase its borrowing
from the Suburban National Bank to $399,000 in the spring of 1996. The maximum loan that
Suburban National would make to any one borrower was $400,000 and Clarkson had been able to
stay within this limit only by relying very heavily on trade credit. In addition, Suburban was now
asking that Mr. Clarkson guarantee the loan personally. Keith Clarkson, sole owner and president of
the Clarkson Lumber Company, was therefore actively looking elsewhere for a new banking
relationship where he would be able to negotiate a larger loan that did not require a personal
guarantee.
Mr. Clarkson had recently been introduced by a friend to George Dodge, an officer of a much
larger bank, the Northrup National Bank. The two men had tentatively discussed the possibility that
the Northrup bank might extend a line of credit to Clarkson Lumber up to a maximum amount of
$750,000. Mr. Clarkson thought that a loan of this size would improve profitability by allowing him
to take full advantage of trade discounts. Subsequent to this discussion, Mr. Dodge had arrange for
the credit department of the Northrup National Bank to investigate Mr. Clarkson and his company.
The Clarkson Lumber Company had been founded in 1981 as a partnership by Mr. Clarkson
and his brother-in-law, Henry Holtz. In 1994, Mr. Clarkson bought out Mr. Holtz’s interest for
$200,000. Mr. Holtz had taken a note for $200,000, to be paid off in 1995 and 199 6 in order to give Mr.
Clarkson time to arrange for the necessary financing. This note carried an interest rate of 11%, and
was repayable in semi-annual installments of $50,000, beginning June 30, 1995.
The business was located in a growing suburb of a large city in the Pacific Northwest. The
company owned land with access to a railroad siding, and four large storage buildings had been
erected on this land. The company’s operations were limited to the retail distribution of lumber
products in the local area. Typical products included plywood moldings and sash and door
products. Quantity discounts and credit terms of net 30 days on open account were usually offered
to customers.
Sales volume had been built up largely on the basis of successful price competition, made
possible by careful control of operating expenses and by quantity purchases of materials at
substantial discounts. Most of the moldings and sash and door products, which constituted
significant items of sales, were used for repair work. About 55% of total sales were made in the six
months from April through September. Annual sales of $2,921,000 in 1993, $3,477,000 in 1994, and
$4,51 9,000 in 1995 yielded aftertax profits of $60,000 in 1993, $68,000 in 1994, and $77,000 in 1995.
This case was prepared as the basis for class discussion rather than to illustrate either effective or ineffective handling of an
administrative situation.
Copyright © 1996 by the President and Fellows of Harvard College. To order copies or request permission to
reproduce materials, call 1-800-545-7685 or write Harvard Business School Publishing, Boston, MA 02163. No
part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in
any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the
permi ssion of Harvard Business School.
1
33
297-028
Clarkson Lumber Company
Operating statements for the years 1993-1995 and for the three months ending March 31, 1996 are
given in Exhibit 1.
Mr. Clarkson was an energetic man, 49 years of age, who worked long hours on the job. He
was helped by an assistant, who in the words of the investigator of the Northrup National Bank, “has
been doing and can do about everything that Mr. Clarkson does in the organization.” Other
employees numbered 15 in early 1996, 8 of whom worked in the yard and drove trucks, and 7 of
whom assisted in the office and in sales.
As part of its customary investigation of prospective borrowers, the Northrup National Bank
sent inquiries concerning Mr. Clarkson to a number of firms that had business dealings with him.
The manager of one of his large suppliers, the Barker Company, wrote in answer:
The conservative operation of his business appeals to us. He has not wasted
his money in disproportionate plant investment. His operating expenses are as low
as they could possibly be. He has personal control over every feature of his business,
and he possesses sound judgment and a willingness to work harder than anyone I
have ever known. This, with a good personality, gives him a good turnover; and
from my personal experience in watching him work, I know that he keeps close
check on his own credits.
All the other trade letters received by the bank bore out this opinion.
In addition to owning the lumber business, which was his major source of income, Mr.
Clarkson held jointly with his wife, an equity in their home. The house had cost $72,000 to build in
1979 and was mortgaged for $38,000. He also held a $70,000 life insurance policy, payable to Mrs.
Clarkson. Mrs. Clarkson owned independently a half interest in a house worth about $85,000.
Otherwise, they had no sizable personal investments.
The bank gave particular attention to the debt position and current ratio of the business. It
noted the ready market for the company’s products at all times and the fact that sales prospects were
favorable. The bank’s investigator reported: “Sales are expected to reach $5.5 million in 1996 and
may exceed this level if prices of lumber should rise substantially in the near future.” On the other
hand, it was recognized that a general economic downturn might slow down the rate of increase in
sales. Clarkson Lumber’s sales, however, were protected to some degree from fluctuations in new
housing construction because of the relatively high proportion of its repair business. Projections
beyond 1996 were difficult to make, but the prospects appeared good for a continued growth in the
volume of Clarkson Lumber’s business over the foreseeable future.
The bank also noted the rapid increase in Clarkson Lumber’s accounts and notes payable in
the recent past, especially in 1995 and in the spring of 1996. The usual terms of purchase in the trade
provided for a discount of 2% for payments made within 10 days of the invoice date. Accounts were
due in 30 days at the invoice price, but suppliers ordinarily did not object if payments lagged
somewhat behind the due date. During the last two years, Mr. Clarkson had taken very few purchase
discounts because of the shortage of funds arising from his purchase of Mr. Holtz’s interest in the
business and the additional investments in working capital associated with the company’s increasing
sales volume. Trade credit was seriously extended in the spring of 1996 as Mr. Clarkson strove to
hold his bank borrowing within the $400,000 ceiling imposed by the Suburban National Bank.
Balance sheets at December 31, 1993 through 1995 and March 31, 1996, are presented in
Exhibit 2. Statistics for a sample of lumber outlets are provided in Exhibit 3.
2
34
Clarkson Lumber Company
297-028
The tentative discussions between Mr. Dodge and Mr. Clarkson had been in terms of a
revolving, secured, 90-day note not to exceed $750,000. The specific details of the loan had not been
worked out, but Mr. Dodge had explained that the agreement would involve the standard covenants
applying to such a loan. He cited as illustrative provisions the requirement that restrictions on
additional borrowing would be imposed; that net working capital would have to be maintained at an
agreed level; that additional investments in fixed assets could be made only with the prior approval
of the bank; and that limitations would be placed on withdrawals of funds from the business by Mr.
Clarkson. Interest would be set on a floating rate basis at 2´ percentage points above the pri me rate.
Mr. Dodge indicated that the initial rate to be paid would be approximately 11.0% under conditions
in effect in early 1996. Both men also understood that Mr. Clarkson would sever his relationship with
the Suburban National Bank if he entered into a loan agreement with the Northrup National Bank.
3
35
297-028
Clarkson Lumber Company
Exhibit 1 Operating Expenses for Years Ending December 31, 1993-1995, and for First Quarter 1996
(thousands of dollars)
1994
1995
$2,921
$3,477
$4,519
$1,062
330
2,209
337
2,729
432
3,579
587
819
$2,539
337
$3,066
432
$4,011
587
$1,406
607
$2,202
$2,634
$3,424
$799
Gross profit
b
Operating expenses
719
622
843
717
1,095
940
263
244
Earnings before interest and taxes
Interest expense
$97
23
$126
42
$155
56
$19
13
Net income before income taxes
c
Provision for income taxes
$74
14
$84
16
$99
22
$6
1
Net income
$60
$68
$77
$5
Net sales
Cost of Goods Sold:
Beginning inventory
Purchases
Ending inventory
Total Cost of Goods Sold
a
1st Quarter
1996
1993
a
In the first quarter of 1995, sales were $903,000 and net income was $7,000
b
Operating expenses include a cash salary for Mr. Clarkson of $75,000 in 1993; $80,000 in 1994; $85,000 in
1995; and $22,500 in the first quarter of 1996.
c
Clarkson Lumber was required to estimate its income tax liability for the current tax year and pay four quarterly
estimated tax installments during that year. The first $50,000 of pretax profits were taxed at a 15% rate; the next
$25,000 were taxed at a 25% rate; the next $25,000 were taxed at a 34% rate; and profits in excess of $100,000
but less than $335,000 were taxed at a 39% rate.
4
36
Clarkson Lumber Company
Exhibit 2
297-028
Balance Sheets at December 31, 1993-1995, and March 31, 1996 (thousands of dollars)
1993
$
Current assets
Property, net
$686
233
$ 895
262
$1,249
388
$1,243
384
Total Assets
$919
$1,157
$1,637
$1,627
Notes payable, bank
b
Note payable to Holtz, current portion
Notes payable, trade
Accounts payable
Accrued expenses
c
Term loan, current portion
$
$
60
100
-340
45
20
$ 390
100
127
376
75
20
$ 399
100
123
364
67
20
Current liabilities
c
Term loan
b
Note payable, Mr. Holtz
$275
140
--
$ 565
120
100
$1,088
100
0
$1,073
100
0
Total Liabilities
Net worth
$415
504
$ 785
372
$1,188
449
$1,173
454
$919
$1,157
$1,637
$1,627
Total Liabilities and Net Worth
b
$
56
606
587
1st Quarter
1996
$ 43
306
337
---213
42
20
52
411
432
1995
Cash
Accounts receivable, net
Inventory
a
a
1994
$
53
583
607
Interest is computed on the average outstanding loan balance at the rate of prime plus 2 ´%.
Interest is fixed at 11% times the outstanding balance.
c
Interest is fixed at 10.0% times the outstanding balance; the term loan is secured by the fixed assets and is
repayable in semiannual installments of $10,000.
5
37
297-028
Exhibit 3
Clarkson Lumber Company
Selected Statistics on Lumber Outlets
Low-Profit
a
Outlets
High-Profit
a
Outlets
Cost of goods
76.9%
75.1%
Operating expense
22.0
20.6
Percent of sales:
Cash
1.3
1.1
Accounts receivable
13.7
12.4
Inventory
12.0
11.6
Fixed assets, net
12.1
9.2
Total Assets
39.1
34.3
Percent of Total Assets:
Current liabilities
52.7%
29.2%
Long-term liabilities
34.8
16.0
Equity
12.5
54.8
Current ratio
1.31
2.52
Return on sales
(0.7%)
4.3%
Return on assets
(1.8%)
12.2%
Return on equity
(14.3%)
22.1%
a
Defined as the bottom 25% and as the top 25% of all contributors, based on
return on sales.
6
38
Friday, September 30
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Friday, 30 September
08:30
Marketing Session 4 – Professor Nunes
Topic: Conjoint Analysis and Product Design
Required Reading: Wilcox, Ronald T.: “A Practical Guide to Conjoint
Analysis”, UV0406
09:45
Break
10:00
Marketing Session 5 – Professor Nunes
Topic: Breakout Sessions
This session will include time to complete earlier lectures and ensure
comprehension, as well as time to meet as teams with the instructor.
11:15
Break
11:30
Marketing Session 6 – Professor Nunes
Topic: Pricing, Cost Curves and Market Price Evolution
Required Readings:
-
Dhebar, Anirudh: “Price-Quantity Determination”, HBS 9-191-093
-
Steenburgh, Thomas and Jill Avery: “Marketing Analysis Toolkit: Pricing
and Profitability Analysis”, HBS 9-511-028
Due: Exercise 1 – Markstrat Familiarity
Assignment: Exercise 2 – Reading Perceptual Maps
12:45
Lunch
13:45
Finance Session 3 – Professor Callahan
Topic: Introduction to Capital Structure Choices
This session is the first of three sessions in which we examine the
concepts, theories, and evidence concerning corporate capital structure
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
39
Page: 1 / 3
Global Executive MBA in Shanghai
(i.e., the debt/equity mix, or amount of leverage, used to finance a firm).
We will start by understanding how changes in a firm’s leverage change
the allocation of operating cash flows and operating risk among the firm’s
debt holders and equity holders. Next we will look at how the tax
deductibility of interest payments to debt holders can increase firm value
and decrease a firm’s cost of capital.
Required Reading: RWJ Chapters 15 and 16
Discussion Questions:
1. What are the differences in the cash flow rights, control rights, and risk
profile of debt versus equity?
2. How does an increase in corporate leverage change the allocation of
operating risk between debt and equity?
3. Does US (and most other countries’) corporate tax law favor equity or
debt?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
15:00
Break
15:15
Finance Session 4 – Professor Callahan
Topic: Capital Structure – Additional Considerations
In this session we discuss a host of factors that can play a significant role
in firms’ capital structure choices. Potential downsides of debt include the
risk of bankruptcy and a negative impact on the firm’s relations with
customers, employees, and suppliers. Debt can also restrict a firm’s
operating flexibility. Excessive debt can create incentives for
stockholders to under or overinvest relative to the firm’s optimal
investment policy. Potential upsides of debt include its disciplinary effect
on managers and the confidence it can project to financial markets.
Upsides of equity include maintaining financial flexibility and its use as a
form of compensation. Downsides of equity are the potentially negative
signal its use sends to financial markets.
Required Reading: RWJ Chapters 17 and 30
Discussion Questions:
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
40
Page: 2 / 3
Global Executive MBA in Shanghai
1. Why are financial distress costs generally considered to be much greater
than bankruptcy costs?
2. What are two ways in which equity holders might gain at the expense of
debt holders and what mechanisms can debt holders use to try and
protect themselves?
3. How can debt and equity encourage managers to be thrifty and
hardworking?
4. What inferences do markets make about firm value when a firm issues
debt? When a firm issues equity?
To do before this session:
-
Review readings
-
Prepare answers to discussion questions
16:30
Break
16:45
In-Class Pub session
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
41
Page: 3 / 3
UV0406
A PRACTICAL GUIDE TO CONJOINT ANALYSIS
Introduction
Conjoint analysis is a marketing research technique designed to help managers determine
the preferences of customers and potential customers. In particular, it seeks to determine how
consumers value the different attributes that make up a product and the tradeoffs they are willing
to make among the different attributes or features that comprise the product. As such, conjoint
analysis is best suited for products that have very tangible attributes that can be easily described
or quantified.
While the history of conjoint analysis can be traced to early work in mathematical
psychology,1 its popularity has grown tremendously over the last few years as access to easy-touse software has allowed its widespread implementation. There have been probably hundreds of
applications of conjoint analysis in industrial settings.2 Some of the more important questions
modern conjoint analysis is used to analyze are the following:
1. Predicting the market share of a proposed new product, given the current offerings of
competitors.
2. Predicting the impact of a new competitive product on the market share of any given
product in the marketplace.
3. Determining consumers’ willingness-to-pay for a proposed new product
4. Quantifying the trade-offs customers or potential customers are willing to make among
the various attributes or features that are under consideration in the new product design.
1
R.D. Luce and J. W. Tukey, “Simultaneous Conjoint Measurement: A New Type of Fundamental
Measurement,” Journal of Mathematical Psychology 1 (February 1964): 1–27.
2
P.E. Green, A.M. Kreiger, and Y. Wind, “Thirty Years of Conjoint Analysis: Reflections and Prospects,”
Interfaces 31 (May–June 2001): S56–S73.
This technical note was written by Associate Professor Ronald T. Wilcox. Copyright © 2003 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to
sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a
spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or
otherwise—without the permission of the Darden School Foundation. Rev. 11/03.
42
-2-
UV0406
The Anatomy of a Conjoint Analysis
Literally, conjoint analysis means an analysis of features considered jointly. The idea is
that, while it is difficult for consumers to tell us directly how much each feature of a product is
worth to them, we can infer the value of an individual feature of a product by experimentally
manipulating the features of a product and observing consumers’ ratings for that product or
choices among competing products.
To fix your intuition here, consider the simple example of a sports car. It would be
difficult for the average consumer to tell a market researcher exactly how much more valuable a
car with 240 horsepower is to them relative to one with 220. It is possible that a consumer might
be able to come up with some dollar value, but that value may not really reflect the way they
would make choices if faced with a real marketplace situation. Instead, marketers have found
that it is much more accurate to present individuals from the target market with a series of cars,
described not only by their horsepower but by other attributes as well (color, price,
standard/automatic transmission, etc.) and then ask them to rate each of the cars on a numerical
scale. Alternatively, the researcher presents several competing cars with different attributes and
asks the consumer to choose one. By repeatedly asking the potential customers to rate the cars or
chose a car from a competing set, the researcher can infer the value of each individual attribute.
This is the essence of a conjoint analysis; replacing the relatively inaccurate method of asking
about each attribute in isolation with a model that allows us to infer the attributes’ values from a
series of ratings or choices.
The Experimental Design
A conjoint analysis begins with an experimental design. This design includes all
attributes and the values of the attributes that will be tested. Conjoint analysis distinguishes
between attributes and what are generally called “levels.” An attribute is self-explanatory. It
could be price, color, horsepower, material used for upholstery, or presence of a sunroof, while a
level is the specific value or realization of the attribute. For example, the attribute “color” may
have the levels “red,” “blue,” and “yellow,” while the attribute “presence of a sunroof” will have
levels “yes” and “no.” Before a researcher begins to collect data, it is important that all the levels
of each attribute to be tested are written down. Commercially available software packages
require that the user provide these as input.
43
UV0406
-3-
Continuing with our car example, an experimental design might look like the information
presented in Table 1:
Table 1. Example of experimental design.
Levels
Price
$23,000
$25,000
$27,000
$29,000
Brand
Toyota
Volkswagen
Saturn
Kia
Horsepower
220 HP
250 HP
280 HP
Upholstery
Cloth
Leather
Sunroof
Yes
No
This is a very simple design that contains a total of 15 attribute levels. Real designs often
contain more attributes and levels than are presented here.
When constructing an experimental design, it is important to keep in mind:
1. The more tangible and understandable the levels of each attribute are to the respondents, the
more valid the results of the research will be. For example, attribute levels such as “really
roomy” are vague, meaning different things to different people and should be avoided.
2. The greater the number of attribute levels to be tested, the more data that will be needed
to achieve the same degree of output accuracy.
3. For quantitative variables (price and horsepower in this example), the greater the distance
between any two consecutive levels, the harder it will be to get a good idea of how a
consumer might evaluate something in between the two (i.e., $24,000).
Data Collection
Collecting data for a conjoint analysis has been made relatively simple by the advent of
dedicated off-the-shelf software. The exact nature of the data collected will be dictated by the
type of conjoint analysis that is used. An exhaustive discussion of the benefits and drawbacks of
each of the many different types of conjoint analysis now in use is beyond the scope of this
technical note. However, those interested are encouraged to read Orme for a good discussion of
this topic.3
The state-of-the-art in conjoint data collection involves using personal computers or a
Web-based version of the software to guide respondents through an interactive conjoint survey.
The software creates the hypothetical product profiles using the experimental design provided by
the researcher and estimates the attribute-level utilities from participant ratings or choices.
3
B. Orme, “Which Conjoint Method Should I Use?” Sawtooth Software Technical Paper (2003). Currently, an
Acrobat-readable copy of this paper is available at http://www.sawtoothsoftware.com/techabs.shtml#which.
44
UV0406
-4Interpreting Conjoint Results
Understanding the basic output
The basic results of a conjoint analysis are the estimated attribute level utilities. Keeping
with the example in Table 1, conjoint output might look like Table 2:
Table 2. Conjoint analysis output.
Attribute
Price
Brand
Horsepower
Upholstery
Sunroof
Level
$23,000
$25,000
$27,000
$29,000
Toyota
Volkswagen
Saturn
Kia
220 HP
250 HP
280 HP
Cloth
Leather
Yes
No
Utility (Part-worth)
2.10
1.15
−1.56
−1.69
0.75
0.65
−0.13
−1.27
−2.24
1.06
1.18
−1.60
1.60
0.68
−0.68
t-value
14.00
7.67
10.40
11.27
5.00
4.33
0.87
8.47
14.93
7.07
7.87
10.67
10.67
4.53
4.53
The estimated utilities or part-worths correspond to average consumer preferences for the
level of any given attribute. Within a given attribute, the estimated utilities are generally scaled
in such a way that they add up to zero. So a negative number does not mean that a given level
has “negative utility”; it just means that this level is on average less preferred than a level with an
estimated utility that is positive.
Conjoint analysis output is also often accompanied by t-values, a standard metric for
evaluating statistical significance. Because of the way conjoint utilities are scaled, the standard
interpretation of t-values can yield misleading results. For example, the level “Saturn” of the
attribute “Brand” has a t-value of 0.87. In general, a t-value of this magnitude would fail a test of
statistical significance. However, this t-value is generated because within the attribute “Brand”
the level “Saturn” has neither a very high nor very low relative preference. It is basically in the
middle in terms of overall preference. Because of the scaling, levels that have more moderate
levels of preference within a given attribute are likely to have estimated utilities close to zero,
which will tend to produce very low t-values (recall that the t-test is measuring the probability
that the true value of a parameter is not different from zero).
45
-5-
UV0406
A better way to think about statistical significance in this context is to examine the tvalues of the levels with the highest and lowest preference within a given attribute. An applicable
common practice would be if the sum of the absolute value of these two statistics is greater than
three, then that given attribute is significant in the overall choice process of consumers. At a
practical level, it is rare that an attribute will not be significant, and, if you find one that is, it
means that it probably should not have been included in the experimental design in the first
place, because respondents are not considering that attribute’s information when they make
choices.
Conjoint Analysis Applications
As mentioned previously, there are many different possible applications of conjoint
analysis. We will focus on three very common applications: trade-off analysis, predicting market
share, and determining overall attribute importances.
Trade-off analysis
The utility of any given product that we might consider can be easily computed by simply
summing the utilities of its attribute levels. For example, a Toyota with 280 horsepower, leather
interior, no sunroof, and a price of $23,000 has a utility of 0.75 + 1.18 + 1.60 − 0.68 + 2.10 =
4.95. If the car with same basic specifications were a Volkswagen, the overall utility would drop
to 0.65 + 1.18 + 1.60 − 0.68 + 2.10 = 4.85, a drop of 0.10. This drop can be seen directly by
noticing that the difference between the utility for the brand Toyota (0.75) and Volkswagen
(0.65) is 0.10 and because nothing else in the profile of the car has changed this will be the exact
utility difference between two cars that are the same except for this brand difference.
A natural consequence of the above observation is that we use the utilities to analyze
what average consumers would be willing to give up on one particular attribute to gain
improvements in another. For example, how much money would they be willing to give up
(price) if a sunroof was added to the vehicle? We will now look directly at this issue of the
hypothetical car detailed in the previous paragraph. Adding a sunroof to the (Toyota) car would
yield an overall utility of 0.75 + 1.18 + 1.60 + 0.68 + 2.10 = 6.31. This represents an increase in
utility of 6.31 – 4.95 = 1.36 over the identical car without a sunroof.
The information above directly implies that we can reduce the utility of price by 1.36,
and average consumers would be just as happy as before the sunroof was installed. To find out
how much the price can be raised, we must convert the change in utility with a change in price.
We do this by first noting how much the original car costs ($23,000) and the utility associated
with that figure, 2.10. We know that we can reduce the price utility by 1.36. This is equivalent to
saying that we can reduce the price utility to 2.10 – 1.36 = 0.74. By referring to Table 2, we can
immediately see that this implies a price between $25,000 and $27,000 because −1.56 < 0.74 <
1.15. In fact, if we assume a linear relationship between price and utility in the range between
46
UV0406
-6-
$25,000 and $27,000, we can solve for the exact price by performing a linear interpolation within
this range.4 Specifically, the interpolation yields:
$25,000 +
1.15 − 0.74
× $2,000 = $25,302.58
1.15 − (−1.56)
Utility spread between the
two tested price points (25K
and 27K).
Utility spread between 25K
and the target utility.
This implies that, if the sunroof is added, the price of the vehicle could be raised from
$23,000 to about $25,300, and the average consumer’s attitude would be one of indifference
between the two vehicles. Qualitatively, it shows that the value of a sunroof to consumers is very
substantial.
This same kind of analysis can be performed for other attributes. We could ask how much
additional horsepower we would need to add if the interior was changed from leather to cloth.
This particular question does present a problem, however. Because the current vehicle under
consideration has 280 horsepower, and that is the maximum amount of horsepower tested by the
conjoint analysis, it will be impossible to determine how much consumers will value additional
horsepower. This leads to an important consideration when constructing the experimental design.
That is, if the output is to be used for trade-off analysis, it is important that the range of the levels
tested within each attribute span the entire range of that attribute before management would ever
consider it as a realistic design alternative. If the experimental design takes this into account, we
can perform trade-off analysis between any two attributes in the design.
Market share forecasting
Another common application is forecasting market share. In order to use conjoint output
for this kind of prediction, two conditions must be satisfied:
1. The company must know the other products, besides their own offering, that a consumer
is likely to consider when making a selection in the category.
2. Each of these competitive products’ important features must be included in the
experimental design. In other words, you must be able to calculate the utility of not only
your own product offering but that of the competitive products as well.
4
This is a common way to approximate the relationship between the value of the attribute and its utility for
attribute values that were not directly tested by the conjoint analysis. The closer the tested levels are to each other
the more accurate this approximation. Also notice that this interpolation can only be performed for quantitative
attributes such as price. Interpolating between qualitative attributes, like brand, is nonsensical.
47
UV0406
-7-
Market share prediction relies on the use of a multinomial logit model.5 The basic form of
the logit model is:
Sharei =
∑
eU i
n
j =1
Uj
e
where:
Ui is the estimated utility of product i
Uj is the estimated utility of product j
n is the total number of products in the competitive set, including product i.
To make things clear, consider the following example. Suppose we are interested in
predicting the market share of a car with the following profile: Saturn; $23,000; 220 HP; cloth
interior; no sunroof. We believe that when consumers consider our car they will also consider
purchasing cars that are currently on the market with the following profiles:
1. $27,000; Toyota; 250 HP; cloth interior; no sunroof
2. $29,000; Volkswagen; 280 HP; leather interior; no sunroof
3. $23,000; Kia; 220 HP; cloth interior; no sunroof
For the Saturn and its associated product profile the estimated utility is 2.10 – 0.13 – 2.24
– 1.60 – 0.68 = − 2.55. Similarly, the utilities of the three competing products can be calculated:
1. −1.56 + 0.75 + 1.06 – 1.60 – 0.68 = −2.03
2. −1.69 + 0.65 + 1.18 + 1.60 – 0.68 = 1.06
3. 2.10 – 1.27 – 2.24 – 1.60 – 0.68 = −3.69
With these utilities in hand we can now directly apply the logit model to forecast market share
for the Saturn. This is given by:
ShareSaturn =
e −2.55
= .025 or 2.5%
e − 2.55 + e − 2.03 + e1.06 + e − 3.69
This implies that this particular Saturn vehicle will achieve a 2.5% market share within
the specified competitive set. The market share of any vehicle that can be described by the
experimental design and a set of competitive vehicles, also described by the experimental design,
can be found in a similar manner.
5
A good marketing reference to learn about the basics of the logit model is G. Lilien and A. Rangaswamy,
Marketing Engineering: Computer-Assisted Marketing Analysis and Planning (2nd Ed.), Englewood Cliffs:
Prentice-Hall (2002). Also, most econometrics textbooks will have information on logit models.
48
UV0406
-8Determining attribute importance
A researcher may also be interested in determining the importance of any individual
attribute in the consumers’ decision processes. Quantifying these attribute importances using the
conjoint output is straightforward and can provide both interesting and useful insights into
consumer behavior.
Intuitively, the variance of the estimated utilities within a given attribute tells you
something about how important the attribute is in the choice process. Take, for example, the
attributes “Sunroof” and “Upholstery,” both of which have only two levels. If you understand the
material up to this point, it should be reasonably clear that “Upholstery” is a more important
attribute than “Sunroof.” That is because the utility difference between having a sunroof and not
having a sunroof (2 × 0.68 = 1.36) is smaller than the utility difference between having leather
versus cloth interior (2 × 1.60 = 3.20).
The common metric used to measure attribute importances is:
Ii =
∑
Ui −U i
n
j =1
U j −U
j
where:
Ii = the importance of any given attribute i
U = the highest utility level within a given attribute (subscripts indicate which attribute)
U = the lowest utility level within a given attribute.
This equation is really quite intuitive. In order to calculate the importance of any given
attribute, you just take the difference between the highest and lowest utility level of that attribute
and divide this by the sum of the differences between the highest and lowest utility level for all
attributes (including the one in question). The resulting number will always lie between zero and
one and is generally interpreted as the percent decision weight of an attribute in the overall
choice process.
It also should be clear at this point that this estimated attribute importance depends
critically on your experimental design. In particular, if you increase the distance between the
most extreme levels of any given attribute you will almost certainly increase the overall attribute
importance. For example, if the tested price range was $21K − $31K instead of $23K – $29K
(Table 1), this is very likely to increase the estimate attribute importance of price.
Let’s now consider a concrete example using the attribute “Horsepower.” The importance
of this attribute is calculated as:
49
-9-
I Horsepower =
UV0406
1.18 + 2.24
= .25
((2.10 + 1.69) + (0.75 + 1.27) + (1.18 + 2.24) + (1.60 + 1.60) + (0.68 + 0.68) )
In our example, 25% of the overall decision weight is assigned to horsepower. The reader
may verify through analogous calculations that the decision weight for “Price” is about 27%,
“Brand” about 15%, “Sunroof” about 10%, and “Upholstery” about 23%. The numbers provide a
very intuitive metric for thinking about the importance of each attribute in the decision process.
Final Thoughts
Conjoint analysis has a broad array of possible applications. Many of these applications
are variants of the three very common applications presented here. The increasingly widespread
availability of conjoint analysis software, both PC and Web-enabled, points to its continued
growth as a marketing decision aid.
This note has presented what is generally known as “aggregate-level” conjoint analysis.
That is, all of the respondents are pooled into one group, and a single set of attribute level
utilities are estimated from the ratings or choices provided by the people in this group. Recent
advancements in conjoint analysis have enabled researchers to estimate different utilities for
different groups of respondents and even, in some cases, for individual respondents. While the
mathematics necessary for this procedure is sometimes quite complex, it is now possible to
estimate the attribute level utilities and to compute trade-off analyses for each individual
respondent. This has some significant advantages over aggregate level analysis particularly when
considering marketing segmentation issues. Either way, the data collection and the basic
interpretation of the output remains the same. While there is currently no textbook that can
provide the reader with answers to all of the questions that might arise when applying this
technique in a business setting, there is, as of this writing, a very good and surprisingly
comprehensive collection of technical papers located on the site of a company that markets
conjoint analysis software (http://www.sawtoothsoftware.com/techpap.shtml). These provide
answers to many of the practical implementation questions a user may face.
50
Harvard Business School
9-191-093
Rev. June 23, 1993
Price-Quantity Determination
This note examines some of the economic considerations affecting two basic decisions in
business: what price should a firm charge for a product, and what quantity of the product should it
make? We begin with a discussion of the appropriate decision criterion (Section 1). In Section 2, we
consider the demand curve and see how it establishes a linkage between price and quantity. The
heart of the analysis is in Section 3, in which we establish the following decision rule: a necessary
condition for optimal price and quantity is that marginal revenue must equal marginal cost. Pricequantity determination in the presence of a resource constraint is analyzed in Section 4. Finally, in
Section 5, we comment on the competitive case.
1.
The Appropriate Decision Criterion
What should be the firm’s objective when determining price and quantity?: To maximize
market share? To achieve a certain margin on cost? To maximize return on sales? To attain at least a
minimum return on capital employed?
Let us consider each of these in turn. It is reasonable for a firm to strive to increase market
share. However, should it do so by choosing a price so low that it is not profitable to stay in business?
“Cost-plus-margin” is a decision rule that eliminates most of the anxiety surrounding the pricing
decision, but does it make sense to achieve the same margin on a product of which the firm is the only
supplier as on a product for which there are many suppliers? “Maximize return on sales” and
“return on capital employed” have at least two weaknesses. First, since they measure ratios rather
than absolute amounts, they can be large not only because the numerator is large (which is good), but
also because the denominator is small (which is not necessarily good). Second, both measures can be
unduly sensitive to the definition of “return.” The example in Table 1 illustrates these weaknesses in
the case of “maximize return on sales.”
Should the firm choose Alternative C because it promises the highest return on sales? Clearly
not: Alternative C has a higher return on sales, not because it results in a large income, but because
sales are low (as indicated earlier, a ratio can be high because the numerator is high, the denominator
low, or both; here, the denominator is low).
Professor Anirudh Dhebar wrote this note as the basis for class discussion rather than to illustrate either effective or
ineffective handling of an administrative situation.
Copyright © 1990 by the President and Fellows of Harvard College. To order copies or request permission to
reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to
http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system,
used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying,
recording, or otherwise—without the permission of Harvard Business School.
1
51
191-093
Table 1
Price-Quantity Determination
Maximize Return on Sales?
Sales
Direct costs
Sales less direct costs
Allocated costs*
Net Income
Return on Sales
Alternative A
Alternative B
$95 million
$63 million
$32 million
$19 million
$13 million
$70 million
$42 million
$28 million
$14 million
$14 million
13.7%
20.0%
Alternative C
$50 million
$28 million
$22 million
$10 million
$12 million
24.0%
Note: Company practice requires the allocation of 20% of sales towards the recovery of general and administrative expenses
incurred by the firm as a whole—expenses that would be incurred whether the firm chooses Alternative A, B, or C.
What about Alternative B, which promises the highest return in absolute terms? This
alternative is also not the correct choice: “net income” is the highest under the alternative, but net
income is calculated after subtracting allocated costs—costs that do not relate to any of the
alternatives and would be incurred irrespective of the alternative that is chosen. (Costs that are
specific to an alternative are included in the category “direct costs.”)
The appropriate decision criterion for the firm should be the following: choose the alternative
that maximizes the present value of the net cash flows to the firm. “Net cash flow” is not the same
thing as “net income.” In general, net income is calculated after subtracting allocated costs, expenses
deferred for tax purposes, depreciation charges (but not capital cost), accrued expenses, etc. On the
other hand, net cash flow calculations focus on the actual cash flows arising from the alternative
under consideration. These calculations include depreciation tax benefits and capital costs, but
exclude allocated costs and costs incurred before the point in time when the decision is being made;
they also adjust for the actual receipt of revenues and actual payment of expenses.
We will make the following simplifying assumptions in this note: Revenues and costs are not
deferred, there is no depreciation, the tax rate is zero, and cash flows occur at the same point in time
(and, therefore, their present value is the sum of the cash flows). Given these assumptions, the net
cash flow for each alternative is the difference between “sales” and “direct expenses.”
Returning to the example in Table 1, the correct choice is Alternative A: it will give the
highest net cash flow—$32 million instead of $28 million (Alternative B) or $22 million (Alternative
C).
Admittedly, the above argument is somewhat simplistic. Depending on the decision context,
one can make very valid arguments for the use of decision criteria such as “maximize return on
sales,” or “achieve a certain margin on cost,” or “attain at least a minimum return on capital
employed.” Ultimately—and over the long run—however, the firm’s choice must result in an
acceptable stream of cash flows to the investors. It is this reasoning that motivates the decision
criterion in this note.
2.
Price, Quantity, and the Demand Curve
Suppose the firm is using the appropriate decision criterion and must decide on the price and
quantity for one of its products. Does it have complete freedom in this decision? Usually, it does
not—because of one or more of the following factors: limited demand, restricted supply, and
threatening competition. We consider demand limitations in this section, supply (capacity)
constraints in Section 4, and competition in Section 5.
To motivate the concept of limited demand, assume the firm has chosen a price for the
product under consideration. Consumers learn of this price and decide whether or not to buy the
2
52
Price-Quantity Determination
191-093
product. This decision is made at the level of the individual consumer, and is based on a comparison
of the maximum he or she is willing to pay for the product with the price of the product: the
consumer will buy the product only if his or her willingness-to-pay is greater than or equal to the
product price. (The difference between a consumer’s willingness-to-pay and the price is known as the
consumer surplus; only those with a consumer surplus greater than or equal to zero will buy the
product.)
What determines a consumer’s willingness to pay for a product? Obviously, the product
itself is an important determinant: most of us would pay much more for, say, a car than a wristwatch.
For a given product category, willingness-to-pay depends on product features, quality, performance,
etc. Thus, for a car, we would usually pay a premium for additional power, space, air conditioning,
sun roof, etc. In the case of consumable products (which, unlike durable products, are typically
purchased repeatedly and consumed in quantities greater than one), willingness-to-pay for each
additional unit of the product typically declines as the total quantity consumed increases (the fifth
scoop of ice cream is not worth as much as the first scoop; arguably, it might even be bad for the
consumer!). Whether the product is a durable product or a consumable product, the willingness-topay depends on the alternatives available to the consumer: it is usually lower if cheaper alternatives
are available from competitors or in the form of perfect or imperfect substitutes. Finally, the
consumer’s preferences for the product are determined by his or her income, wealth, consumption
budget, and tastes.
Since a consumer will buy a product only if his or her willingness-to-pay exceeds its price,
and since any population of consumers is typically heterogenous in the willingness-to-pay (some
consumers value the product a great deal, others somewhat, and others hardly at all), it follows that
only a subset of consumers will choose to buy the product. In other words, demand for the product
will be limited.
In general, the higher the price, the smaller will be the number of consumers with
willingness-to-pay greater than or equal to the price, and, therefore, the smaller will be the demand.
On the other hand, the lower the price, the larger will be the number of consumers with willingnessto-pay greater than or equal to the price, and the greater will be the demand. This characteristic can
be represented graphically with price on the horizontal axis and demand on the vertical axis (see
Figure 1). The resulting curve is called the demand curve for the product: it gives the quantity of the
product that consumers will demand as a function of its price. Since demand usually decreases as
price increases, demand curves are typically downward-sloping. You can think of the demand curve
as a constraint on the firm’s price-quantity decision: once a price is chosen, the quantity is
automatically determined, and vice versa.
Figure 1
Typical Demand Curve
D
e
m
a
n
d
Price
3
53
191-093
Price-Quantity Determination
The shape of the demand curve plays an important role in the determination of product price
and quantity. An important measure in this context is the elasticity of demand: the per cent increase
(decrease) in demand that would result from a one per cent decrease (increase) in price.
Figure 2 illustrate two extreme cases of demand elasticity. The curve in Figure 2 (a) is a
demand curve for a product with highly elastic demand. Demand is said to be highly elastic with
respect to price if small changes in price leads to relatively large changes in demand. Demand for
luxury items, e.g., Rolls-Royce cars, tends to be highly elastic. Figure 2 (b) illustrates the demand
curve for a product with relatively inelastic demand. Demand is said to be inelastic with respect to
price if large changes in price lead to relatively small changes in demand. Demand for the basic
necessities of life, e.g., essential food items, tends to be relatively inelastic.
Figure 2 (a)
Demand Curve for Product
with Elastic Demand
D
e
m
a
n
d
Figure 2 (b)
D
e
m
a
n
d
Price
3.
Demand Curve for Product
with Inelastic Demand
Price
Determining the Optimal Price and Quantity
We saw in Section 1 that the firm’s goal should be to maximize the difference between the
revenues and the costs attributable to a product. In this section, we examine how the firm should
choose an optimal price-quantity combination using the information in the product’s demand curve.
Before beginning with the analysis, however, we should understand that the firm must make a “gono go” decision as well, i.e., it must also decide whether or not to offer the product in the first place.
Interestingly, while in real time, the “go-no go” step must occur first, in the analytical sequence, the
firm first chooses the price and the quantity, and then decides whether or not to go ahead with the
product.
Proceeding with the analysis, calculating revenue is simple: it equals price times quantity. As
for direct costs, it is useful to distinguish between costs that vary with quantity (variable costs; e.g., cost
of raw materials and direct labor) and costs that do not vary with quantity (direct fixed costs; e.g.,
product launching costs, cost of “indirect” labor and supervisory product management, rent for space
obtained specifically for the product under consideration). Since direct fixed costs do not vary with
quantity, they cannot have a bearing on the quantity decision; only variable costs are relevant to this
decision, and the decision should be based on the maximization of the difference between revenues
and variable costs, i.e., contribution. Of course, direct fixed costs are not completely irrelevant. If the
maximum contribution is not sufficient to cover these costs, then the firm should stop making the
product. In other words, direct fixed costs are relevant to the “go-no go” decision, but not the
quantity decision. (Costs that the firm incurs in connection with its other products and businesses
and overall corporate administration do not change whatever the firm decides about the product
under consideration; they are not relevant to the quantity decision or the go-no go decision.)
4
54
Price-Quantity Determination
191-093
The firm can determine the contribution-maximizing price and quantity by adopting the
following procedure (see Figure 3 and Table 2): Start with a price P1 , read the demand D1 from
the demand curve, and calculate the contribution for price P1 ; choose a new price P2 , read the
demand D2 from the demand curve, and calculate the contribution for price P2 ; choose a new price
P3 , read the demand D3 from the demand curve, and calculate the contribution for price P3 ;
choose a new price P4 , read the demand D4 from the demand curve, and calculate the contribution
for price P4 ;…; stop when a price is chosen such that the total contribution is as large as possible.
The greater the number of points along the demand curve that are sampled, the closer to the true
optimum the final answer will be.
To see how the above procedure works, let us go back to the demand curve (see Figure 4).
The decision maker starts with a low price (not lower than the variable cost per unit, however), and
increases it by small amounts, always checking if the benefits from doing so outweigh the costs. Let
us say the decision maker has reached point A and is considering whether or not to increase the price
to point B.
At point A, the area (A+C) represents the firm’s total contribution. Similarly, at point B, the
area (B+C) represents the total contribution. Should the firm move from point A to point B? Yes, if
area (B+C ) exceeds area (A+C), and not, otherwise. Observe that
area (B+C) - area (A+C) = area B - area A
= “increase in revenue on account of a higher price”
- “decrease in contribution because of decreased demand.”
Figure 3
Determining Optimal Price and Quantity
D1
D
e
m
a
n
d
D2
D3
D4
v
(P3-v)
Unit
Unit Contribution
Variable
a t Pric e P3
Cost v
P1
P2
P3
P4
Price
Table 2
Determining Optimal Price and Quantity
Price
Demand
Unit
Variable Cost
Unit
Contribution
Total
Contribution
P1
P2
P3
P4
•
D1
D2
D3
D4
•
v
v
v
v
•
[P1 - v]
[P2 - v]
[P3 - v]
[P4 - v]
•
[P1 - v] x D1
[P2 - v] x D2
[P3 - v] x D3
[P4 - v] x D4
•
5
55
191-093
Figure 4
Price-Quantity Determination
Finding Optimal Price and Quantity: "Marginal Analysis"
Unit
Variable
Cost
A
Demand at Price PA
D
e
m
a
n
d
A
C
Unit Contribution
B
B
Price
at Price P 3
The decision maker should go on increasing price by “one cent” (or “one dollar” or “one per
cent”) as long as the “increase in revenue on account of a higher price” exceeds the “decrease in
contribution on account of decreased demand.” Ultimately, a point will be reached when the two
terms are equal. This is the optimal point. For any further increase in price, “increase in revenue on
account of a higher price” will fall short of “decrease in contribution on account of decreased
demand.” This is what marginal analysis is all about.
Once the firm knows the optimal price and quantity, it should check whether or not the
resulting total contribution exceeds the direct fixed costs.
The above discussion assumed a constant variable cost per unit of the product: the
incremental cost to the firm of the 2nd (but not the 1st!) unit of product is the same as that of the 3rd
unit, the 4th unit, the 45th unit, the 7363rd unit, etc. In other words, the cost of producing an
additional unit of the product is independent of the total quantity that is produced. In practice, this
assumption usually only holds over limited ranges of production quantities. Over large quantity
ranges, the unit variable cost may increase (because of some production diseconomies) or decrease
(because of production economies) as the total production quantity increases. Furthermore, in some
situations, for example, those involving joint products and costs, it may be difficult, if not impossible,
to determine the unit variable cost at any level of production. In this note, we will restrict the
discussion to the case where the unit variable cost does not change with total production quantity.1
Example
You are thinking about launching a new high-tech gizmo, a digital AirTime Hogmeter®. The
market for the new product is limited: you do not expect anyone other than first-year students to buy
the product, and this is going to be a one-time venture (next year, you will be in the second year and,
presumably, will have better things to do with your time). You could give the meters away, in which
case all 800 students will pick up the freebies. But you would not want to do that: Oisac Corporation,
your Napajese supplier from the Pacific rim, is going to charge you $20 for each meter. What should
be the selling price for the meter, and how many meters should you order?
1This assumption does not imply that there are no scale economies. If the direct fixed costs are greater than
zero, total cost (variable plus direct fixed) per unit (often used to measure scale economies) will decline even if
the unit variable cost is constant.
6
56
Price-Quantity Determination
191-093
The first thing you do is conduct some market research: Will anyone buy the meter for a
zillion dollars? No way. How about $1000? No: it is cheaper to keep tab of air time consumption
using a $16.95 watch, a sheet of paper, and a pencil. $100? There is some interest, but no one has yet
said a definite yes. $80? People are about to sign up, but not quite yet. $79? 10 people have
indicated that they will buy the meter for $79. You try $70. Now demand is building up: 100 have
indicated a willingness to buy (in the population of 800, 100 have a willingness-to-pay greater than or
equal to $70). You discover that 200 will buy if you price the meter at $60, 300 at $50, 400 at $40, 500
at $30, 600 at $20, 700 at $10 (of course, you would not want to do that; the meter costs you $20 a
piece!), and 800 for free. You plot this information on a sheet of graph paper, and you have a
demand curve (see Figure 5).
Figure 5
Demand Curve for AirTime Hogmeter
600
D
e
m
a
n
d
500
400
300
200
100
0
$20
$30
$40
$50
$60
$70
$80
Price
You can find the optimal price and quantity using the procedure described earlier. Table 3
reproduces the methodology in Table 2, and Figure 6 presents a graph of total contribution as a
function of price. From Table 3 and Figure 6, we see that contribution is maximized at a price of $50
and a quantity equal to 300. The maximum contribution is $9,000. Not bad. Will it cover your
product launching costs and your administrative costs? If so, you should go ahead.
Table 2
Determining Optimal Price and Quantity for AirTime Hogmeter
Price
Demand
$20
$30
$40
$50
$60
$70
$80
600
500
400
300
200
100
0
Unit
Variable Cost
$20
$20
$20
$20
$20
$20
$20
Unit
Contribution
Total
Contribution
$0
$10
$20
$30
$40
$50
$60
$0
$5,000
$8,000
$9,000
$8,000
$5,000
$0
7
57
191-093
Price-Quantity Determination
Figure 6
4.
Determining Optimal Price and Quantity for AirTime Hogmeter
Price-Quantity Determination with a Resource Constraint
In Section 3, we saw how the demand curve acts as a constraint on the firm’s choice of price
and quantity. In this section, we consider price-quantity determination in the presence of a resource
constraint. Specifically, we examine a case with two products—an existing product, Product A, and a
new product, Product B. While the two products are unrelated, they are made with the help of a
common resource, only a limited amount of which is available.
Suppose the firm is selling Product A for $50. The direct fixed cost of making the product is
$15,000 and the variable cost per unit is $10 (the first unit costs the firm $15,010; each additional unit
costs an incremental $10). Demand for Product A is very high—demand the firm is unable to meet
because of limited manufacturing capacity: it takes four hours to machine each unit of the product,
and only 2,000 hours of machine time are available, restricting production to 500 units. At this
production level, the firm makes a total contribution of $20,000 (unit contribution $40 times 500
units), and a net cash flow of $5,000 (after paying the $15,000 direct fixed cost for Product A).
Consider Product B. The firm must first determine whether it is beneficial to divert any
machining capacity from Product A to Product B. Suppose the firm diverts four hours of machining
capacity from Product A to Product B. This would result in a drop in the production of Product A by
one unit (from 500 to 499 units), and a reduction in total contribution of $40. Clearly, the firm should
divert capacity to Product B only if it can manufacture enough Product B in the four hours to offset
the contribution loss. Suppose it takes two hours to machine each unit of Product B. Then, in four
hours, the firm can machine two units of the product. It follows that Product B should be made only
if the contribution realized from the sale of one unit of the product exceeds $20 ($40, the drop in
contribution from making one unit less of Product A, divided by the two units of Product B that can
be made in the same amount of time).
We could repeat the above analysis, now for a diversion of one hour of machine time from
Product A to Product B. In that case, the firm would make one-quarter units less of Product A, and
the total contribution would go down by $10 (unit contribution of $40 times one-quarter units;
assume that the firm can sell fractional units of the product). In the one hour that is freed up, the firm
can make one-half units of Product B, from the sale of which it must obtain a contribution of at least
$10 for the whole exercise to be worthwhile.
We conclude that the “breakeven” contribution for one hour of machine time is $10 ($40, the
unit contribution for the existing product, Product A, divided by 4 hours, the machining time for
Product A); we call this breakeven value the opportunity cost of the scarce machining resource. The
breakeven contribution for Product B is $20 (opportunity cost for the machine constraint times two
hours, the machining time for Product B).
8
58
Price-Quantity Determination
191-093
We have seen that the firm should manufacture Product B only if the unit contribution is $20
or greater. Suppose the unit variable cost of Product B is $6 (there is also a direct fixed cost; we will
consider it later). Then, machining capacity should be diverted from Product A to Product B only if
each unit of the latter can be sold for more than $26.
Of course, the firm can charge a higher price. To determine the optimal price, the decision
maker should work with the demand curve for Product B and use the procedure described in Section
3. Suppose this price is $36 and the corresponding optimal quantity 600 units. Then, the maximum
contribution that the firm can realize from Product B is $18,000 ($36, selling price, minus $6, unit
variable cost, times 600, quantity).
Making 600 units of Product B will take 1,200 hours of machine time (600 units times two, the
machining time per unit). The firm has 2,000 hours of machine time at its disposal. Should it use the
balance 800 hours to make more Product B? The answer is “No”: 600 is the optimal production
quantity—the level at which contribution is maximized.
Should the remaining 800 hours of machine time be used to produce Product A, the demand
for which is unlimited? In 800 hours, the firm can machine 200 units of Product A (800 divided by
four, the machining time for each unit of the product), and sell these units for a total contribution of
$8,000 (200 units times $40, the unit contribution for Product A). Unfortunately, the direct fixed cost
for Product A is $15,000. And $8,000 is less than $15000! The conclusion: the firm should not use the
balance time to make any Product A.
The analysis is not yet complete. We have not answered the basic question: Should the firm
make the new product, Product B? Here is where the direct fixed cost for Product B is relevant.
Clearly, Product B should not be made if the direct fixed cost for the product exceeds $18,000, the
maximum contribution from the product. Suppose the direct fixed cost for Product B is less than
$18,000. Should the firm make Product B? It depends. Remember that the firm is already making
$20,000 in contribution from Product A, for a net cash flow of $5,000 after paying for its direct fixed
cost. Clearly, Product B should only be made if the net cash flow realized from it will exceed $5,000.
For this, the direct fixed cost for Product B must be less than $13,000 ($18,000, the contribution from
600 units of the product sold at a price of $36, minus $5,000, the existing cash flow from Product A).
It would be worth your while repeating the above analysis with a different set of numbers.
Explore alternative scenarios: When should the firm make Product A alone? Product B alone? Both
Product A and Product B? The answers to these questions will depend on the machining times for the
two products, the total amount of machine time available, the respective variable and fixed costs,
Product A’s selling price, and the contribution-maximizing price and quantity for Product B.
5.
Competition
How does competition affect the firm’s choice of price and quantity? In general, this is a
difficult question to answer with any degree of analytical precision. Still, formal—albeit stylized—
analysis often helps, offering insight and indicating beneficial courses of action for the firm.
To be sure, the competitive decision context is usually quite complex: the different
competitors may have different costs, capacities, and market strengths; their products may be
differentiated; they may have imperfect information about each other; they may differ in their
assessment of consumer demand; and they may be unwilling or unable (for antitrust reasons?) to
coordinate their actions. The analysis, interesting (and exasperating!) as it can be in the “singlemove” case (where each competitor decides independently on price and quantity and lives with the
decision forever), can get more even complicated in the “multi-move” case (where one firm decides,
the other responds, the first firm counter-responds, the second firm counter-counter-responds, and so
on and so forth). The multi-move model is particularly interesting because it gives the competitors
9
59
191-093
Price-Quantity Determination
repeated opportunities to learn about each other and about each other’s information, costs, strategies,
etc.; it also allows competitors to signal to each other—nudging prices up or down, punishing firms
for not cooperating, rewarding through cooperative moves, etc. In both the single- and the multimove cases, the firm may need to go through a fair amount of “I-think-that-you-think-that-I-thinkthat-you-think-...” type of analysis when determining price and quantity.
Even then, the final
outcome might be unpredictable.
Our goal is somewhat modest. We will examine an “industry” with two competitors, Firm A
and Firm B, offering similar but not identical products (if the products were identical, all the demand
would go to the competitor with the lowest price—provided, of course, it has the capacity to meet the
demand). The competitors have different variable costs; for simplicity, we assume that neither firm
has any fixed costs. Each competitor has adequate capacity to meet total market demand. The two
firms have the same objective—to maximize contribution for the firm—and have complete
information about each other’s costs and capacities.
At what price should each firm offer its product, and what quantity should it expect to sell?
The answer to this question depends on two important factors: (1) the dynamics of the industry (Are
both firms already in the industry, or is one an incumbent and the other a new entrant? Do both
decide at the same time, or does one firm take the leadership, and the other follows? Do the two
firms decide only once, or repeatedly in a sequence of moves and countermoves? Et cetera.), and (2)
the way in which the two firms share the total demand.
We consider the second factor first. The two firms offer somewhat differentiated products.
Accordingly, it is reasonable to anticipate that they will not split the total demand equally even if
their products are offered at identical prices; indeed, because of product superiority and consumer
loyalty, a firm may enjoy a significant market share even if its product is more expensive than its
competitor’s product. In the general case, what is the relationship between total demand, market
share, and the two firms’ prices? This is usually a very difficult question to answer, and firms expend
a lot of resources (money, time, and people) trying to understand how consumers do or will react to
price and product differences. For our purpose, let us assume that a total demand-market share-price
difference model exists for the products in question. Both firms have access to this model.
Let us turn to the first factor, industry dynamics. We begin by considering the case where
Firm A is the leader and Firm B the follower. Once the two firms have chosen prices, they must live
with their decisions forever. How should each firm set its price?
We know that both firms maximize contribution. We also know that each firm is aware of its
own variable cost and the variable cost of the competing firm. Finally, each firm knows what the total
market demand and its own market share will be given its price and its competitor’s price. In this
context, each firm can construct what is called a payoff matrix: a table of contributions for the two
firms as a function of their prices. Each row in the table represents a price chosen by Firm A; each
column, a price chosen by Firm B. There are two entries in each cell of the table: one for Firm A’s
contribution and one for Firm B’s contribution (for the values of Firm A’s and Firm B’s prices to
which the cell corresponds).
What does Firm A do? It says to itself: “Firm B is not stupid. For any price I choose, Firm B
is going to choose a price that maximizes its contribution given my price. Each row of the payoff
matrix represents a choice of price on my part, and the different cells in the row give values of my
contribution and Firm B’s contribution for each price that Firm B might choose. Firm B is going to
choose a price that maximizes its contribution. So, in each row, I should look for the cell in which
Firm B’s contribution is maximized. That cell also gives my contribution if I were to choose the price
represented by that row and Firm B were to choose its contribution-maximizing price given my price.
What price should I choose? I should choose the price that maximizes my contribution given that
Firm B chooses a price that maximizes its contribution given my price.”
10
60
Price-Quantity Determination
191-093
An example will help. Let us say Firm A has a variable cost of $1 per unit; Firm B, a variable
cost of $2 per unit. The two firms are restricted to only two choices of prices—$5 and $6—and total
market demand does not depend on the individual firm’s prices (in reality, overall lower industry
prices will stimulate additional consumer demand). The total market demand is 1,000 units. Firm A
has a superior product; it expects a 60% market share if it and Firm B offer products at the same price,
a 100% share if its price is $5 and Firm B’s price is $6 (i.e., Firm A can expect a share shift from Firm B;
the shift comes at a price, however: the unit contribution is diluted because of the lower price), and a
20% share if its price is $6 and Firm B’s price is $5. The payoff matrix is reproduced in Table 4.
Table 4
Payoff Matrix for Firm A and Firm B
Firm B’s Price
$5.00
Firm A’s Price
$6.00
$5.00
$2,400; $1,200
$4,000; $0
$6.00
$1,000; $2,400
$3,000; $1,600
Note: The first number in each cell of the “2 x 2” payoff matrix is Firm A’s contribution; the second,
Firm B’s contribution.
Firm A is the price leader; Firm B, the follower. Firm A says to itself: “If I choose a price of
$5, Firm B, which I believe is of sound mind and quite rational, is bound to choose $5 too—because its
contribution is higher for a price of $5 than for a price of $6. In that case, my contribution would be
$2,400. On the other hand, if I choose $6, Firm B would still choose $5—once again, because its
contribution is higher at that price than with a price of $6. In that case, my contribution would be
only $1,000, a good deal less than $2,400. Therefore, I should choose a price of $5.”
What about Firm B, the follower? If Firm A chooses $5, Firm B maximizes its contribution by
also choosing a price of $5.
Since both firms choose the same price, Firm A gets a 60% share of the market; Firm B, 40%.
Thus, Firm A can expect a demand of 600; Firm B, 400.
Given our assumption of the dynamics, the two firms have to live with these decisions
forever. What if we allowed Firm A to respond to Firm B’s response to its lead. Now Firm A can
examine its alternatives given Firm B’s price ($5) and respond with a price that maximizes
contribution. In our example, that price happens to be $5—the price Firm A had chosen in the first
place. So, Firm A does not wish to change its price. We have an interesting situation here: Firm A,
the leader, is satisfied with its choice given the follower’s response. Since Firm A does not wish to
counter-respond, Firm B has no need to counter-counter-respond. In other words, the original prices
represent an equilibrium.
This need not be the case in general. For different economics and total demand-market shareprice relationships, it is possible that Firm A might have chosen a price other than $5 in a counterresponse to Firm B’s response, and Firm B a price other than $5 in a counter-counter-response. The
two players will play a game over time. The game may or may not have an equilibrium.
Of course, there is no need to actually play the game to determine the equilibrium—if it
exists. In our original leader-follower model, we noted that Firm A chooses a price to maximize its
contribution assuming Firm B chooses a price to maximize its contribution given Firm A’s price. In
doing so, Firm B was being myopic. Instead, it could have thought about the possibility of Firm A
counter-responding to its price, and incorporated that counter-response in its decision. This is fine,
expect that Firm A’s counter-response depends on an analysis of Firm B’s counter-counter-response,
and the latter hinges on Firm A’s counter-counter-counter-response, and so on and so forth.
11
61
191-093
Price-Quantity Determination
Returning to our example, for the [A:$5, B:$5] price pair, Firm A has a contribution of $2,400,
and Firm B has a contribution of $1,200. This is rather unfortunate: both firms would prefer the price
pair [A:$6, B:$6], where Firm A will have a contribution of $3,000; Firm B, $1,600. The problem, of
course, is that it is difficult to get to this price pair: if Firm A chooses a price of $6, Firm B can do
better by choosing a price of $5 (its contribution will be $2,400), in which case Firm A is unhappy (its
contribution is only $1,000). Similarly, if Firm B chooses a price of $6, Firm A can do better by
choosing a price of $5 (its contribution will be $4,000), in which case Firm B is unhappy (its
contribution drops to zero). So, the two firms are in a bind. Both prefer the price pair [A:$6, B:$6]
over the pair [A:$5, B:$5], but how can they get there? To get there, they need cooperation and
credibility (otherwise, each has the incentive to defect—unless the other firm can offer a credible threat
to punish the defector).
In the microeconomics literature, this situation is often referred to as the prisoner’s dilemma. In
this version of the game, there are two prisoners, Prisoner A and Prisoner B, who are suspected of
having committed a crime. The prisoners are kept in separate cells and questioned by the local
sheriff, who goes first to Prisoner A and asks him to fess up and threatens him with a severe penalty
if Prisoner B confesses but Prisoner A does not. Then, the sheriff goes to Prisoner B and asks her to
fess up and threatens her with a severe penalty if Prisoner A confesses but Prisoner B does not. If
both prisoners confess, they will be thrown into the slammer, but the sentence would not be as severe.
Finally, if neither confesses, the sheriff will file some minor charge against them (for the trouble
caused!) and the two will receive only a minor reprimand. Draw up the payoff matrix for this case
and convince yourself it has the same structure as the matrix in Table 4. Will both prisoners confess?
Most probably, yes.
All payoff matrices do not have the “prisoner’s dilemma” form; a different set of economics
and market sharing rule can result in a matrix with a different structure and a very different set of
incentives for the two players.
In our example, the total market demand was constant. If we had considered the payoffs, not
in terms of contribution, but market shares, we would have observed that Firm A could only benefit
at Firm B’s expense: Firm A’s gain of market share is Firm B’s loss of market share. Similarly, Firm
B’s gain of share must be accompanied by Firm A’s loss of share. The market share game is, hence, a
zero-sum game. The contribution game, on the other hand, is a nonzero-sum game: by charging higher
prices, both firms can benefit through cooperation. Who loses in that case? You got it: we, the
consumers.
12
62
9-511-028
REV: JULY 26, 2011
THOMAS STEENBURGH
JILL AVERY
Marketing Analysis Toolkit:
Pricing and Profitability Analysis
Introduction
Pricing is one of the most difficult decisions marketers make and the one with the most direct and
immediate impact on the firm’s financial position. Thi s marketing analysis toolkit introduces the
fundamental terminology and calculations associated witth pricing and profitability analysis.
Determining Demand
For most goods, the price of a product determines wh ether customers will buy it; such that as the
price goes up, the quantity demanded by customers go
oes down, and as the price goes down, the
quantity demanded by customers goes up. The re lationship between price and demaand is
graphically represented by a demand curve, which is line ar in its most basic form.
A demand curve plots customers’ probable purchase quantities at various prices. The slope of a
linear demand curve is calculated by the following formu la:
Slope of the demand curve (m) =
c hange in price
change in quantity deemanded
The point at which the demand curve crosses the y -axis indicates the price above which any
customer will not buy a product because it is too expen sive. This indicates the outer boundary of
what customers are willing to pay for the product. If th e firm prices the product at this price, they
will sell zero units.
The point at which the demand curve crosses the x-aaxis indicates the maximum number of units
the firm can sell if the price is zero, the maximum quantiity customers are willing to buy at any price.
It is important to remember that a demand curve is only a model of what will happen in the market
__________________________________________________
_______________ ______________ __________________________________
Professors Thomas Steenburgh and Jill Avery (Simmons School of Management) p repared this note a s the basis for class discussion.
Copyright © 2010, 2011 President and Fellows of Harvard Colleg
e e. To order copi es or request perm ission to reproduce materials, call 1--800-5457685, write Harvard Business School Publishing, Boston, MA 02163, or go to ww w.hbsp.harvard.e du/educators. This publication may not be
digitized, photocopied, or otherwise reproduced, posted, or transmitted, without t he permission of H arvard Business School.
63
511-028
Marketing Analysis Toolkit: Pricing and Profitability Analysis
at different prices, and we should be more willing to trust its predictions for the prices that have been
observed in the market than we would for prices that are either extremely high or extremely low.
12,000
maximum reservation price “b”
11,000
10,000
At a price of $9,000, consumers are
likely to buy 140 units.
Price in Dollars
9,000
8,000
Demand Curve
7,000
6,000
At a price of $5,000, consumers are
likely to buy 325 units.
5,000
4,000
3,000
2,000
maximum quantity
customers willing to buy
at any price
1,000
0
100
200
300
400
500
Quantity Demanded
Source: Casewriters.
Marketers can graph their demand curves if they have information on how many units customers
will buy at two different price points and if they assume that demand is linear. For example, if
customers will buy 250 units at a price of $30 and 500 units at a price of $10, then the slope of the
demand curve will be:
slope
=
$30-$10
=
-0.08
250-500
Once the slope is known, marketers can calculate the y-intercept point using one of the known
price/volume relationships using the basic formula for a line:
y = m * x + b (where y is the price, x is the quantity demanded, m is the slope of the demand curve,
and b is the y-intercept)
30 = (-0.08 * 250) + y-intercept
2
64
Marketing Analysis Toolkit: Pricing and Profitability Analysis
511-028
Rearranging the terms algebraically yields:
y-intercept = $50
Given the slope and the y-intercept, the demand curve line can be drawn. In this example, the
demand curve equation would be:
y = -0.08 * x + $50
This equation can then be used to calculate any price/quantity combination by substituting in any
price for y and calculating x to understand how many units will be sold at that price. For example,
the maximum quantity consumers are willing to buy at any price (the x-intercept) can be calculated
by substituting a zero in for the y value and solving for x.
0 = -0.08 * x + $50
x=
650
Marketers need to understand how responsive, or elastic, customers’ demand for a product is to a
change in price at a certain point on the demand curve. The price elasticity of demand ratio helps
illuminate this:
Price Elasticity =
of Demand
Percentage change in quantity demanded
Percentage change in price
The price elasticity of demand measures the percentage change in quantity demanded by
consumers as a result of a percentage change in price. Note that this formula usually yields a
negative number. The negative sign is traditionally ignored, as the magnitude of the number is the
focus for interpretation.
Demand is considered “elastic” if consumers’ demand for a product considerably changes when
price is changed by a small amount. Demand is considered “inelastic” if consumers’ demand for the
product hardly changes when price is changed by a small amount. The price elasticity of demand
ratio helps us understand whether our demand is elastic or inelastic. The following chart
summarizes the range of elasticities that exist in the marketplace to guide interpretation:
Elasticity Ratio
Type of Elasticity
Description
E=∞
Perfectly Elastic
Any very small change in price results
in a very large change in quantity
demanded.
1<E<∞
Relatively Elastic
Small changes in price cause
large changes in quantity demanded.
E=1
Unit Elastic
Any change in price is matched by an equal
change in quantity.
0<E<1
Relatively Inelastic
Large changes in price cause
small changes in quantity demanded.
E=0
Perfectly Inelastic
price is changed.
Quantity demanded does not change when
3
65
511-028
Marketing Analysis Toolkit: Pricing and Profitability Analysis
Under relatively elastic demand conditions, quantity is very responsive to price. For example,
suppose the price elasticity equals 2. A 10% increase in price will lead to a 20% drop in the quantity
demanded. Under relatively inelastic demand conditions, the opposite is true, as the quantity
demanded is not very responsive to price. For example, suppose the price elasticity is 0.5. In this
case, a 10% drop in price will lead to only 5% increase in the quantity demanded.
Inelastic demand curves have a steeper slope and are therefore more vertical than elastic demand
curves, indicating that the quantity demanded by consumers does not change very much when the
price is increased or decreased.
12,000
12,000
10,000
10,000
Price in Dollars
8,000
7,000
6,000
At a price of $5,000, consumers are
likely to buy 325 units.
5,000
8,000
7,000
6,000
4,000
3,000
3,000
2,000
2,000
1,000
1,000
100
200
300
400
At a price of $5,000, consumers are
likely to buy 210 units.
5,000
4,000
0
At a price of $9,000, consumers are
likely to buy 180 units.
9,000
At a price of $9,000, consumers are
likely to buy 140 units.
9,000
Price in Dollars
Inelastic Demand Curve
11,000
Elastic Demand Curve
11,000
0
500
100
200
300
400
Quantity Demanded
Quantity Demanded
Source: Casewriters.
Marketers can attempt to shift their product’s demand curve outwards (i.e. creating higher levels
of quantities demanded at the same price point) by increasing customer preference for the product
through branding or other marketing initiatives that increase consumers’ desire for the product.
Demand curves can also shift outwards when substitute products offered by competitors become
more expensive, or when consumers’ incomes increase. However, demand curves can also shift the
other way (i.e. lower quantity demanded at the same price point) if competitors offer compelling
substitutes or consumers’ incomes decrease.
Technical Note: If we assume that demand for a product is linear (i.e. a straight line with constant
slope across all price/quantity combinations, then the price elasticity of demand changes at each
price point along the demand curve. As the focal price decreases, the price elasticity of demand gets
smaller (in absolute value).
An example, which is relatively easy to calculate for a linear demand curve, helps make this idea
concrete. Continuing with the previous demand curve, suppose that the product was initially priced
at $30. We want to determine the percentage change in demand that occurs following a percentage
change in price from $30. The demand curve equation predicts that the company would sell 250 units
if the company were to set its price at $30. Furthermore, if the company was to set its price at $20, the
demand curve equation predicts that it would sell the following quantity, calculated using the linear
demand formula outlined above:
4
66
500
Marketing Analysis Toolkit: Pricing and Profitability Analysis
$20
511-028
= -0.08(x) + $50
x = 375 units
Given this, the demand model predicts that a 33% drop in price (from $30) would result in a 50%
increase in demand. Therefore, the price elasticity of demand at $30 is:
Price Elasticity =
of Demand
(375-250)/250
($20-$30)/$30
=
50%
-33%
=
-1.5
In contrast, suppose the initial price point was $20 instead of $30. The demand curve equation
predicts that the firm would sell 375 units at a price point of.$20 and 500 units at a price point of $10.
Therefore, a 50% drop in price (from $20) would result in a 25% increase in demand. Therefore, the
price elasticity of demand at $20 is:
Price Elasticity =
of Demand
(500-375)/375
($10-$20)/$20
=
33%
-50%
=
-0.67
For most demand curves, like the linear demand curve with which we are working, the price
elasticity of demand takes different values depending on price. This, however, is not always the case.
The constant elasticity demand curve, which is slightly more difficult to work with because it takes
the form of an arc, has the same elasticity regardless of price.
Calculating Total Revenue
Once a manager measures her product’s demand curve and its resulting price elasticity of
demand, she can begin to calculate the total revenue the firm will achieve at various price points.
Total revenue measures the amount of money coming into the firm from the sale of its products. It is
based on the total number of units sold to customers (quantity) and the price at which each unit is
sold.
Total Revenue = Price per unit * Quantity Sold
For firms who sell directly to consumers, the price used in the Total Revenue calculation is the
retail price at which the consumer purchases the product. However, for firms who do not sell
directly to consumers but rather sell their products through a retailer, the price used in the Total
Revenue calculation is the price at which the firm sells its products to its distribution channel
partners.
Calculating Total Costs
A major part of any pricing decision is analyzing the costs of the product. The product’s demand
curve sets a ceiling on the price that the firm can charge for its product, and the cost structure of the
firm sets a floor. In general, firms want to charge a price that covers its costs of producing,
distributing, and selling the product, including a fair return for its effort and risk.
There are two types of costs incurred by the firm: fixed costs and variable costs. Fixed costs do
not change as the quantity of units produced and sold changes. Fixed costs usually include things
like factory and equipment costs, management salaries, and advertising. Variable costs vary as the
5
67
511-028
Marketing Analysis Toolkit: Pricing and Profitability Analysis
unit quantity produced and sold changes. Variable costs are directly linked to the quantity produced
and include things like direct manufacturing labor and raw materials. Variable costs generally
appear in an income statement in the Cost of Good Sold (COGS) line item, while fixed costs generally
appear in line items like Property, Plant, and Equipment (PPE), Selling, General, and Administrative
(SG&A), and Research and Development (R&D). However, please note that the COGS line item
includes variable costs plus a portion of fixed costs associated with making the product. In the U.S.,
Generally Accepted Accounting Principles (GAAP) requires firms to estimate all product costs (fixed
and variable) and allocate them to the number of product units produced. When units are sold, those
allocated costs (both fixed and variable) become the cost of goods sold.
Total costs measure the amount of money the firm has to spend. Total costs are the combination
of fixed and variable costs.
Total Costs = Fixed Costs + Variable Cost per unit * Quantity Sold
Remember, that total costs change with the unit quantity sold, because of variable costs, so if you
are running different quantity scenarios, you need to adjust both Total Revenue and Total Costs.
Calculating Profit
In general, “profit” or “margin” is the difference between the revenues taken in by the firm and
the costs incurred in achieving those revenues. Profit is often calculated as a raw dollar amount and
as a percentage of revenue. In marketing, there are four profit measures that are used frequently:
contribution margin, gross margin, direct marketing contribution, and net income.
Contribution margin is the term used to capture the firm’s top line profitability. Contribution
margin is the difference between its total revenue and its variable costs. While contribution margin is
calculated as a dollar amount, most firms also calculate a contribution margin percentage, which
calculates contribution margin as a percentage of total revenue.
Contribution Margin
=
Total Revenue - (Variable Cost per unit * Quantity Sold)
Contribution Margin % =
Total Revenue - (Variable Cost per unit * Quantity Sold)
Total Revenue
Contribution margin may also be calculated on a per unit basis:
Contribution margin/unit
=
Revenue per unit - Variable Cost per unit
One can also calculate the firm’s gross margin. Gross margin is the difference between the firm’s
total revenue and its cost of goods sold (COGS) and is often reported on the income statement.
Gross Margin
=
Total Revenues - Cost of Goods Sold
Gross Margin %
=
Total Revenues - Cost of Goods Sold
Total Revenues
Gross margin may also be calculated on a per unit basis:
Gross margin/unit
= Revenue per unit - COGS per unit
6
68
Marketing Analysis Toolkit: Pricing and Profitability Analysis
511-028
Direct marketing contribution (DMC) includes variable costs plus fixed costs associated with
marketing expenditures, such as advertising expense and sales promotion expense. Note that direct
marketing contribution generally does not include fixed costs labeled as Selling, General and
Administrative (SG&A on an income statement). Again a direct marketing contribution ratio can be
calculated.
Direct
Marketing
=
Contribution
Total Revenue - (Variable Cost per unit *Quantity) - Marketing Expenses
Direct
Marketing
=
Contribution
%
Total Revenue - (Variable Cost per unit *Quantity) - Marketing Expenses
Total Revenue
Net income is the term used to capture the firm’s bottom line profitability and is often reported on
the income statement. Unlike the contribution margin, net income includes fixed costs in addition to
variable costs. Again, a net income percentage can be calculated too.
Net Income
=
Total Revenue - Total Costs
Net Income %
=
Total Revenue - Total Costs
Total Revenue
Retailer Margins and Penny Profit
Just like firms, retailers and other distribution channel partners use profit and margin analysis to
measure their profitability. A retailer’s “penny profit” calculates the amount of profit a retailer
makes in dollars. The “retailer margin” is the ratio of the amount of profit a retailer makes in dollars
as a percentage of the retail selling price. The “retail selling price” is the amount of money the retailer
receives from a consumer when it sells the item in its store. The “cost to the retailer” is the money
that the retailer pays to the manufacturer to purchase the item for resale to the consumer. Hence, the
“cost to the retailer” becomes the “revenue per unit” in the firm profit calculations above.
Penny Profit
=
Retail Selling Price - Cost to the Retailer
Retailer Margin =
Retail Selling Price - Cost to the Retailer
Retail Selling Price
Notice that penny profit and retailer margin calculations are merely contribution margin/unit
calculations from the retailer’s perspective.
7
69
511-028
Marketing Analysis Toolkit: Pricing and Profitability Analysis
Using Pricing and Profitability Analysis to Guide Marketing Decision
Making
It is crucial when making any marketing decision to understand the financial implications of the
decision for the firm. Whether you are working in a for-profit or a non-profit organization,
profitability matters! Even not-for-profit organizations need to manage their revenues and costs.
Managers should assess changes in profitability in the following situations:

When recommending a price change, whether permanent or temporary (i.e. a temporary
price reduction, coupon promotion, rebate promotion).

When recommending a marketing expense.

When recommending changes to the product which change its variable cost structure.

When recommending changes to the distribution strategy that change the price the firm
will receive from its retailer partners or that change the price consumers will pay for the
product.

When evaluating sales demand from various consumer segments to understand how the
varying sales estimates from different customer segments contribute to profitability. Very
often, customers who buy lots of unit volume get special pricing, which reduces their
profitability to the firm. Customers who buy lower unit quantities but who buy those
units at full price are often more profitable.
8
70
Saturday, October 1
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Saturday, 1 October
08:30 Finance Session 5 – Professor Callahan
Topic: Capital Structure Choices – Putting It All Together
In this session we will use a case discussion to review and summarize
the important tax, risk, information, and agency consequences of a major
change in a firm’s capital structure policy.
Required Reading: Blaine Kitchenware, Inc.: Capital Structure case
Case Discussion Questions:
1. Do you believe Blaine’s capital structure and payout policies are
appropriate? Why or why not?
2. Should Dubinski recommend a large share repurchase to Blaine’s board?
What are the primary advantages and disadvantages of such a move?
3. Consider the following share repurchase proposal: Blaine will use
$209M of cash from its balance sheet and $50M in new interest-bearing
debt at a rate of 6.75% to repurchase 14.0M shares at a price of $18.50
per share. How would such a buyback affect Blaine? Consider the impact
on, among other things, Blaine’s earnings per share and ROE.
4. As a member of Blaine’s controlling family, would you be in favor of this
proposal? Would you be in favor of it as a non-family shareholder?
To do before this session: Prepare the case for discussion
09:45
Break
10:00
Finance Session 6 – Professor Callahan
Topic: Leverage and the Cost of Capital for a Firm or Project/Division
In this session we will learn how to calculate a firm’s or project’s cost of
capital. (They may or may not be the same!) We will discuss the
appropriate cost of debt financing, preferred stock financing, and
common equity financing and how each is commonly derived. We will
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
71
Page: 1 / 3
Global Executive MBA in Shanghai
also look at the interaction between capital structure choices and the
firm’s cost of capital.
Required Readings:
-
RWJ Chapter 13 (review, especially 13.9)
-
RWJ Chapter 18 (with an emphasis on 18.3, 18.5, and 18.7)
Discussion Questions:
1. Intuitively, what does a stock beta represent?
2. What is the difference between a stock beta, a debt beta, and an asset
beta?
3. What are three primary determinants of a company’s stock beta (i.e.,
what determines the relative exposure of a company stock to economy
wide fluctuations)?
4. Why might a project beta or division beta differ from a company’s stock
beta? Give a real world example of a project with the same beta as the
parent company and a real world example of a project with a different
beta than the parent company.
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
11:15
Break
11:30
Finance Session 7 – Professor Callahan
Topic: Estimating the Cost of Capital – An Example
In this session we will apply what we have learned in the previous
session to the Midland Energy Resources, Inc. case and further refine
our financial analysis skills.
Required Reading: Midland Energy Resources, Inc.: Cost of Capital
Discussion Questions:
1. How are Mortensen’s estimates of Midland’s cost of capital used? How, if
at all, should these anticipated uses affect the calculations?
2. Calculate Midland’s corporate WACC. Be prepared to defend your
specific assumptions about the various inputs to the calculations. Is
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
72
Page: 2 / 3
Global Executive MBA in Shanghai
Midland’s choice of EMRP appropriate? If not, what recommendations
would you make and why?
3. Should Midland use a single corporate hurdle rate for evaluating
investment opportunities in all of its divisions? Why or why not?
4. Compute a separate cost of capital for the E&P and Marketing & Refining
divisions. What causes them to differ from one another?
5. How would you compute a cost of capital for the Petrochemical division?
To do before this session: Prepare the case for discussion
12:45
Lunch
13:45
Marketing Session 7 – Professor Nunes
Topic: Saurer: The China Challenge
Required Readings: Saurer: The China Challenge (A) IMD399
Discussion Questions:
1. Should Saurer enter the market for lower functionality twisting machines
in China? What are the major advantages and disadvantages to doing
this?
2. If Saurer were to introduce the proposed machine, what marketing
strategy would you recommend for the product in China?
3. If Saurer were to not introduce the machine, what changes would you
recommend for the company’s current strategy in China?
Optional Reading: Gadiesh, Orit, Philip Leung and Till Vestring: “The Battle for
China’s Good Enough Market”, HBR Reprint R0709E
Due: Exercise 2 – Reading Perceptual Maps
15:00
Break
15:15
Marketing Session 8 – Professor Nunes
Topic: Breakout Sessions
This session will include time to review Exercise 2 before allowing teams
to work on their first practice decision.
Assignment: Make Markstrat Practice Decision
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
73
Page: 3 / 3
4040
OCTOBER 08, 2009
TIMOTHY LUEHRMAN
JOEL HEILPRIN
Blaine Kitchenware, Inc.: Capital Structure
On April 27, 2007, Victor Dubinski, CEO of Blaine Kitchenware, Inc. (BKI), sat in his office
reflecting on a meeting he had had with an investment banker earlier in the week. The banker, whom
Dubinski had known for years, asked for the meeting after a group of private equity investors made
discreet inquiries about a possible acquisition of Blaine. Although Blaine was a public company, a
majority of its shares were controlled by family members descended from the firm’s founders
together with various family trusts. Family interests were strongly represented on the board of
directors as well. Dubinski knew the family had no current interest in selling—on the contrary,
Blaine was interested in acquiring other companies in the kitchen appliances space—so this overture,
like a few others before it, would be politely rebuffed.
Nevertheless, Dubinski was struck by the banker’s assertion that a private equity buyer could
“unlock” value inherent in Blaine’s strong operations and balance sheet. Using cash on Blaine’s
balance sheet and new borrowings, a private equity firm could purchase all of Blaine’s outstanding
shares at a price higher than $16.25 per share, its current stock price. It would then repay the debt
over time using the company’s future earnings. When the banker pointed out that BKI itself could do
the same thing—borrow money to buy back its own shares—Dubinski had asked, “But why would
we do that?” The banker’s response was blunt: “Because you’re over-liquid and under-levered. Your
shareholders are paying a price for that.” In the days since the meeting, Dubinski’s thoughts kept
returning to a share repurchase. How many shares could be bought? At what price? Would it sap
Blaine’s financial strength? Or prevent it from making future acquisitions?
Blaine Kitchenware’s Business
Blaine Kitchenware was a mid-sized producer of branded small appliances primarily used in
residential kitchens. Originally founded as The Blaine Electrical Apparatus Company in 1927, it
produced then-novel electric home appliances, such as irons, vacuum cleaners, waffle irons, and
cream separators, which were touted as modern, clean, and easier to use than counterparts fueled by
oil, coal, gas, or by hand. By 2006, the company’s products consisted of a wide range of small kitchen
appliances used for food and beverage preparation and for cooking, including several branded lines
of deep fryers, griddles, waffle irons, toasters, small ovens, blenders, mixers, pressure cookers,
steamers, slow cookers, shredders and slicers, and coffee makers.
________________________________________________________________________________________________________________
HBS Professor Timothy A. Luehrman and Illinois Institute of Technology Adjunct Finance Professor Joel L. Heilprin prepared this case solely as
a basis for class discussion and not as an endorsement, a source of primary data, or an illustration of effective or ineffective management.
This case, though based on real events, is fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional
references to actual companies in the narration.
Copyright © 2009 Harvard Business School Publishing. To order copies or request permission to reproduce materials, call 1-800-545-7685, write
Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored
in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or
otherwise—without the permission of Harvard Business Publishing.
Harvard Business Publishing is an affiliate of Harvard Business School.
74
4040 | Blaine Kitchenware, Inc.: Capital Structure
Blaine had just under 10% of the $2.3 billion U.S. market for small kitchen appliances. For the
period 2003–2006 the industry posted modest annual unit sales growth of 2% despite positive market
conditions including a strong housing market, growth in affluent householders, and product
innovations. Competition from inexpensive imports and aggressive pricing by mass merchandisers
limited industry dollar volume growth to just 3.5% annually over that same period. Historically, the
industry had been fragmented, but it had recently experienced some consolidation that many
participants expected to continue.
In recent years, Blaine had been expanding into foreign markets. Nevertheless in 2006, 65% of its
revenue was generated from shipments to U.S. wholesalers and retailers, with the balance coming
from sales to Canada, Europe, and Central and South America. The company shipped approximately
14 million units a year.
There were three major segments in the small kitchen appliance industry: food preparation
appliances, cooking appliances, and beverage-making appliances. Blaine produced product for all
three, but the majority of its revenues came from cooking appliances and food preparation
appliances. Its market share of beverage-making appliances was only 2%. Most of BKI’s appliances
retailed at medium price points, at or just below products offered by the best-known national brands.
BKI’s market research consistently showed that the Blaine brand was well-known and well-regarded
by consumers. It was associated somewhat with “nostalgia” and the creation of “familiar,
wholesome dishes.”
Recently, Blaine had introduced some goods with “smart” technology features and sleeker styling,
targeting higher-end consumers and intended to compete at higher price points. This strategy was in
response to increased competition from Asian imports and private label product. The majority of
BKI’s products were distributed via a network of wholesalers, which supplied mass merchandisers
and department stores, but its upper-tier products were sold directly to specialty retailers and
catalogue companies. Regardless of the distribution channel, BKI offered consumers standard
warranty terms of 90 days to one year, depending on the appliance.
Blaine’s monthly sales reached a seasonal peak during October and November as retailers
increased stock in anticipation of the holiday season. A smaller peak occurred in May and June,
coinciding with Mother’s Day, a summer surge in weddings, and the seasonal peak in home
purchases. Historically, sales of Blaine appliances had been cyclical as well, tending to track overall
macroeconomic activity. This also was the case for the industry as a whole; in particular, changes in
appliance sales were correlated with changes in housing sales and in home renovation and household
formation.
BKI owned and operated a small factory in Minnesota that produced cast iron parts with specialty
coatings for certain of its cookware offerings. Otherwise, however, Blaine, like most companies in the
appliance industry, outsourced its production. In 2006 BKI had suppliers and contract manufacturers
in China, Vietnam, Canada, and Mexico.
Victor Dubinski was a great-grandson of one of the founders. An engineer by training, Dubinski
served in the U.S. Navy after graduating from college in 1970. After his discharge, he worked for a
large aerospace and defense contractor until joining the family business in 1981 as head of operations.
He was elected to the board of directors in 1988 and became Blaine’s CEO in 1992, succeeding his
uncle.
Under Dubinski’s leadership, Blaine operated much as it always had, with three notable
exceptions. First, the company completed an IPO in 1994. This provided a measure of liquidity for
certain of the founders’ descendants who, collectively, owned 62% of the outstanding shares
2
BRIEFCASES | HARVARD BUSINESS PUBLISHING
75
Blaine Kitchenware, Inc.: Capital Structure | 4040
following the IPO. Second, beginning in the 1990s, Blaine gradually moved its production abroad.
The company began by taking advantage of NAFTA, engaging suppliers and performing some
manufacturing in Mexico. By 2003, BKI also had established relationships with several Asian
manufacturers, and the large majority of its production took place outside the United States. Finally,
BKI had undertaken a strategy focused on rounding out and complementing its product offerings by
acquiring small independent manufacturers or the kitchen appliance product lines of large
diversified manufacturers. The company carefully followed changes in customer purchasing
behavior and market trends. Victor Dubinski and the board were eager to continue what they
believed had been a fruitful strategy. The company was particularly keen to increase its presence in
the beverage appliance segment, which demonstrated the strongest growth and where BKI was
weakest. Thus far, all acquisitions had been for cash or BKI stock.
Financial Performance
During the year ended December 31, 2006, Blaine earned net income of $53.6 million on revenue
of $342 million. Exhibits 1 and 2 present the company’s recent financial statements. Approximately
85% of Blaine’s revenue and 80% of its operating income came from the sale of mid-tier products,
with the line of higher-end goods accounting for the remainder. The company’s 2006 EBITDA
margin of nearly 22% was among the strongest within the peer group shown in Exhibit 3. Despite its
recent shift toward higher-end product lines, Blaine’s operating margins had decreased slightly over
the last three years. Margins declined due to integration costs and inventory write-downs associated
with recent acquisitions. Now that integration activities were completed, BKI executives expected the
firm to achieve operating margins at least as high as its historical margins.
The U.S. industry as a whole faced considerable pressure from imports and private label products,
as well as a shift in consumer purchasing preferences favoring larger, “big box” retailers. In response,
some of Blaine’s more aggressive rivals were cutting prices to maintain sales growth. Blaine had not
followed suit and its organic revenue growth had suffered in recent years, as some of its core
products lost market share. Growth in Blaine’s top line was attributable almost exclusively to
acquisitions.
Despite the company’s profitability, returns to shareholders had been somewhat below average.
Blaine’s return on equity (ROE), shown below, was significantly below that of its publicly traded
peers.1 Moreover, its earnings per share had fallen significantly since 2004, partly due to dilutive
acquisitions.
Companies
2006 ROE
Home & Hearth Design
AutoTech Appliances
XQL Corp.
Bunkerhill Incorporated
EasyLiving Systems
Mean
11.3%
43.1%
19.5%
41.7%
13.9%
25.9%
Median
19.5%
Blaine
11.0%
1 ROE is computed here as net income divided by end-of-period book equity.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
76
3
4040 | Blaine Kitchenware, Inc.: Capital Structure
During 2004–2006, compounded annual returns for BKI shareholders, including dividends and
stock price appreciation, were approximately 11% per year. This was higher than the S&P 500, which
returned approximately 10% per year. However, it was well below the 16% annual compounded
return earned by shareholders of Blaine’s peer group during the same period.
Financial Policies
Blaine’s financial posture was conservative and very much in keeping with BKI’s long-standing
practice and, indeed, with its management style generally. Only twice in its history had the company
borrowed beyond seasonal working capital needs. The first time was during World War II, when it
borrowed from the U.S. government to retool several factories for war production. The second time
was during the first oil shock of the 1970s. On both occasions the debt was repaid as quickly as
possible.
At the end of 2006, Blaine’s balance sheet was the strongest in the industry. Not only was it debtfree, but the company also held $231 million in cash and securities at the end of 2006, down from $286
million two years earlier. Given such substantial liquidity, Blaine had terminated in 2002 a revolving
credit agreement designed to provide standby credit for seasonal needs; the CFO argued that the fees
were a waste of money and Dubinski agreed.
In recent years the company’s largest uses of cash had been common dividends and cash
consideration paid in various acquisitions. Dividends per share had risen only modestly during
2004–2006; however, as the company issued new shares in connection with some of its acquisitions,
the number of shares outstanding climbed, and the payout ratio rose significantly, to more than 50%
in 2006.
2004
Net income
Dividends
Average shares outstanding
Earnings per share
Dividend per share
Payout ratio
$ 53,112
$ 18,589
41,309
$ 1.29
$ 0.45
35.0%
2005
$ 52,435
$ 22,871
48,970
$ 1.07
$ 0.47
43.6%
2006
$ 53,630
$ 28,345
59,052
$ 0.91
$ 0.48
52.9%
The next largest use of funds was capital expenditures, which were modest due to Blaine’s
extensive outsourcing of its manufacturing. Average capital expenditures during the past three years
were just over $10 million per year. While they were expected to remain modest, future expenditures
would be driven in part by the extent and nature of Blaine’s future acquisitions. In recent years,
after-tax cash generated from operations had been more than four times average capital expenditures
and rising, as shown in the table below.
EBITDA
Less: Taxes
After-Tax Operating Cash Flow
4
2004
2005
$ 69,370
24,989
44,380
$ 68,895
24,303
44,592
2006
$ 73,860
23,821
50,039
AVG.
46,337
BRIEFCASES | HARVARD BUSINESS PUBLISHING
77
Blaine Kitchenware, Inc.: Capital Structure | 4040
Reassessing Financial Policies in 2007
In 2007 Blaine planned to continue its policy of holding prices firm in the face of competitive
pressures. Consequently, its managers were expecting top line growth of only 3% for fiscal year 2007.
However, this growth rate assumed no acquisitions would be made in 2007, unlike the previous two
years. While the board remained receptive to opportunities, Dubinski and his team had no target in
mind as yet at the end of April.
As he reflected on the possibility of repurchasing stock, Dubinski understood that he could
consider such a move only in conjunction with all of BKI’s financial policies: its liquidity, capital
structure, dividend policy, ownership structure, and acquisition plans. In addition, he wondered
about timing. Blaine’s stock price was not far off its all-time high, yet its performance clearly lagged
that of its peers. A summary of contemporaneous financial market information is provided in
Exhibit 4.
Dubinski had begun to suspect that family members on the board would welcome some of the
possible effects of a large share repurchase. Assuming that family members held on to their shares,
their percentage ownership of Blaine would rise, reversing a downward trend dating from BKI’s IPO.
It also would give the board more flexibility in setting future dividends per share. Both Dubinski and
the board knew that the recent trend in BKI’s payout ratio was unsustainable and that this concerned
some family members.
On the other hand, a large repurchase might be unpopular if it forced Blaine to give up its war
chest and/or discontinue its acquisition activity. Perhaps even more unsettling, it would cause
Blaine to borrow money. The company would be paying significant interest expense for only the
third time in its history. As Dubinski turned his chair to face the window, he glanced at the framed
photo behind his desk of his great grandfather, Marcus Blaine, demonstrating the company’s first
cream separator—its best-selling product during Blaine’s first decade. A real Blaine Electrical Cream
Separator sat in a glass case in the corner; the last one had been manufactured in 1949.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
78
5
4040 | Blaine Kitchenware, Inc.: Capital Structure
Exhibit 1
Blaine Kitchenware, Inc., Income Statements, years ended December 31, ($ in Thousands)
Operating Results
2004
2005
2006
$291,940
204,265
$307,964
220,234
$342,251
249,794
Gross Profit
Less: Selling, General & Administrative
87,676
25,293
87,731
27,049
92,458
28,512
Operating Income
Plus: Depreciation & Amortization
62,383
6,987
60,682
8,213
63,946
9,914
EBITDA
69,370
68,895
73,860
EBIT
Plus: Other Income (expense)
62,383
15,719
60,682
16,057
63,946
13,506
Earnings Before Tax
Less: Taxes
78,101
24,989
76,738
24,303
77,451
23,821
53,112
$ 18,589
52,435
$ 22,871
53,630
$ 28,345
5.5%
11.1%
Revenue
Less: Cost of Goods Sold
Net Income
Dividends
Margins
Revenue Growth
3.2%
Gross Margin
30.0%
28.5%
27.0%
EBIT Margin
21.4%
19.7%
18.7%
21.6%
EBITDA Margin
23.8%
22.4%
Effective Tax Ratea
32.0%
31.7%
30.8%
Net Income Margin
15.7%
14.7%
14.2%
Dividend payout ratio
35.0%
43.6%
52.9%
a.
6
Blaine's future tax rate was expected to rise to the statutory rate of 40%.
BRIEFCASES | HARVARD BUSINESS PUBLISHING
79
Blaine Kitchenware, Inc.: Capital Structure | 4040
Exhibit 2
Blaine Kitchenware, Inc. Balance Sheets, December 31, ($ in Thousands)
Assets
2004
2005
2006
Cash & Cash Equivalents
$ 67,391
$ 70,853
$ 66,557
Marketable Securities
218,403
196,763
164,309
Accounts Receivable
40,709
43,235
48,780
Inventory
47,262
49,728
54,874
Other Current Assets
2,586
3,871
5,157
Total Current Assets
376,351
364,449
339,678
99,402
138,546
174,321
Property, Plant & Equipment
8,134
20,439
38,281
Other Assets
Goodwill
13,331
27,394
39,973
Total Assets
$497,217
$550,829
$592,253
$ 26,106
$ 28,589
$ 31,936
Liabilities & Shareholders' Equity
Accounts Payable
Accrued Liabilities
22,605
24,921
27,761
Taxes Payable
14,225
17,196
16,884
Total Current Liabilities
62,935
70,705
76,581
1,794
3,151
4,814
Other liabilities
Deferred Taxes
15,111
18,434
22,495
Total Liabilities
79,840
92,290
103,890
Shareholders' Equity
Total Liabilities & Shareholders' Equity
Note:
417,377
458,538
488,363
$497,217
$550,829
$592,253
Many items in BKI’s historical balance sheets (e.g., Property, Plant & Equipment) have been affected by the firm’s
acquisitions.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
80
7
81
$ 350,798
372,293
475,377
Net debtb
Total debt
Book equity
45.18%
31.12%
Net Debt/Equity
Net Debt/Enterprise Value
b. Net debt is total long-term and short-term debt less excess cash.
a. Net working capital excludes cash and securities.
1.91x
10.56x
9.46x
1.63x
1.03
LTM Trading Multiples
MVIC/Revenue
MVIC/EBIT
MVIC/EBITDA
Market/Book equity
Equity beta
776,427
$1,127,226
$ 21,495
54,316
900,803
$ 976,613
Cash & securities
Net working capitala
Net fixed assets
Total assets
Market capitalization
Enterprise value (MVIC)
$ 589,747
106,763
119,190
$ 53,698
Home &
Hearth Design
31.74%
24.10%
1.02x
7.35x
6.03x
4.26x
1.24
13,978,375
$18,415,689
$4,437,314
4,973,413
3,283,000
$ 536,099
1,247,520
7,463,564
$9,247,183
$18,080,000
2,505,200
3,055,200
$1,416,012
AutoTech
Appliances
17.97%
15.23%
1.45x
8.65x
7.84x
2.51x
0.96
5,290,145
$6,240,947
$ 950,802
972,227
2,109,400
$ 21,425
353,691
3,322,837
$3,697,952
$4,313,300
721,297
796,497
$ 412,307
XQL Corp.
6.01%
5.67%
1.14x
7.42x
6.88x
4.93x
0.92
3,962,780
$4,200,836
$ 238,056
391,736
804,400
$ 153,680
334,804
815,304
$1,303,788
$3,671,100
566,099
610,399
$ 335,073
Bunkerhill, Inc.
-15.47%
-18.31%
1.87x
18.05x
15.15x
4.41x
0.67
418,749
$ 353,949
$ (64,800)
177,302
94,919
$ 242,102
21,220
68,788
$ 332,110
$ 188,955
19,613
23,356
$ 13,173
EasyLiving
Systems
Selected Operating and Financial Data for Public Kitchenware Producers, 12 months ended December 31, 2006, ($ in Thousands)
Revenue
EBIT
EBITDA
Net income
Exhibit 3
-24.06%
-31.68%
2.13x
11.40x
9.87x
1.96x
0.56
959,596
$ 728,730
$(230,866)
488,363
$ 230,866
32,231
174,321
$ 592,253
$ 342,251
63,946
73,860
$ 53,630
Blaine
Kitchenware
4040
-8-
Blaine Kitchenware, Inc.: Capital Structure | 4040
Exhibit 4
Contemporaneous Capital Market Data (April 21, 2007)
Yields on U.S. Treasury Securities
Maturity
30 days
60 days
90 days
1 year
5 years
10 years
20 years
30 years
4.55%
4.73%
4.91%
4.90%
4.91%
5.02%
5.26%
5.10%
Seasoned corporate bond yields
Moody's Aaa
5.88%
Aa
6.04%
A
6.35%
Baa
6.72%
Ba
7.88%
B
8.94%
Default spread
0.86%
1.02%
1.33%
1.70%
2.86%
3.92%
HARVARD BUSINESS PUBLISHING | BRIEFCASES
82
9
4129
JUNE 19, 2009
TIMOTHY A. LUEHRMAN
JOEL L. HEILPRIN
Midland Energy Resources, Inc.: Cost of Capital
In late January 2007, Janet Mortensen, senior vice president of project finance for Midland Energy
Resources, was preparing her annual cost of capital estimates for Midland and each of its three
divisions. Midland was a global energy company with operations in oil and gas exploration and
production (E&P), refining and marketing (R&M), and petrochemicals. On a consolidated basis, the
firm had 2006 operating revenue and operating income of $248.5 billion and $42.2 billion,
respectively.
Estimates of the cost of capital were used in many analyses within Midland, including asset
appraisals for both capital budgeting and financial accounting, performance assessments, M&A
proposals, and stock repurchase decisions. Some of these analyses were performed at the division or
business unit level, while others were executed at the corporate level. Midland’s corporate treasury
staff had begun preparing annual cost of capital estimates for the corporation and each division in the
early 1980s. The estimates produced by treasury were often criticized, and Midland’s division
presidents and controllers sometimes challenged specific assumptions and inputs.
In 2002, Mortensen, then a senior analyst reporting to the CFO, was asked to estimate Midland’s
cost of capital in connection with a large proposed share repurchase. Six months later she was asked
to calculate corporate and divisional costs of capital that the executive and compensation committees
of the board could incorporate in planned performance evaluations. Since then, Mortensen had
undertaken a similar exercise each year and her estimates had become widely circulated de facto
standards in many analyses throughout the company, even ones in which they were not formally
required. By 2007 Mortensen was aware that her calculations had become influential and she
devoted ever more time and care to their preparation. Lately she wondered whether they were
actually appropriate for all applications and she was considering appending a sort of “user’s guide”
to the 2007 set of calculations.
Midland’s Operations
Midland Energy Resources had been incorporated more than 120 years previously and in 2007
had more than 80,000 employees. Exhibits 1 and 2 present Midland’s most recent consolidated
financial statements. Exhibit 3 presents selected business segment data for the period 2004-06.
________________________________________________________________________________________________________________
HBS Professor Timothy A. Luehrman and Illinois Institute of Technology Adjunct Finance Professor Joel L. Heilprin prepared this case
specifically for the Harvard Business Publishing Brief Case Collection. Though inspired by real events, the case does not represent a specific
situation at an existing company, and any resemblance to actual persons or entities is unintended. Cases are developed solely as the basis for
class discussion and are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management.
Copyright © 2009 Harvard Business Publishing. To order copies or request permission to reproduce materials, call 1-800-545-7685 or go to
www.hbsp.harvard.edu/educators. This publication may not be digitized, photocopied, or otherwise reproduced, posted, or transmitted,
without the permission of Harvard Business Publishing.
Harvard Business Publishing is an affiliate of Harvard Business School.
83
4129 | Midland Energy Resources, Inc.: Cost of Capital
Exploration & Production
Midland engaged in all phases of exploration, development, and production, though the last of
these, production, dominated the E&P division’s reported operating results. During 2006, Midland
extracted approximately 2.10 million barrels of oil per day—a 6.3% increase over 2005 production—
and roughly 7.28 billion cubic feet of natural gas per day—an increase of slightly less than 1% over
2005. This represented $22.4 billion of revenue and after-tax earnings of $12.6 billion. E&P was
Midland’s most profitable business, and its net margin over the previous five years was among the
highest in the industry.
Midland expected continued global population and economic growth to result in rising demand
for its products for the foreseeable future. Nevertheless, the fraction of production coming from nontraditional sources such as deepwater drilling, heavy oil recovery, liquefied natural gas (LNG), and
arctic technology was expected to increase. Further, the geographic composition of output was
shifting, marked by increases from places such as the Middle East, Central Asia, Russia, and West
Africa.
With oil prices at historic highs in early 2007, Midland anticipated continued heavy investment in
acquisitions of promising properties, in development of its proved undeveloped reserves, and in
expanding production. In particular, continued high prices underlay plans to boost investment in
sophisticated extraction methods that extended the lives of older fields and marginal properties.
Capital spending in E&P was expected to exceed $8 billion in 2007 and 2008.
Refining and Marketing
Midland had ownership interests in 40 refineries around the world with distillation capacity of 5.0
million barrels a day. Measured by revenue, Midland’s refining and marketing business was the
company’s largest. Global revenue for 2006 was $203.0 billion—a slight decrease of approximately
1.8% over 2005. The division faced stiff competition, as its products were highly commoditized.
After-tax earnings for refining and marketing totaled only $4.0 billion. The relatively small margin
was consistent with a long-term trend in the industry; margins had declined steadily over the
previous 20 years.
Though most of Midland’s refinery output was gasoline and was sold as fuel for automobiles, the
company also had manufacturing capacity to produce approximately 120,000 barrels of base-stock
lubricants per day. Midland believed its capacity was as technologically advanced as any in the
industry. Advanced technology and vertical integration combined to make Midland a market leader
in this business.
Midland projected capital spending in refining and marketing would remain stable, without
substantial growth in 2007-08. This reflected both the historical trends of low and shrinking margins
and the difficulty of obtaining the myriad approvals necessary to expand or to build and operate a
new refinery. However, most analysts projected a longer-term global shortage of refining capacity
that would eventually spur investment in this segment.
Petrochemicals
Petrochemicals was Midland’s smallest division, but was a substantial business nonetheless.
Midland owned outright or had equity interests in 25 manufacturing facilities and five research
centers in eight countries around the world. The company’s chemical products included
polyethylene, polypropylene, styrene and polystyrene as well as olefins, 1-hexene, aromatics, and
2
BRIEFCASES | HARVARD BUSINESS PUBLISHING
84
Midland Energy Resources, Inc.: Cost of Capital | 4129
fuel and lubricant additives. In 2006, revenue and after-tax earnings were $23.2 billion and $2.1
billion, respectively.
Capital spending in petrochemicals was expected to grow in the near-term as several older
facilities were sold or retired and replaced by newer, more efficient capacity. Much of the new
investment would be undertaken by joint ventures outside the United States in which Midland’s
Petrochemicals Division owned a substantial minority interest.
Midland’s Financial and Investment Policies
Midland’s financial strategy in 2007 was founded on four pillars: (1) to fund significant overseas
growth; (2) to invest in value-creating projects across all divisions; (3) to optimize its capital structure;
and (4) to opportunistically repurchase undervalued shares.
Overseas Growth
The most easily exploited domestic resources had been put into production decades previously.
Consequently, overseas investments were the main engine of growth for most large U.S. producers,
and Midland was no exception. Midland usually invested in foreign projects alongside either a
foreign government or a local business as partner. Often, these investments had specialized financial
and contractual arrangements similar in many respects to project financing. In most cases, Midland
acted as the lead developer of the project, for which it collected a management fee or royalty.
Midland and its foreign partner shared the equity interest, with the foreign partner generally
receiving at least 50% plus a preferred return. Despite the fact that the investments were located
abroad, Midland analyzed and evaluated them in U.S. dollars by converting foreign currency cash
flows to dollars and applying U.S. dollar discount rates. In 2006, Midland had earnings from equity
affiliates of approximately $4.75 billion. The majority of these earnings, 77.7%, came from non-US
investments.
Value-creating Investments
Midland used discounted cash flow methodologies to evaluate most prospective investments.
Midland’s DCF methods typically involved debt-free cash flows and a hurdle rate equal to or derived
from the WACC for the project or division. However, Midland’s interests in some overseas projects
were instead analyzed as streams of future equity cash flows and discounted at a rate based on the
cost of equity.
The performance of a business or division over a given historical period was measured in two
main ways. The first was performance against plan over 1- , 3- , and 5-year periods. The second was
based on “economic value added” (EVA), in which debt-free cash flows1 were reduced by a capital
charge, and expressed in dollars.2 The capital charge was computed as the WACC for the business or
division times the amount of capital it employed during the period.
1 For purposes of EVA calculations, the company defined debt-free cash flows as net operating profit after taxes (NOPAT),
which is EBIT(1-t).
2The basic EVA equation employed by Midland was EVA = NOPAT – (r
HARVARD BUSINESS PUBLISHING | BRIEFCASES
85
wacc
)(Invested Capital).
3
4129 | Midland Energy Resources, Inc.: Cost of Capital
Optimal Capital Structure
Midland optimized its capital structure in large part by prudently exploiting the borrowing
capacity inherent in its energy reserves and in long-lived productive assets such as refining facilities.
Debt levels were regularly reevaluated and long-term targets set accordingly. In particular, changes
in energy price levels were correlated with changes in Midland’s stock price, and necessitated regular
reassessments of optimal borrowing. In 2007 both oil prices and Midland’s stock price were at
historic highs, which—all else equal—increased the company’s borrowing capacity. This in turn
represented an opportunity to shield additional profits from taxes.
Each of Midland’s divisions had its own target debt ratio. Targets were set based on
considerations involving each division’s annual operating cash flow and the collateral value of its
identifiable assets. Targets themselves tended to be “sticky,” but changes in the market value of
specific collateral, such as oil reserves, or the market capitalization of the company as a whole could
drive actual debt ratios away from corresponding targets.
Mortensen’s team did not set targets—they were set in consultations among division and
corporate executives and the board—but Mortensen did estimate a debt rating for each division
based on its target, and a corresponding spread over treasury bonds to estimate divisional and
corporate costs of debt.3 Mortensen’s preliminary estimates for 2007 are shown below in Table 1.
Table 1
Business Segment:
Consolidated
Credit
Rating
A+
Exploration & Production
Refining & Marketing
Petrochemicals
A+
BBB
AA-
Debt/ Spread to
Value Treasury
42.2%
1.62%
46.0%
31.0%
40.0%
1.60%
1.80%
1.35%
Note: Debt/Value is based on market values.
At December 31, 2006, the company’s debt was rated A+ by Standard & Poor’s. Table 2 gives
yields to maturity for U.S. Treasury bonds in January 2007.
Table 2
Maturity:
1-Year
10-Year
30-Year
Rate:
4.54%
4.66%
4.98%
Finally, although prudent use of Midland’s debt capacity was a primary determinant of capital
structure, other considerations played important roles as well. In particular, the strength of
Midland’s consolidated balance sheet and its access to global financial and commodity markets
sometimes presented attractive opportunities to trade securities and commodities. Midland was
conservative compared to some of its large competitors, but it did have a group of traders in-house
3 The spread to Treasury refers to the amount the borrower will have to pay in interest cost above U.S. Treasury securities of a
similar maturity.
4
BRIEFCASES | HARVARD BUSINESS PUBLISHING
86
Midland Energy Resources, Inc.: Cost of Capital | 4129
who actively managed currency, interest rate, and commodity risks within a set of guidelines
approved by the board. The desire to manage certain risks, or to take advantage of private
information or unusual pricing relationships, was an additional reason that the actual capital
structure sometimes departed, temporarily, from planned targets.
Stock Repurchases
In the past, Midland had repurchased its own shares on occasion, and had stated that it would do
so again whenever attractive opportunities arose. Consequently, the company regularly estimated
the intrinsic value of its shares by subtracting the market value of its debt from the fundamental
value of the enterprise and dividing the result by the number of shares outstanding. The
fundamental value of the consolidated enterprise was estimated using DCF analyses and a
comparison of the company’s trading multiples with those of its peers. When the stock price fell
below the stock’s intrinsic value, Midland considered repurchasing its shares. Small numbers of
shares could be purchased on the open market; larger blocks would be bought via self-tenders.
Midland had not repurchased shares in large numbers since 2002 and no large purchases were
anticipated by analysts in the near future, given the company’s high stock price. Nevertheless,
Midland executives pointed out that the mere fact that the stock price had risen did not mean the
shares were not undervalued – intrinsic value had clearly risen as well, and Midland remained
committed to repurchasing shares when they were undervalued. Exhibit 4 shows Midland’s
historical stock prices, dividends per share, and selected financial data for the period 2001-06.
Estimating the Cost of Capital
Mortensen’s primary calculations were based on the formula for WACC shown below.
In this expression, D and E are the market values of the debt and equity respectively, and V is the
firm’s or division’s enterprise value (V = D + E). Similarly, rd and re are the costs of debt and equity,
respectively, and t is the tax rate.
⎛D⎞
⎛E⎞
WACC = rd ⎜ ⎟(1 − t ) + re ⎜ ⎟
⎝V ⎠
⎝V ⎠
Cost of Debt
Mortensen computed the cost of debt for each division by adding a premium, or spread, over U.S.
Treasury securities of a similar maturity. The spread depended on a variety of factors, including the
division’s cash flow from operations, the collateral value of the division’s assets, and overall credit
market conditions. For some of Midland’s operations, long-term expected cash flow and collateral
value were affected by political risk. This risk was most apparent, for example, in the exploration
and production division. A significant fraction of E&P’s productive assets and proven reserves were
located in politically volatile countries in which the risk of nationalization or a forced renegotiation of
production rights was significant. All else equal, such properties supported less borrowing than
might otherwise be expected.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
87
5
4129 | Midland Energy Resources, Inc.: Cost of Capital
Cost of Equity
To estimate the cost of equity, Mortensen used the Capital Asset Pricing Model (CAPM), shown
below, in which rf denotes the risk-free rate of return, β is a measure of systematic risk, and EMRP
denotes the equity market risk premium, that is, the amount by which the return on a broadly
diversified portfolio of risky assets is expected to exceed the risk-free return over a specific holding
period.
re = rf + β(EMRP)
Mortensen was aware that betas were measured, with error, from regressions of individual stock
returns on market returns. She and her team used betas published in commercially available
databases, rather than running their own regressions. Midland’s beta was 1.25, for example.
However, betas for Midland’s divisions were not observable, since the divisions did not have traded
shares of stock. To estimate betas for the divisions, Mortensen relied on published betas for publicly
traded companies she deemed comparable to each division’s business. A selection of these, along
with related financial data, is presented in Exhibit 5.
In 2006 Midland used an equity market risk premium of 5.0%, but higher EMRPs—6.0% to 6.5%—
had been used by Midland at various times in the past. Historical data on stock returns and bond
yields, such as those presented in Exhibit 6, supported the higher estimates of the EMRP. Other data,
such as the survey results shown in Exhibit 6, suggested lower figures. Midland adopted its current
estimate of 5.0% after a review of recent research and in consultation with its professional advisors—
primarily its bankers and auditors—as well as Wall Street analysts covering the industry.
6
BRIEFCASES | HARVARD BUSINESS PUBLISHING
88
Midland Energy Resources, Inc.: Cost of Capital | 4129
Exhibit 1
Midland Income Statements, Years ended December 31 ($ in millions)
Operating Results:
Operating Revenues
Plus: Other Income
Total Revenue & Other Income
Less: Crude Oil & Product Purchases
Less: Production & Manufacturing
Less: Selling, General & Administrative
Less: Depreciation & Depletion
Less: Exploration Expense
Less: Sales-Based Taxes
Less: Other Taxes & Duties
Operating Income
Less: Interest Expense
Less: Other Non-Operating Expenses
Income Before Taxes
Less: Taxes
Net Income
HARVARD BUSINESS PUBLISHING | BRIEFCASES
89
2004
2005
2006
$201,425 $249,246 $248,518
2,817
3,524
1,239
202,664 252,062 252,042
94,672 125,949 124,131
15,793
18,237
20,079
9,417
9,793
9,706
6,642
6,972
7,763
747
656
803
18,539
20,905
20,659
27,849
28,257
26,658
29,005
41,294
42,243
10,568
8,028
11,081
528
543
715
17,910
32,723
30,447
7,414
12,830
11,747
$10,496 $19,893 $18,701
7
4129 | Midland Energy Resources, Inc.: Cost of Capital
Exhibit 2
Midland Balance Sheets, at December 31 ($ in millions)
Assets:
Cash & Cash equivalents
Restricted Cash
Notes Receivable
Inventory
Prepaid Expenses
Total Current Assets
2005
$16,707
3,131
18,689
6,338
2,218
47,083
2006
$19,206
3,131
19,681
7,286
2,226
51,528
Investments & Advances
Net Property, Plant & Equipment
Other Assets
Total Assets
30,140
156,630
10,818
244,671
34,205
167,350
9,294
262,378
Liabilities & Owners' Equity:
Accounts Payable & Accrued Liabilities
Current Portion of Long-Term Debt
Taxes Payable
Total Current Liabilities
24,562
26,534
5,723
56,819
26,576
20,767
5,462
52,805
Long-Term Debt
Post Retirement Benefit Obligations
Accrued Liabilities
Deferred Taxes
Other Long-Term Liabilities
82,414
6,950
4,375
14,197
2,423
81,078
9,473
4,839
14,179
2,725
Total Shareholders' Equity
Total Liabilities & Owners' Equity
8
97,280
77,493
$244,671 $262,378
BRIEFCASES | HARVARD BUSINESS PUBLISHING
90
Midland Energy Resources, Inc.: Cost of Capital | 4129
Exhibit 3
Midland Segment Data ($ in millions)
Exploration & Production:
Operating Revenue
After-Tax Earnings
2004
$15,931
6,781
Capital Expenditures
Depreciation
Total Assets
6,000
7,180
7,940
4,444
4,790
5,525
$76,866 $125,042 $140,100
Refining & Marketing:
Operating Revenue
After-Tax Earnings
2005
$20,870
13,349
2006
$22,357
12,556
2004
2005
2006
$166,280 $206,719 $202,971
2,320
4,382
4,047
Capital Expenditures
Depreciation
Total Assets
1,455
1,620
$60,688
1,550
1,591
$91,629
1,683
1,596
$93,829
Petrochemicals:
Operating Revenue
After-Tax Earnings
2004
$19,215
1,394
2005
$21,657
2,162
2006
$23,189
2,097
Capital Expenditures
Depreciation
Total Assets
305
578
$19,943
330
591
$28,000
436
642
$28,450
HARVARD BUSINESS PUBLISHING | BRIEFCASES
91
9
4129 | Midland Energy Resources, Inc.: Cost of Capital
Exhibit 4
Stock Prices, Dividends and Selected Financial Data
Stock Prices:
Fourth Quarter
Third Quarter
Second Quarter
First Quarter
Dividends Per Share:
Fourth Quarter
Third Quarter
Second Quarter
First Quarter
Annual Dividend
Selected Financial Data:
Net income ($ in millions)
Shares Outstanding
EPS
Payout Ratio
DPS
2001
$27.16
$27.90
$28.33
$24.13
2002
$31.29
$30.41
$27.80
$26.85
2003
$32.59
$29.42
$32.45
$31.57
2004
$34.37
$35.78
$36.98
$31.28
2005
$38.32
$40.29
$37.52
$34.58
2006
$44.11
$39.75
$46.32
$38.81
$0.28
$0.28
$0.28
$0.28
$1.11
$0.29
$0.29
$0.29
$0.29
$1.14
$0.30
$0.29
$0.29
$0.29
$1.17
$0.31
$0.31
$0.31
$0.31
$1.24
$0.34
$0.34
$0.34
$0.34
$1.35
$0.36
$0.36
$0.36
$0.36
$1.46
$15,303
2,049
7.47
14.8%
$1.11
$11,448
2,025
5.65
20.2%
$1.14
$11,848
2,035
5.82
19.9%
$1.16
$10,496
2,055
5.11
24.2%
$1.24
$19,893
2,945
6.75
20.0%
$1.35
$18,701
2,951
6.34
23.0%
$1.46
Note: Results have not been adjusted for a divestiture at the end of 2001 and an acquisition at the beginning of 2005.
10
BRIEFCASES | HARVARD BUSINESS PUBLISHING
92
Midland Energy Resources, Inc.: Cost of Capital | 4129
Exhibit 5
Comparable Company Information ($ in millions)
Exploration & Production:
Jackson Energy, Inc.
Wide Plain Petroleum
Corsicana Energy Corp.
Worthington Petroleum
Average
Refining & Marketing:
Bexar Energy, Inc.
Kirk Corp.
White Point Energy
Petrarch Fuel Services
Arkana Petroleum Corp.
Beaumont Energy, Inc.
Dameron Fuel Services
Average
Midland Energy Resources
Equity
Market Value
$57,931
46,089
42,263
27,591
60,356
15,567
9,204
2,460
18,363
32,662
48,796
$134,114
Net
Debt
$6,480
39,375
6,442
13,098
6,200
3,017
1,925
(296)
5,931
6,743
24,525
$79,508
D/E
11.2%
85.4%
15.2%
47.5%
39.8%
Equity
Beta
0.89
1.21
1.11
1.39
1.15
LTM
LTM
Revenue Earnings
$18,512
$4,981
17,827
8,495
14,505
4,467
12,820
3,506
10.3%
19.4%
20.9%
-12.0%
32.3%
20.6%
50.3%
20.3%
1.70
0.94
1.78
0.24
1.25
1.04
1.42
1.20
160,708
67,751
31,682
18,874
49,117
59,989
58,750
9,560
1,713
1,402
112
3,353
1,467
4,646
59.3%
1.25
$251,003
$18,888
Market values are based on 12/31/06 closing prices. The average stock price for MIDLAND during 2006 was $42.31, and the average
shares outstanding were 2,951 million.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
93
11
4129 | Midland Energy Resources, Inc.: Cost of Capital
Exhibit 6
The Equity Market Risk Premium—Selected Data and Studies
A. Selected historical data on U.S. stock returns minus Treasury bond yields
Period
Average excess return
US Equities – T-Bonds
Standard error
1987-2006
6.4%
3.7%
1967-2006
4.8%
2.6%
1926-2006
7.1%
2.2%
1900-2006
6.8%
1.9%
1872-2006
5.9%
1.6%
1798-2006
5.1%
1.2%
Source: Pratt & Grabowski, Cost of Capital, Applications and Examples, John Wiley & Sons, 2008. Data are
extracted from Exhibit 9.1, p. 95.
B. Selected market risk premium survey results
Researcher
Survey Subjects
Dates
Respondents’ Risk Premia
Welcha
500+ finance & economics
professors
2001
Median: 3.6%
~400 U. S. CFOs
Quarterly
2000-2006
Range: 2.5% - 4.7%
2006
Range: 2% - 4%
Graham & Harveyb
c
Greenwich Associates
US pension fund managers
Interquartile range: 2.6%-5.6%
Most recent survey (4Q2006): 3.3%
Notes:
a. Ivo Welch, “The Equity Premium Consensus Forecast Revisited,” Cowles Foundation Discussion Paper No. 1325, September
2001.
b. John Graham & Campbell Harvey, “The Equity Risk Premium in 2006: Evidence from the Global CFO Outlook Survey,”
downloaded at http://www.duke.edu/~charvey.
c. Greenwich Associates, “Market Trends, Actuarial Assumptions, Funding, and Solvency Ratios,” Fall 2006.
12
BRIEFCASES | HARVARD BUSINESS PUBLISHING
94
IMD399
v. 20.07.2006
INTERNATIONAL
SAURER: THE CHINA CHALLENGE (A)
Professor
Adrian
Ryans
prepared this case as a basis
for class discussion rather than
to illustrate either effective or
ineffective handling of a
business
situation.
Cost,
market share and margin data
have been changed to protect
confidentiality.
KREFELD, GERMANY: In December 2003 the management team
at Saurer Twisting Systems was facing a series of difficult
choices with respect to its machine business. The market for
twisting machines had migrated rapidly to Asia, particularly
China. The global market for these machines was in recession
and Volkmann, one of the two brands in Saurer Twisting
Systems, was seeing a significant decline in sales of machines
for making staple yarn for apparel.
Volkmann was the global leader in the high end of the market,
but saw few opportunities to grow its business in that segment
by taking market share away from its Japanese and European
competitors. The lower end segment of the market had grown
rapidly in recent years. In the critically important Chinese
market, Volkmann’s lowest cost twisting machine, which was
made in China, was about twice the price of the machines
produced by its local Chinese competitors. In addition, Rifa, its
major Chinese competitor, was investing in research and
development and was rapidly upgrading its machines.
Dr Dirk Burger, the CEO, and Wolfgang Leupers, the executive
vice president, of Saurer Twisting, along with their management
team and managers from Saurer China, were considering the
introduction of a lower cost machine targeted at Chinese and
Asian customers who could not afford their high-end solution.
Margins on the new product were likely to be significantly lower
than those on their existing products, even if they managed to
achieve some very aggressive cost targets. In addition, the new
machine could significantly cannibalize their existing high-end
machine. If they did decide to introduce the new product, they
would have to make some very difficult decisions about
positioning, pricing, naming the product, and sales strategy. It
was also not clear how the Chinese competitors would react to
the launch of the new product.
Copyright © 2005 by IMD - International Institute for Management
Development, Lausanne, Switzerland. Not to be used or reproduced without
written permission directly from IMD.
95
-2-
IMD-5-0688
INTERNATIONAL
Saurer
Franz Saurer founded Saurer in 1853. The business began as a small foundry and
engineering workshop. About 15 years later Saurer began manufacturing
embroidery machines in Arbon in eastern Switzerland. Over the next one hundred
years it diversified into a variety of industries. In the late 1980s, after significant
financial difficulties and restructuring, it refocused, largely on the textile industry.
Over the next decade Saurer made a series of acquisitions, many in textile
machinery. Among the companies acquired were Hamel, Volkmann, Allma,
Schlafhorst, Melco and Zinser. In 1999 Saurer purchased Barmag and Neumag,
leading global players in machines for producing and finishing chemical fibers (as
opposed to natural fibers, like cotton or wool). Most of the acquired companies
were based in Germany.
In 2003 Saurer was expected to have revenues of about €1.7 billion and net profit
of over €45 million (refer to Exhibit 1 for some key financial data). About 73% of
Saurer’s revenues were generated by Saurer Textile Solutions (STS).
Saurer Textile Solutions
STS was comprised of Saurer’s nine textile strategic business units. In 2002 the
business units were placed under one management structure reporting to Henry
Fischer, the CEO of Saurer. The business units shared centralized resources for
marketing and sales, technology, and administration and human resources. STS’s
mission was “to be the undebated market leader for full service solutions in textile
engineering and set continuously new benchmarks for efficient production.” In
line with this mission statement, STS was putting increasing emphasis on
providing total solutions for customers in selected textile markets (refer to
Exhibit 2 for ad). A Saurer total solutions package included the textile
manufacturing equipment, project finance, turnkey plant and process support
(including plant design, process integration, plant/machine start-up, and
production support), local service, Internet support and marketing support.
In early 2002 STS began the Tempus program--which aimed to radically reengineer business processes and the corporate culture--to enhance customer
satisfaction and reduce fixed costs by at least €50 million by 2005. During 2003
STS completed a major program to outsource parts manufacturing. Some
production was relocated to the Czech Republic and China. These moves gave it a
much greater ability to weather the highly volatile cycles in the textile machinery
industry and to adjust more rapidly to any changes in market demand.
STS had a number of important strengths. Many of its business units had industryleading products, which incorporated state-of-the-art engineering. In the early
years many of the STS predecessor companies had developed a deep understanding
of textile production from their very close contact and working relationships with
their European customers, who were often the leaders in their industries. Saurer’s
stable of brands, such as Barmag, Volkmann and Schlafhorst, was very strong and
highly regarded by many customers all over the world. Saurer also had a welldeveloped local sales and service presence around the world. The Tempus program
had helped Saurer develop quite a flexible supply chain, so that STS had the best
capital turns in the textile machinery industry. Fischer was still concerned that the
company was not as customer focused and responsive to customer needs as it
96
-3-
IMD-5-0688
INTERNATIONAL
might be, and he believed there were still opportunities to drive more costs out of
the business and to improve the company’s responsiveness.
Textiles and the Textile Machinery Industry
As shown in Exhibit 3, the textile industry can be broadly thought of as having
four major value-added steps: fibers, yarns, fabrics, and apparel and made-up
textile articles. Fibers come from two sources: natural and manmade. The yarns
made from fibers are then made into fabrics, which are converted or incorporated
into a variety of products, such as apparel, home textiles, vehicle tires, carpets,
disposable clothing, baby diapers and luggage.
Worldwide fiber consumption was growing by 2% to 3% per year, with manmade
fibers, such as polyester and acrylic, growing at about 5% per year and natural
fibers, such as wool, cotton and silk, showing little or no growth. The production
of textile raw materials was increasingly concentrated in Asia. Over 50% of the
world’s cotton was produced in Asia, with almost 25% in China alone. The only
major producers outside of Asia were the United States and Brazil. In manmade
fibers the situation was similar. In the case of polyester fiber, 83% of global
manufacturing was done in Asia, with almost 50% of global manufacturing in
China alone. Asian countries accounted for 85% of global exports of textiles and
apparel in 2001. China and India also had huge domestic markets. China was
expected to account for about 25% of the global textile industry in 2004. Some
industry observers believed that this might rise to 50% by 2010 as a result of the
elimination of quotas on China’s exports to North America and the European
Union in January 2005.
STS was the largest manufacturer of textile machinery in the world with a major
presence in many steps of the textile value-added chain (refer to Exhibit 4). In
most of the markets in which it competed, the Saurer business units had the
leading market share (refer to Exhibit 5).
Strategic Importance of China
The globalization of the textile industry accelerated in the early 1990s. Decisions
about where to locate new textile production facilities depended not only on the
raw materials available in a geographic area, but also on the growth potential of
the area and labor costs. Turkey, India and particularly China became growth
markets for textile machinery. These new customers had different requirements
than traditional Western customers. The latest technology and automation were no
longer key choice factors. More emphasis was placed on simple and productionproven equipment and low prices.
Given that China was a major source of textile raw materials, had a large domestic
market and a large and growing position in textile exports, success in China was
crucial for Saurer. Some steps in the textile manufacturing process, such as
weaving and sewing, were still very labor intensive. Labor costs in the textile
industry in China were between $0.40 and $0.70 per hour, versus about $15 per
hour in the United States and over $20 per hour in Germany. Also the Chinese
infrastructure, labor markets and productivity were better than those in other
developing countries. It was theoretically cheaper to manufacture yarn, a highly
97
-4-
IMD-5-0688
INTERNATIONAL
automated part of the textile value chain, in the United States due to lower energy
costs and cost of capital. However, since there were very few customers for yarn
left in the United States, yarn manufacturing had also moved to Asia, and
particularly to China.
Historically, the textile machinery business was about one-third in Europe, onethird in North and South America and one-third in Asia. However, by 2003, more
than 70% of textile manufacturing investments were being made in Asia, with
about half of those in China. As a result of these developments, Saurer was
increasingly dependent on Asian markets, especially China, for its sales (refer to
Exhibit 6). This trend was expected to continue.
Saurer in China
Some of Saurer’s business units (or predecessor companies) were early
participants in the Chinese market. Barmag had established a wholly owned sales
subsidiary in Hong Kong in 1973, and for a period of time was the only company
allowed to export machines for spinning manmade fibers to China. Service was
provided out of Hong Kong. In 1984 Barmag began technical cooperation
agreements with Shanghai Enfangi Company in Shanghai and with Wuxi Hong
Yuan in Wuxi. During this period it had representatives in China, but all contracts
had to be signed in Hong Kong. It also began to add service engineers in China,
opening a service center in Beijing in 1985. In 1992 it established a joint venture
(JV) in Beijing (initially 25% Barmag-owned, but later this was increased to
60%). In 1995 and 1996 Barmag established JVs with its Chinese partners in
Shanghai (51% Barmag-owned) and Wuxi (53% Barmag-owned). These JVs were
successful in further developing the Chinese market and were profitable.
However, they were not without problems.
JV partners set up parallel companies to develop their own products based on the
knowledge they had gained from Barmag, and they even hired key personnel from
the JV for the wholly owned company. Some of the parts supplied by the JV
partners were of inferior quality, which impacted the image of the products. The
partners often had a conflict of interest in making the best decisions for the joint
companies. While Barmag nominally had the controlling interest in the JVs, the
Chinese partners had the relationships that were so essential to business success in
China in the 1990s. By 2000 it was clear that this approach, although successful in
many respects, was not allowing Barmag to fully capitalize on the Chinese market
opportunity.
When China made a decision to allow wholly owned foreign subsidiaries in the
late 1990s, Barmag established Barmag Textile Machinery (Suzhou) in 2001. It
then proceeded to liquidate its Shanghai JV and took over the one in Wuxi.
Taking over certain assets of the Wuxi JV was highly contentious, but ultimately
successful. At the same time as two of Barmag’s JVs were being wound up,
Saurer corporately decided to focus on establishing a direct presence for most of
its business units in China. The joint ventures and other forms of cooperation with
Chinese companies had helped create strong new Chinese competitors, which was
viewed by Fischer as a “cost of learning about the Chinese market.”
By late 2003 Saurer had about 1,000 employees in China with significant
operations in Beijing, Suzhou and Wuxi. Over the previous two or three years,
98
-5-
IMD-5-0688
INTERNATIONAL
Saurer had developed an extensive sales and service organization in China. In
2005 Saurer planned to build a major new facility in Suzhou that would allow it to
consolidate in one location much of its procurement activities, machine assembly
and administration for all the business units operating in China.
Saurer Volkmann
History of Volkmann
Volkmann was founded in 1904 in Krefeld, Germany by two Volkmann brothers.
Initially, Volkmann manufactured textile machinery for the local silk and velvet
industry. After World War II, Volkmann began producing twisting machines for
the manufacture of yarns. Twisting is a mechanical finishing process for two or
more spinning threads which are twisted around a common axis. The twisted yarn
is higher quality with higher density and greater strength than single fibers; it has
a noticeable twist structure that improves the appearance of woven and knitted
goods. In 1954 Volkmann developed the breakthrough “Two-for-One” twisting
machine, which resulted in one turn of the spindle inserting two turns in the yarn.
This machine was significantly more productive and cost effective than
conventional twisting machines. This gave the company a technological lead over
its competitors. Largely as a result of this breakthrough, sales grew to over DM 10
million in the early 1960s.
Over the next two decades, Volkmann expanded globally. By the end of the
1980s, Volkmann had evolved into a healthy, mid-sized enterprise, but felt that it
needed to be part of a larger textile machinery company if it was to continue to be
successful. This led to the merger with Saurer in 1990. Volkmann and Allma
became Saurer Twisting Systems, but the two brands were retained. After the
merger Volkmann’s activities were focused on twisting systems for staple fiber
yarn production (mostly cotton, cotton blends, wool and wool blends) and carpet
yarn twisting.
In 1994 Volkmann introduced a new machine generation of Two-for-One twisters
called the CompactTwister, which was designed to meet the requirements of
customers in the new emerging textile markets (refer to Exhibit 7). The machine
dramatically lowered customer costs. The benefits of the CompactTwister were
also valued by customers in many Western markets, particularly Italy, and most of
the initial sales were outside Asia. The machines were produced in Krefeld. The
design of the product was also imitated by several of Volkmann’s competitors.
Project “Dragon”
Although the CompactTwister enjoyed some success in China, it only succeeded
in capturing about 20% of the market (in units) for twisting machines in 1996
(refer to Exhibit 8). In early 1997 planning began for the production of
CompactTwister in Asia (ultimately called Project “Dragon”). Volkmann set up a
plant in Suzhou, a city in Jiangsu Province west of Shanghai. The product was
initially targeted at the Chinese market, but, over time, the plan was to export it to
99
-6-
IMD-5-0688
INTERNATIONAL
other markets in the Far East. The plant was to have a capacity of 48,000 spindles
per year (equivalent to 240 machines).1 The machines would be identical to those
produced in Germany, but fewer options would be offered, thus reducing the
complexity of the manufacturing and sourcing processes. The goal was to produce
machines of the same quality as those made in Krefeld. The initial investment was
about €3 million.2
After detailed analysis of the likely production costs, the management team
estimated that by 2002 total product cost per spindle in China would be 25%
lower than in Germany. The bulk of the parts for the Chinese machine would be
sourced in China, but some key components would continue to come from
Germany. In addition, as Volkmann developed component suppliers in China,
some of them would be used to supply the Krefeld plant, replacing the more costly
parts made in Germany.
Project Dragon was not without risks. Among them were the political risks of
investing in China, cost inflation in China, productivity levels in the Chinese
plant, potential loss of the Volkmann quality image in Europe and North America,
loss of intellectual property to Chinese competitors, and the impact on the cost
structure in the Krefeld plant of the transfer of production volume to China. The
reaction of the German work force to the loss of jobs to China was also an issue.
As a result of Project Dragon, Twisting Systems was the first Saurer business unit
to establish its own manufacturing operation in China. It began producing the
CompactTwister in May 1998. It was priced in renmimbi and sold at a price about
15% below that of a German-built machine. The product was sold on the basis of
superior German engineering, high quality and excellent performance.
Sales grew rapidly from 1998 to 2002, when Volkmann sold over 40,000 spindles
in China, accounting for about 40% of its global CompactTwister sales. Sales
worldwide were down in 2003, but Chinese CompactTwister sales accounted for
almost 50% of global sales. By 2003 it had succeeded in capturing about 15% of
the Chinese market in terms of machine market share, but about 30% of the
Chinese market in value terms.3 The market was much larger than had been
projected in 1997, with about 1,000 machines expected to be sold in 2004 rather
than the 550 anticipated in the 1997 business plan. The price realized and the
degree of the CompactTwister’s success were pleasant surprises for Saurer
management. As Fischer remarked:
We were pleasantly surprised at how well such an expensive machine sold in China, and
we realized that part of its success was due to the fact that it was priced in local currency,
so that customers did not have to acquire foreign currency to make the purchase.
1
Since twisting machines varied in the number of spindles, number of spindles was a standard way
of measuring market sizes and unit market shares.
2
Most figures are stated in euro or renmimbi (RMB) to simplify the presentation. The exchange
rate in late 2003 was about €1 = RMB 10.
3
Volkmann tended to sell machines with more spindles than average. The combination of higher
prices per spindle and more spindles per machine explains the difference between unit and RMB
market shares.
100
-7-
IMD-5-0688
INTERNATIONAL
In some cases customers had to get a license from the government to import a
machine, and this barrier was obviously removed when the machine was made in
China.
Volkmann global sales in 2003 were expected to be about €50 million, with just
over half of that in staple yarn twisting machines and parts sold under the
Volkmann brand. Gross margins on machine sales were typically about 35%.
Market in China
Volkmann’s Customers
One of the challenges that Volkmann management faced was understanding the
needs of the local Chinese customers. Until the late 1990s all textile companies
had been state-owned. Some of these state-owned companies were very large,
vertically integrated textile companies with thousands of employees; many of
them provided social services such as education and health care for the workers
and their families. There were two major segments of state-owned companies:
those that produced for export markets and those that produced for the local
Chinese market. The export-oriented firms were typically able to get licenses from
the ministry responsible for textiles to buy imported textile machines. As stateowned companies began to be restructured in the 1990s, private, more
entrepreneurial, companies were allowed to produce textile products. These
companies tended to be much smaller, with 50--or fewer--to perhaps 500
employees. This led to a dramatic increase in the number of potential customers,
with many of them initially buying one or two machines. Volkmann estimated that
by 2003 there were perhaps 1,000 to 1,500 potential customers for staple fiber
twisting machines in China, of which it had relationships with 250.
The private companies paid much less attention than Western customers to total
cost of ownership; rather, they were very focused on how long it would take a
machine to pay for itself (“When will I get my money back?”). They felt that this
would reduce their financial risk and give them the flexibility to redeploy their
capital to a totally different industry if a better opportunity came up in a couple of
years. In some cases, Chinese companies had to buy Western or Japanese
equipment to meet their customers’ quality requirements. However, as soon as a
local manufacturer could meet these standards and provide a faster payback, the
companies tended to switch to the lower cost local manufacturer. Customers often
ordered new machines only after they had received yarn orders. They then wanted
quick delivery of the machines.
Connections and relationships were still very important in many suppliercustomer interactions in China, as in the rest of Asia. In addition, given the low
labor costs and the ease of replacing workers, most Chinese companies had little
interest in machine ergonomics and automation, areas that were important to
Western customers. However, the old attitudes were starting to change as more
and more Chinese managers began to appreciate the efficiency of the machines
and the quality of the products they produced, which would enable them to
increase revenue and make more money.
The attitude of most Chinese customers to after-sales service and support was also
different. In North America and Europe, customers wanted no unplanned
101
-8-
IMD-5-0688
INTERNATIONAL
downtime. They invested in preventative maintenance and had planned annual
shutdowns for maintenance. Customers sometimes contracted with Saurer to do
this annual maintenance to minimize machine downtime. These service contracts,
consumables and spare parts were a profitable business. In China, there was little
interest in preventative maintenance or annual contracts. Some Saurer executives
believed that this attitude was rooted in the old state-owned enterprise mentality
of basing mill managers’ bonuses more on minimizing outside cash expenditures
on spare parts than on machine uptime and efficiency. This was deeply ingrained
in the mindset of some of the older managers, who had often worked for stateowned companies for most of their careers. The attitude was to repair the machine
when it broke down and only to replace a part when it failed. This resulted in a
need for quick replacement when a machine did come to a halt. Many of the parts
could be fabricated very quickly in local machine shops close to the customers’
factories, although some of the electronic controls and higher technology parts
still had to be bought from the original equipment manufacturer.
The result of these attitudes and behaviors was that Saurer’s after-sales spare parts
and service revenues per machine in China were only about one-third the level
achieved in Western countries. This was very worrying to Fischer and other senior
Saurer managers, since after sales-service and support were not only profitable but
also a powerful way to develop close relationships with customers, identify new
sales opportunities and gain a deep understanding of their evolving needs.
Volkmann’s Competition in China
Volkmann faced two major traditional twisting machine competitors in China,
Muratec from Japan and Savio from Italy. Muratec was a diversified Japanese
company with total revenues of less than €1 billion. It had traditionally been
Volkmann’s strongest competitor in China, with almost twice the market share of
Volkmann in the mid-1990s. Savio’s market share in the mid-1990s was slightly
smaller than Volkmann’s. With the introduction of the Chinese-made
CompactTwister, Volkmann became the market leader in China, although
Muratec remained a significant player.
The domestic Chinese competitors had become an increasingly important factor in
the textile machinery market, and the twisting machine market was no exception
to this trend. They had a dominant position in the lower end of the market and
showed little respect for the intellectual property of Western and Japanese
manufacturers. In the mid-1990s most of the locally made twisting machines in
China were copies of either the Murata or Volkmann machines. After the
introduction and success of the Volkmann CompactTwister, it was the one that
was most imitated. In some cases even the colors of the machines and the
brochures were reproduced (including printing errors).
The copying was increasingly sophisticated. One of Saurer’s business units had
delivered a new, state-of-the-art machine to one of its Chinese customers. A
month later, the Saurer salesperson asked how the machine was performing and
was told that the customer was not using it for production yet, since a team of
engineers from a textile technology institute had taken the machine apart and were
making computer-aided design (CAD) drawings of all the parts, so that they could
be made available to Chinese manufacturers. Fischer had become philosophical
about the inability to protect intellectual property in China, saying: “For the
102
-9-
IMD-5-0688
INTERNATIONAL
moment, it is a cost of doing business in China, just as dealing with the tort
system is a cost of doing business in the US.” Some Chinese competitors were
surprised that Western companies were so upset by the copying of intellectual
property and responded with remarks such as: “You should be proud that we copy
your machines--we only copy the best!” and “Every great painter starts by
copying the great masters before developing a style of his own.”
By 2003, there were three major Chinese competitors in the twisting market, with
another 10 to 15 local companies in the market at any one time. Rifa Textile
Machines, a private company, was the largest of the three. Rifa was founded in
1993 and by 2003 it had five subsidiary companies producing different types of
textile machines. It was part of a large holding company. Rifa had originally copied
Murata twisting machines, but partly in response to the demands of its own
customers, it had steadily upgraded its machines--often using principles from the
CompactTwister--and was now incorporating its own technology. It had recently
invested in a 7,000 square meter research and development center, and was highly
efficient from a production perspective. Rifa had a relatively large sales force with
good coverage all over China. Taitan, the second Chinese competitor, was smaller
than Rifa but followed a similar strategy. Wanli, the third major local competitor,
was viewed as less of a threat by Volkmann management. Volkmann’s Chinese
competitors were very focused on winning contracts, even if margins were very
low. They seemed to be driven primarily by revenue growth rather than profit
growth.
The situation in India was similar to that in China. A strong local competitor,
Veejay Lakshmi, had emerged and was following a strategy very similar to Rifa’s.
It had captured about 70% market share in the Indian market and was exporting its
twisting machines to some other developing countries, particularly those with
expatriate Indian business people in the textile business. Muratec and Savio were
also active in India. In Pakistan, Muratec had the dominant market share, but Rifa
was beginning to gain some. Volkmann had very low market shares in both
countries.
Responding to the Local Chinese Challenge
Although the CompactTwister was the clear market leader in the high end of the
Chinese market, the product did not meet the needs of the lower-end customers in
China and other Asian countries, who bought twisting machines from local
competitors. Volkmann managers took some comfort from the fact that a few
Chinese customers were attempting to compete in higher-end markets where
better-quality and high-performance yarns were key and where the capabilities
and flexibility of the CompactTwister were important. However, at the same time
the quality and capabilities of the Rifa and Taitan machines were improving
rapidly. Rifa was starting to supply twisting machines for some of the coarser
wool and wool-blend yarns, a market which Volkmann and the other high-end
competitors had traditionally owned. While the cotton and cotton blend market
was still showing good growth in Asia, the wool and wool blend market was
stagnant.
After a trip to China in September 2003, Burger sent a memo to members of his
management team which said in part:
103
- 10 -
IMD-5-0688
INTERNATIONAL
Besides the market slowdown in China, competition is getting worse and worse. Local
competitors have reduced their prices by 30% per spindle. Machines from local competitors
have less performance, but they are doing the job for counts lower than Ne 80.4 Our sales
price is twice as much… It will be difficult to hold onto our market share with our existing
pricing strategy… Therefore, it seems absolutely necessary to develop a new product for
cotton that is less than Ne 80.
A team with members from both Germany and China, under the leadership of
Burger and Leupers, began investigating the opportunity. The team had had one
face-to-face meeting in Krefeld to exchange information and do some preliminary
planning. They attempted to quantify the size of the opportunity in the low end,
customer requirements and price levels. They concluded that the CompactTwister,
although well suited to wool and very fine cotton yarns, was over-engineered for
less demanding segments. They estimated that the market size for the next few
years would average about 1,000 machines (with 150 spindles per machine) per
year, with about 300 in the high end and 700 in the low end. The market in the
rest of Asia was expected to be almost as large.
From their research the team put together the preliminary specification for a new
twisting machine that would meet the needs of many lower-end customers making
cotton and cotton blend yarns with medium and fine yarn counts. On some
specifications, such as spindle speed, the proposed machine was higher than Rifa
and the other local Chinese competitors, matched Muratec, but was lower than
CompactTwister. On other specifications it was lower than all competitors, and on
some other features that were standard on competitive machines, it offered the
feature as an option for an additional charge. Overall, the team believed that the
proposed product was superior to all of the products made by its Chinese
competitors. They believed that the key elements of the value proposition for
customers would be about 20% higher productivity, globally competitive yarn
quality, lower energy and spare parts consumption, easy maintenance and high
reliability. However, convincing the customers of the value of these benefits
would take effective selling. The CompactTwister would continue to be the
machine of choice for wool and wool blends, as well as very fine cotton yarn.
The variable manufacturing cost per spindle for the proposed machine was
estimated to be 25 to 30% lower than the CompactTwister in 2005, and that cost
was expected to be reduced by about 3% per year. Other variable costs, including
sales, installation and warranty, were expected to be about 15% of the sales price.
VAT was 17%. In addition, there were fixed production and administration costs
that had to be covered. Normally, these additional costs were about 8% of sales.
Decision Time
As they prepared for the crucial decision meeting on December 8, Burger and
Leupers reviewed the arguments for and against the launch of the new product.
They also looked at implementation issues the Volkmann and Saurer China team
would have to resolve if the product was to be a profitable addition to the
Volkmann product line.
4
The higher the count the finer was the yarn. Ne 80-count yarn was fine, but some Volkmann
twisters could handle extremely fine yarn with counts as high as Ne 200.
104
- 11 -
IMD-5-0688
INTERNATIONAL
They could see several reasons for not moving ahead with the value solution.
Volkmann’s name was synonymous with leading-edge engineering, and it had
always positioned itself as the technology leader in twisting systems. If Volkmann
introduced a value solution, it would almost certainly lead to some cannibalization
of CompactTwister sales. Since Volkmann had historically focused on customers
with high-end needs, its sales force had almost no contact with the target
customers for the new machine, and it was used to selling the CompactTwister
against low-end solutions. In fact, it would be quite difficult to identify the
potential customers. A quick analysis had suggested that about 85% of the
potential demand for the new machine would be in 6 of China’s 22 provinces.
These six provinces had a population of almost 400 million.
The biggest challenge would be making the low-end business a profitable one for
Volkmann. Even at relatively high prices for the new machine and with an
aggressive cost reduction program, the contribution margin was likely to be a
fraction of that for CompactTwister. There was also the issue of how Rifa and the
other Chinese and foreign competitors would react to Volkmann’s entry into this
segment. Would they drop their prices in response? Rifa, in particular, was more
vertically integrated than Volkmann and this might give it a cost advantage. In
some cases, Volkmann used higher-quality components that resulted in higher
costs. Volkmann spindle bearings were designed to last for 100,000 hours, and
some had now been in machines for 15 years with no failures.
If Volkmann decided to introduce the new product, it needed to develop a
comprehensive marketing strategy. Among the major issues that needed to be
addressed was the exact value proposition that would be communicated to the
target customers, the pricing of the new product, the branding and naming of the
product and the sales and marketing communications strategy. The pricing of the
CompactTwister and competitive machines in late 2003 is shown in Exhibit 9.
Some small local players had even lower prices than Taitan. Several names for the
new product were also under consideration, including VolkTwister, CottonTwister
and E-Twister (E for easy or economical).
Volkmann currently had four regional managers, who also acted as sales
representatives, together with three additional sales representatives to cover all of
China. These sales representatives were paid a fixed salary and a commission
based on the sales revenue the individual generated. Typically, about two-thirds of
the representative’s total compensation came from the fixed salary. Saurer’s
Suzhou organization would be responsible for installing and commissioning new
machines and providing spare parts support. An obvious opportunity to launch the
new machine was at the biannual CITME (China International Textile Machinery
Exhibition), which was to be held in October 2004 in Beijing. At the 2002
CITME, the exhibition space occupied 50,000 square meters and attracted 850
exhibitors and 120,000 visitors. If a decision was made quickly and product
development moved very quickly, a prototype could probably be built by mid2004 and the first regular production machines could be assembled and ready
before CITME.
105
106
INTERNATIONAL
47
Net profit (loss)
671
580
142
468
Current assets
Non-current assets
Third-party loans
Shareholders’ equity
Source: Company information
Capital expenditures
51
10.7%
1,251
Total assets
Return on equity (ROE)
2003
Balance Sheet
125
88
EBIT
Cash flow (from op. activities)
166
1,746
Sales
EBITDA
2003
Income Statement
50
7.7%
431
206
630
671
1,301
2003
145
33
69
149
1,697
2002
Exhibit 1
Selected Financial Data (€ million)
- 12 -
64
-12.3%
415
342
692
695
1,386
2001
117
-51
-36
100
1,593
2001
74
9.4%
470
276
688
646
1,334
2000
110
44
77
151
1,426
2000
IMD-5-0688
107
Source: Company information
INTERNATIONAL
Exhibit 2
Saurer Total Solutions Ad
- 13 IMD-5-0688
108
Filament
ƒ $UWLILFLDO
5D\RQ
$FHWDWH
ƒ 6\QWKHWLF
3RO\HVWHU
1\ORQ
$FU\OLF
Chemical industry
(manmade fibers)
ƒ 7RZHOV
ƒ 6KHHWVSLOORZFDVHV
ƒ &XUWDLQVDQGGUDSHV
Home textiles
ƒ 6KLUWVDQGEORXVHV
ƒ 7URXVHUVDQGVKRUWV
ƒ 6NLUWVDQGGUHVVHV
ƒ 8QGHUZHDU
Apparel
Apparel and
Made-up
Textile
Articles
ƒ /XJJDJH
ƒ 7HQWV
ƒ %DJV
Other made-ups
Industrial fabrics Carpets and rugs
Nonwoven
Knit
ƒ &RWWRQ PDQPDGH
ILEHUV
ƒ :RRODQGILQHDQLPDO
KDLU
ƒ 0DQPDGHILEHUV
ƒ 6LON
ƒ 'HQLP
ƒ 3ULQWFORWK
ƒ %URDGFORWK
ƒ 6KHHWLQJ
Spun
Agricultural sector
(natural)
ƒ &RWWRQ
ƒ :RRODQGILQHDQLPDOKDLU
ƒ 6LON
ƒ 5DPLH
Woven
Yarns
Fabrics
Exhibit 3
Textile Industry Value-Added Steps
- 14 -
Fibers
Source: US International Trade Commission
INTERNATIONAL
IMD-5-0688
109
Source: Company information
INTERNATIONAL
6DXUHU
Neumag
Nonwoven
Schlafhorst
Open-end
Spinning
Barmag
Staple Fiber Spinning
Spinning Preparation
Zinser
Ring Spinning
Schlafhorst
Winding
Allma & Volkmann
Twisting
Weaving/Knitting
Finishing/Printing/Dyeing
Melco Saurer
Embroidery
Final Consumer Market
Air-Jet Spinning
Exhibit 4
Saurer’s Participation in the Textiles Value-Added Chain
- 15 -
POY Preparation
Barmag &
Neumag
Filament
Spinning
Barmag
Texturizing
IMD-5-0688
110
Neumag
Saurer
Melco
Allma, Hamel,
Volkmann
30%
40%
40%
35%
38%
60%
25%
Market
Share
400-600
1
1
1
1
1
1,2
1
2,1
Saurer
Pos.
10-50% 3
300-500
50%
(excl. multihead)
500-750
650-1000
200-400
350-500
500-800
400-800
500-1000
Market Volume
in CHF m
Note: 1 Swiss Franc (CHF) was approximately € 0.63
Airlaid/Spunbond
NONWOVENS
Embroidery
Systems
EMBROIDERY
Twisting
Systems
TWISTING
Filament Spinning
Synth. Staple fiber
Texturing
Barmag
Neumag
Barmag
Schlafhorst
Schlafhorst
Winding
Rotor Spinning
Brands
Zinser
FILAMENT
- 16 -
Reifenhäuser, Fleissner
Lässer,
Tajima, Barudan
Murata, Rieter, Leewa
Savio
TMT, Rieter
Lurgi Zimmer, HILLS, Fare
TMT, Rieter
Rieter, Toyoda, Marzoli,
LMW
Murata, Savio
Rieter, Savio
Main Competitors
Exhibit 5
Saurer’s Position in the Textile Value-Added Steps
(2003)
Ringspinning
SPINNING
Source: Company information
INTERNATIONAL
IMD-5-0688
111
Percentage of total sales
0%
10%
20%
30%
40%
50%
60%
70%
80%
Source: Company information
INTERNATIONAL
2001
2000
Asia
2002
2001
2000
Region
America
2003
1995
2003
2002
Exhibit 6
Percentage of Saurer Textile Solution’s Sales by Region
- 17 -
2001
2000
Europe
2003
2002
1995
1995
IMD-5-0688
112
Source: Company information
INTERNATIONAL
Exhibit 7
CompactTwister
- 18 IMD-5-0688
113
$FWXDO
3URMHFWHG
$VVXPHVWKDWVWDSOHILEHU PDUNHWJURZVDWSHU\HDU
Saurer Germany
Saurer China
Muratec
Savio
Other
Second hand
Exhibit 8
Actual and Projected Sales of Staple Fiber Twisting Machines in China
- 19 -
1995 1996 1997 1998 1999 2000 2001 2002
0
100
200
300
400
500
600
Source: Company information
INTERNATIONAL
Machines per year
IMD-5-0688
114
:DQOL
5LID
7DLWDQ
9RONPDQQ
&RPSDFW7ZLVWHU
0XUDWHF
Price per spindle
(in RMB)
Source: Company information
Exhibit 9
Approximate Pricing in China in Late 2003
- 20 -
Manufacturer
1RWH$OOSULFHVLQFOXGH9$7
INTERNATIONAL
IMD-5-0688
www.hbrreprints.org
Multinationals and local
firms for the first time are
squaring off in China’s rapidly
growing middle market—a
critical staging ground for
global expansion and the
segment from which worldbeating companies will
emerge.
The Battle for China’s
Good-Enough Market
by Orit Gadiesh, Philip Leung, and Till Vestring
Included with this full-text Harvard Business Review article:
1 Article Summary
The Idea in Brief—the core idea
The Idea in Practice—putting the idea to work
2 The Battle for China’s Good-Enough Market
12 Further Reading
A list of related materials, with annotations to guide further
exploration of the article’s ideas and applications
Reprint R0709E
115
The Battle for China’s Good-Enough Market
The Idea in Brief
The Idea in Practice
Between now and 2030, China will account
for one-third of the world’s GDP growth.
Yet many multinationals are losing share in
this critical market. That’s because local
businesses are targeting China’s ballooning
cohort of midlevel consumers with reliable,
low-cost products that are displacing
multinationals’ premium offerings. And the
regional upstarts making these “good
enough” products plan to use the same
strategy to challenge incumbents in other
emerging markets.
Gadiesh, Leung, and Vestring offer these
guidelines for entering China’s goodenough space:
COPYRIGHT © 2007 HARVARD BUSINESS SCHOOL PUBLISHING CORPORATION. ALL RIGHTS RESERVED.
To defend your China position and prevent
local competitors from becoming global
threats, say Gadiesh, Leung, and Vestring,
consider entering China’s good-enough
space. For instance, attack the competition
from above by lowering your costs and
distributing simplified, reasonable-quality
offerings. If you can’t reduce your costs
quickly, use acquisitions to gain a toehold
in this space.
By managing the risks and opportunities
inherent in China’s middle market, you’ll
claim your share of this pivotal market. And
you’ll strengthen your competitive position
elsewhere around the globe.
ATTACK FROM ABOVE
Moving to the good-enough segment in
China is risky if you’re already thriving in the
premium space. For instance, your new offerings could cannibalize your high-end
products. To mitigate the risks:
• Analyze the differences between China’s
premium and good-enough segments. You
may discover strong geographic distinctions you can capitalize on.
Example:
GE Healthcare expanded sales of its MRI
equipment in China by creating a line of
simplified machines targeted at hospitals in
China’s remote and financially constrained
second- and third-tier cities.
• Determine which capabilities and resources
you’ll need to seize opportunities in the
good-enough space.
Example:
GE Healthcare assigned a special team to
observe target hospitals’ operations. Members also talked with administrators and
physicians to determine the kinds of medical equipment they wanted, features they
needed, possible price points, and required distribution and services. GE then
reconfigured its existing networks of sales,
distribution, and services to serve this
new market.
• Stake claims in the good-enough space to
box out emerging local players and global
competitors that might be eyeing the same
target market. By entering this space ahead
of the pack, GE Healthcare defended its position against local upstarts, capturing 52%
of the $238 million market in 2004.
USE ACQUISITIONS
If you can’t alter your cost structure or business processes quickly enough to compete
with local players, consider mergers and
acquisitions.
Example:
Anheuser-Busch owned 27% of Tsingtao
Brewery, one of China’s largest brewers. It
outbid competitor SABMiller to acquire
Harbin, the fourth-largest brewer in China.
The acquisition enabled Anheuser-Busch to
reach the masses while preventing Harbin
from swimming upstream.
Note, though, that non-Chinese acquirers are
facing tougher M&A approval processes. To increase your chances of gaining regulatory and
political approval:
• Draft a compelling business case for the
acquisition, citing benefits for local companies and authorities.
• Be willing to adjust the structure, terms, and
conditions of the deal.
• Engage in heavy-duty relationship building,
to woo critical players.
Also, to ensure that each acquisition delivers
the maximum possible value:
• Select a target company that offers cost
and distribution synergies with your firm
and whose products won’t cannibalize your
premium brands.
• Overinvest in the due diligence process.
• Take a systematic approach to postmerger
integration.
page 1
116
Multinationals and local firms for the first time are squaring off in
China’s rapidly growing middle market—a critical staging ground for
global expansion and the segment from which world-beating
companies will emerge.
The Battle for China’s
Good-Enough Market
COPYRIGHT © 2007 HARVARD BUSINESS SCHOOL PUBLISHING CORPORATION. ALL RIGHTS RESERVED.
by Orit Gadiesh, Philip Leung, and Till Vestring
Caterpillar, the world leader in construction
equipment, is having trouble making deeper
tracks in China. The U.S.-based manufacturer
of tractors, backhoes, road graders, and other
devices began selling equipment in China in
1975, a year before the death of Chairman
Mao. As the Chinese government invested
massively in infrastructure, Caterpillar helped
pave the way, literally, for economic growth
and modernization in the world’s fastestgrowing market for construction equipment.
Like many foreign players in any number of
industries, Caterpillar got its start in China by
selling goods to the Chinese government—
the only possible customer before the era of
economic reform—and then began selling
high-quality equipment to the private sector
as a premium segment of the market
emerged. But it never broadened its focus to
include other segments, and by the early
2000s, Komatsu, Hitachi, Daewoo, and other
competitors from Japan and Korea were in
the middle market with tools and equipment
that cost less but were still reliable. Mean-
harvard business review • september 2007
while, a tranche of local manufacturers that
had previously been focused only on the low
end of the market were burrowing up to battle the established players, designing and releasing their own products targeted squarely
at middle-market consumers.
As the experiences of Caterpillar and other
multinationals suggest, a critical new battleground is emerging for companies seeking
to establish, sustain, or expand their presence
in China: It’s the “good-enough” market segment, home of reliable-enough products
at low-enough prices to attract the cream
of China’s fast-growing cohort of midlevel
consumers.
Harvard professor Clay Christensen, author
of The Innovator’s Dilemma, has used the
phrase “good enough” to suggest that start-up
companies developing and releasing new
products and services don’t necessarily need
to aim for perfection to make inroads against
established players. The phrase can be similarly applied to middle-market players in
China that have been able to steal a march on
page 2
117
The Battle for China’s Good-Enough Market
incumbents by developing and releasing
good-enough products that are displacing
premium ones.
These forward-thinking companies (multinational and domestic firms alike) are doing
more than just seizing share of wallet and
share of mind in China’s rapidly expanding
middle market—in and of itself a major
achievement. They are conditioning themselves for worldwide competition tomorrow:
They’re building the scale, expertise, and
business capabilities they’ll need to export
their China offerings to other large emerging
markets (India and Brazil, for instance) and,
ultimately, to the developed markets. Given
China’s share of global market growth (Goldman Sachs estimates that China will account
for 36% of the world’s incremental GDP
between 2000 and 2030) and the country’s
role in preparing companies to pursue
opportunities in other developing regions,
it’s becoming clear that businesses wanting
to succeed globally will need to win in
China first.
In the following pages, we’ll explore the
importance of China as a lead market. We’ll
describe the surge of activity in China’s
middle market; when (and whether) multinationals and Chinese companies should
enter this vibrant arena for growth; and,
most important, how they can compete
effectively in the good-enough segment. As
Caterpillar and other foreign players have
learned, achieving leadership in China’s
middle market isn’t easy.
An Evolving Opportunity
Orit Gadiesh (orit.gadiesh@bain.com)
is the chairman of Bain & Company in
Boston. Philip Leung (philip.leung@
bain.com), a Bain partner in Shanghai,
leads the firm’s Greater China health care
practice. Till Vestring (till.vestring@
bain.com), a Bain partner based in
Singapore, leads the firm’s Asia-Pacific
industrial practice.
Historically, there has been a simple structure to China’s markets: at the top, a small
premium segment served by foreign companies realizing solid margins and rapid growth;
at the bottom, a large low-end segment
served by local companies offering lowquality, undifferentiated products (typically
40% to 90% cheaper than premium ones)
that often lose money—when producers do
their accounting right. Between the two is
the rapidly expanding good-enough segment.
(For an example of how one market sector
breaks out, see the exhibit “The Structure of
China’s Market for Televisions.”)
The good-enough space in China is growing for many reasons, not the least of which
are recent shifts in consumer buying patterns
harvard business review • september 2007
and preferences. These shifts are coming
from two directions: Consumers with rising
incomes are trading up from the low-end
products they previously purchased. At the
same time, higher-income consumers are
moving away from pricey foreign brands and
accepting less expensive, locally produced
alternatives of reasonable quality. The same
holds true on the B2B front.
Consequently, China’s middle market is
growing faster than both the premium and
low-end segments. In some categories, the
good-enough space already accounts for
nearly half of all revenues. Eight out of every
ten washing machines and televisions now
sold in China are good-enough brands. It
should come as no surprise, then, that
China—and, in particular, its opportunityrich middle market—is increasingly capturing multinational executives’ resources and
attention. As Mark Bernhard, chief financial
officer of General Motors’ Shanghai-based
GM China Group, recently told the Detroit
News: “For GM to remain a global industry
leader, we must also be a leader in China.”
The automaker’s strategy in China embodies
that belief. GM had traditionally been an
underperformer in the market for small
cars. However, its acquisition of Korea’s ailing
Daewoo Motor in 2002 enabled it to compete
and ultimately take a leadership position in
China. The deal allowed GM to develop new
models for half what it would cost the company to develop them in the West. Daewoodesigned cars now make up more than 50%
of GM’s sales in China, currently its second
biggest market. What’s more, GM is using
these vehicles to compete against Asian
automakers and sell small cars in more than
150 markets around the world, from India to
the United States.
Colgate-Palmolive made similar moves in
China. It entered into a joint venture in the
early 1990s with one of China’s largest toothpaste producers, and it acquired China’s market leader for toothbrushes a decade later,
allowing it to scale up and then leverage its
production processes to compete in other
parts of the world. As a result, Colgate more
than doubled its oral hygiene revenues in
China between 1998 and 2005, and it now
exports its China products to 70 countries.
Local Chinese competitors pose the biggest
challenge to multinationals seeking to capi-
page 3
118
The Battle for China’s Good-Enough Market
talize on their business ventures in China
and beyond. In the auto industry, for instance, domestic carmakers like Geely and
Chery have eaten away at Western companies’ market share in China by introducing
good-enough cars for local consumption.
Several of these automakers have started
exhibiting vehicles at car shows in the
United States and Europe, buying available
Western brands, and exporting vehicles to
other emerging markets. True, these players
face enormous challenges in meeting safety
and emissions standards and in building up
the required distribution networks to compete in Europe and North America. But no
Western company should underestimate the
determination of Chinese firms to figure out
how to meet international quality standards
and make their global mark.
European and North American companies
producing major appliances, microwaves, and
televisions know this all too well. They abdicated China to low-cost local competitors in
the 1990s and now find themselves struggling
to compete globally against those same Chinese companies. Haier, which started making
refrigerators in 1984, went on to become one
of China’s best-known brands and then used
its hard-won scale advantages and manufacturing skills to crack, and then dominate,
foreign markets. Today, it is one of the largest
refrigerator companies in the world, controlling 8.3% of the highly fragmented global
market. The company sells products in more
than 100 markets, including the United States,
Africa, and Pakistan.
Obviously, the stakes in China have
changed. Local companies are using booming domestic markets to hone their strategies
at home before taking on the world. Multinationals, therefore, need to defend their
positions in China not only to profit from the
economic growth in that country but also to
prevent local competitors from becoming
global threats. The good-enough space is
where multinationals and Chinese firms are
going head-to-head—and it’s the market
segment from which the world’s leading
companies will emerge.
Making an Entrance
It’s one thing to recognize the importance of
China’s middle market; it’s another thing entirely to turn that awareness into action. The
first step in winning the battle for China’s
good-enough market is determining when—
or when not—to enter the fray. That will depend on the attractiveness of the premium
segment: Is it still growing? Are companies
still achieving high returns or are returns
eroding? Another consideration is your company’s market position: Are you a leader or a
niche player? (See the exhibit “Should You
Enter the Middle Market?”)
Foreign companies grappling with the goodenough decision in China will need to consider
these factors and perform thorough market
and competitor analyses, along with careful
The Structure of China’s Market for Televisions
Premium (Narrow)
Good-Enough (Rapidly Expanding)
Low-End (Evolving Base)
Definition: High-end products
purchased by discerning
customers with significant
purchasing power.
Definition: Products of good quality,
produced by local companies for a
rapidly expanding group of value-seeking
consumers with midlevel incomes.
Leading Vendors:
Panasonic, Philips, Sony
Leading Vendors:
Hisense, Skyworth, TCL
Definition: Products of lower
quality, meeting basic needs,
produced by local firms for a
large group of consumers with
low incomes.
Product Features: LCD and
plasma screens, many stateof-the art user features, priced
according to their status as
international brands.
Product Features: LCD, plasma, and
large cathode-ray tube screens, with
limited user features, priced to undercut
foreign brands.
Product Features: Cathode-ray
tube screens with basic standard user features and low-cost
components, priced to sell.
Share of Market in 2005: 13%
Share of Market in 2005: 62%
Share of Market in 2005: 25%
harvard business review • september 2007
Leading Vendor: Konka
page 4
119
The Battle for China’s Good-Enough Market
customer segmentation and needs analyses—
classic strategy tools, of course, but applied in
the context of a rapidly changing economy
that may lack historical data on market share,
prices, and the like. Senior managers will need
to establish the factors that are key to success
in everything from branding to pricing to distribution. This knowledge will inform important decisions about whether companies
should expand organically into the middle
market, acquire an existing player in that
space, or find a good-enough partner.
Generally speaking, competing in the goodenough space is neither necessary nor wise for
multinationals operating in stable premium
segments. These companies should instead
focus on lowering their costs and innovating to
maintain their premium or niche positions and
to sustain their margins. We studied one large
manufacturer of automation equipment, for
example, that wisely decided to stand pat in
the premium segment. Market research suggested that its customers were still willing to
pay more for reliability, even with a variety of
lower-cost choices out there. The company
continued to invest in R&D, hoping to further
differentiate its products from those of local
players; it expanded its distribution and service
networks to improve its responsiveness to customers; and it cut costs by taking advantage of
local production resources.
Few multinationals find themselves in such
a fortunate position, however. If growth in the
premium segment is slowing and returns are
eroding, multinational corporations will need
to enter the good-enough space. Even those
companies that because of their strong competitive position initially abstain from entering
the middle market should revisit their decision
frequently to guard against emerging competitive threats. For their part, Chinese companies
will need to move upmarket as the lower-end
segment becomes increasingly competitive.
Our research and experience indicate that
companies contemplating a move into the
good-enough space go about it in one of three
basic ways: Leading multinationals in the premium segment attack from above. The goal for
Should You Enter the Middle Market?
Multinationals deciding whether to move
into China’s middle market need to first
consider the attractiveness of the premium
segment and their current market position.
If conditions warrant, they can attack aggressively from above. Chinese firms can
burrow up from below. Both can acquire
their way into the good-enough space.
STATE OF THE PREMIUM MARKET SEGMENT
STRONG
WEAK OR ERODING
Maintain strong premium
status
Attack from above or
buy your way in
Hold off on entering the goodenough segment of the market—
Premium players employ an
for now. Drop prices as required
to remain competitive; lower costs
and innovate to defend premium
enter the middle market. That is,
they enter the good-enough segment in order to defend against the
status and sustain margin.
rise of local competitors and the
erosion of the premium segment.
Regularly reevaluate the decision
not to enter.
WEAK OR ERODING
COMPANIES’ COMPETITIVE POSITION
STRONG
offensive-defense approach to
Innovate to maintain
current premium status
Burrow up from below or
buy your way in
Hold off on entering the goodenough segment of the market—
Value players enter the goodenough segment using a break-
for now. Increase innovation efforts
to capture a niche position in the
premium segment.
through approach—with a merger,
for instance, or by developing
China-specific products or busi-
Regularly reevaluate the decision
not to enter.
harvard business review • september 2007
ness models—to steal share from
incumbents and attain market
leadership.
page 5
120
The Battle for China’s Good-Enough Market
these organizations is to lower their manufacturing costs, introduce simplified products
or services, and broaden their distribution
networks while maintaining reasonable quality. Meanwhile, Chinese challengers in the
low-end segment tend to burrow up from
below. These companies aim to take the legs
out from under established players by providing new offerings that ratchet up quality but
cost consumers much less than the premium
products do. And, finally, multinationals that
can’t reduce their costs fast enough, and
domestic players looking for more skills,
technology, and talent, buy their way in.
Each of these moves comes with its own set
of traps. The challenge, then, for companies
eyeing the middle market is to understand why
those that went before them failed in this
space—and how to sidestep the pitfalls they
encountered. Let’s take a closer look at these
three approaches.
Attacking from Above
Whether they’re selling toothpaste or power
transmission equipment, multinational companies dominate China’s small but highmargin premium segment—the only one in
which foreign players have traditionally been
able to compete successfully. So a move toward
the middle certainly holds a fair amount of
risk for those already thriving in the premium
space. A chief concern is cannibalization. After
all, selling to consumers in less-than-premium
segments could negatively affect sales of highend products. These companies also run the
risk of fueling gray markets for their wares. If,
say, a business sells a T-shirt for $10 in China
but $20 in the United States, there’s a good
chance an enterprising distributor will find a
way to buy that T-shirt in China and export it
to the United States for sale there.
Multinational managers, therefore, need to
conduct careful market analyses to understand
the differences between China’s premium and
good-enough segments. There may be, for
instance, strong geographic distinctions a company can capitalize on. Consider the strategy
GE Healthcare employed to expand sales of
its MRI equipment in China. The company
created a line of simplified machines targeted
at hospitals in China’s remote and financially
constrained second- and third-tier cities—
places like Hefei and Lanzhou, where other
multinationals rarely ventured. That good-
harvard business review • september 2007
enough territory had all the right conditions: It
was a fast-growing market whose customers’
purchasing criteria weren’t likely to change
soon. GE’s cost structure allowed it to compete
with other middle-market players in the industry. And there was little risk that the company
would cannibalize its premium line of diagnostic machines; large city hospitals were not keen
on downgrading their MRI equipment.
Markets are dynamic, and there’s no place
on the planet where they are shifting as
quickly or as dramatically as in China. So
multinational executives also need to think
about the degree to which the premium and
good-enough segments will converge over time.
Managers can use traditional forecasting
methods (scenario planning, war gaming,
consultations with leading-edge customers,
and so on) to pick up on emerging threats
and impending opportunities. Which brings
us back to GE Healthcare’s MRI expansion
plans: The company’s long-held commitment
to health care development in China meshed
perfectly with Chinese leaders’ publicly
stated desire to improve health services in
less-privileged areas of the country. Given
the government’s aims, GE Healthcare understood there would eventually be some
overlap of the premium, middle, and lowend markets—and profitable opportunities
in the good-enough space.
After weighing the risks that cannibalization and dynamic markets pose to their
company’s premium positioning, managers
in multinationals need to consider their
possible opportunities in the good-enough
space: Can they take advantage of their
lower purchasing costs, greater manufacturing scale, and distribution synergies? Then
they have to determine which capabilities
they may need to develop: How adept is
their organization at designing products,
services, brands, and sales approaches that
will attract customers in the middle market
without diminishing their company’s position
in the premium space? They may need to
convene teams dedicated solely to studying
the opportunities and resources required in
the good-enough segment, as GE Healthcare
did. (See the sidebar, “Penetrating the GoodEnough Market, One County Hospital at a
Time.”) They may also want to recruit local
management talent—individuals with experience competing in the middle space—
page 6
121
The Battle for China’s Good-Enough Market
or purchase local companies to gain new
technologies or expertise.
Those multinationals that decide to enter the
middle market tend to employ an “offensivedefense” strategy—aggressively staking claims
in the good-enough space to box out emerging local players and established global
competitors seeking to gain their own scale
advantages. By entering the good-enough
space ahead of the pack, for instance, GE
Healthcare was able to defend its position
against local upstarts, including Mindray,
Wandong, and Anke. The company is still
trying to develop the optimal product portfolio
and is addressing such issues as how best to
service the equipment. Even so, GE captured
52% of the $238 million market in 2004,
generating roughly $120 million in sales.
Having honed its approach to the goodenough space, GE is replicating the strategy in
new markets in several developing countries,
including India.
Multinationals are bound to find it tough to
jump in from above. Apart from the risks of
cannibalization and all the challenges always
associated with going down-market, companies will need to adapt fast, as customers’
preferences change and competitors react.
And they will probably need to tear apart the
cost structure of their good-enough competitors to understand how those firms make
money while charging such low prices. Just
switching to local sourcing, for instance, may
not be sufficient for large multinationals to
match the lower production costs of their
domestic competitors.
Burrowing Up from Below
Multinationals for years underestimated the
ability and desire of local players in the low
end of the market to move up and compete—
a miscalculation that may now be coming
home to roost. Recent developments have
strengthened local competition in China and
facilitated Chinese companies’ moves upmarket and beyond.
Let’s start with consolidation. For years,
there were often hundreds of companies in a
single industry catering primarily to customers in the low end of the market and typically focusing on regional needs. Many of
those companies operated unprofitably—
think of Red Star Appliances or Wuhan Xi
Dao Industrial Stock. Because of China’s
free-market reforms, however, the weakest
of those competitors are folding, and industries are experiencing waves of consolidation. Red Star, Wuhan Xi Dao, and 16 other
money-losing concerns shifted and reshifted
throughout the 1990s to form appliance
maker Haier. A competent player or two,
like Haier, have risen in each industry, often
benefiting from national support. China’s
booming economy has enabled these survivors to build scale and develop market capabilities such as R&D and branding. As we
Penetrating the Good-Enough Market, One County Hospital at a Time
GE Healthcare already had a successful business selling high-end medical equipment in
China when the Chinese government set a
goal for the next decade of improving the
health care available in less-privileged locales.
To support the government’s efforts and also
to break out of the high end of the market, GE
developed a business case for manufacturing
and selling medical devices for China’s goodenough market. CEO Jeff Immelt’s visits and
conversations with Chinese leaders motivated
the company to pursue the opportunity. In the
end, GE’s research and analysis identified a
substantial demand from thousands of
midtier and low-end Chinese hospitals in less
affluent provinces that were not served by
multinationals. GE knew that it could design
new products and business models to serve this
market. GE also knew that by using techniques
like Six Sigma to eliminate manufacturing
waste, it could make its costs competitive.
A team was charged with observing operations in the target hospitals and meeting with
the hospital administrators and physicians to
help determine what sort of medical equipment customers wanted, the specific features
they needed, possible price points, and the
kinds of distribution and services that would
be required. Armed with this information, the
fact-finding team considered stripping out
some of the expensive equipment features and
adding others that these target customers
valued more. For instance, doctors in China’s
high-end hospitals preferred to program the
harvard business review • september 2007
medical equipment themselves, whereas physicians in the midlevel and low-end hospitals,
who considered themselves less computer
savvy, preferred preprogrammed machines.
The team worked with staffers in GE’s R&D
and manufacturing groups to build the right
products at the right price points for the
good-enough market. Because GE’s existing
sales, distribution, and service systems were
not geared to the target customers, the company also had to reconfigure its networks of
existing representatives and recruit new
ones. This middle-market initiative is still a
work in progress, but GE Healthcare has
taken an enormous first step in establishing
itself—and defending itself against rivals—in
the good-enough segment.
page 7
122
The Battle for China’s Good-Enough Market
Chinese customers are
becoming less willing to
shell out 70% to 100%
premiums for
international products.
At most they may pay
20% to 30% more for
world-class brands.
have seen, over time, several of these emerging domestic champions have become direct
challengers to global companies in a variety
of industries.
Next, look at the rapidly expanding customer base in the middle space. Chinese
customers—whether individual consumers,
businesses, or government agencies—are becoming less willing to shell out 70% to 100%
premiums for international products. At
most, they may pay 20% to 30% more for
world-class brands. The Italian dairy giant
Parmalat discovered exactly that when it tried
selling fruit-flavored yogurt for the equivalent
of 24 cents a cup. Instead, consumers went
with local brands at half the price. It seemed
that brand, innovation, and quality—the
hallmarks of multinationals in China—were
no longer critical points of differentiation
in customers’ minds. This price sensitivity
is opening up new ground for ambitious
Chinese companies traditionally focused on
the low end. These firms are designing and
releasing good-enough products that overcome buyers’ skepticism about quality at
much lower prices, which generate higher
margins than their low-end products. The
often brutal competitive dynamics in the lowend segment also serve as a huge incentive for
the better-managed local companies to move
up. Until consumer demand began to explode
in China, however, there really wasn’t anywhere for these firms to go. Now there is.
The journey from low end to good-enough
to global usually takes a decade and then
some—but more and more Chinese companies are embarking on it. For instance,
Lenovo, founded in 1984, entered the goodenough segment via a joint venture, flourished in the middle market, and then went
on to establish its international brand with
the purchase of IBM’s PC division in 2005
for $1.75 billion. It is currently the world’s
third-largest PC maker. Similarly, Huawei
Technologies has grown since 1988 to the
point where 31 of the first 50 firms on Standard & Poor’s ranking of the world’s top
telecom companies are clients of the Chinese
maker of mobile and fixed telecommunications networks.
Just as foreign players approaching the
market from above come face-to-face with
their shortcomings—high costs, limited distribution capabilities, and the possibility of
harvard business review • september 2007
cannibalizing their own products—local
companies moving up encounter their own
limitations. Foremost is the shortage of managerial talent, especially for international
businesses. Growing numbers of Chinese
students are pursuing MBAs and studying
abroad. They are slowly distinguishing themselves from the large cohort of current Chinese managers, whose command-and-control
leadership style dominates local manufacturing houses. But catching up remains difficult,
as China’s surging economic growth outpaces the country’s ability to educate and
apprentice twenty-first-century managers.
Another obstacle for Chinese companies is
their inability to compete with global players
through innovation or by establishing a strong
brand because of their limited size and their
lack of management tools and experience. A
question like “How much should we spend on
advertising?” can stymie local managers looking at expansion. Long used to competing
solely on price, they have little experience
in understanding and addressing segmentspecific needs, linking those needs to R&D and
brand-building efforts, and creating the required infrastructure in sales and distribution.
Consider the early successes enjoyed by
Chinese handset manufacturer Ningbo Bird.
It was among a group of small, local companies that took 20% to 30% of the telecom
market between 2000 and 2002 from the
likes of Nokia and Motorola. Ningbo Bird
prevailed by competing on price. But its success was short-lived, its march toward global
expansion thwarted. The company just didn’t
have the expertise and resources the foreign
corporations had in customer segmentation,
R&D, innovation, and distribution.
By contrast, Huawei has been able to successfully navigate such roadblocks. Initially
established as a network equipment distributor, Huawei has built and acquired the technical and managerial capabilities it needed
to rise up from the low end of the market.
From its inception, Huawei invested 10% of its
sales in R&D. It developed its own products
to penetrate new segments in China and
forged technical alliances to further broaden
its product mix. With government support,
Huawei prompted consolidation in the domestic market, gaining massive scale in the
process. The company now controls 14% of
the local market for telecom networks.
page 8
123
The Battle for China’s Good-Enough Market
Firmly established in the good-enough space
at home, Huawei built brands to meet the
requirements of global customers. It established 12 R&D centers around the world,
pioneering next-generation technologies
(customized communication networks and
voice access systems) and partnering with global brands such as 3Com to build customer
awareness of its own brands.
Huawei has broadened its reach in stages
over 14 years. The company first focused on
establishing itself in developing regions of
China, where multinationals had less incentive to compete. It then penetrated countries
with emerging economies, such as Russia
and Brazil. Finally, it attacked the developed
countries. It has expanded internationally
through aggressive sales and marketing, by
taking advantage of low-cost China-based
R&D, and by leveraging its ability to outsource some of its manufacturing processes
to other players in China. A little more than a
decade ago, Huawei was a regional company
in a local market that few multinationals
considered important. With 2005 revenues
of $8.2 billion, it is now second only to Cisco,
according to InfoTech Trends’ ranking of the
networking hardware industry. It could
never have ascended the way it has without
using China’s good-enough segment as a
springboard for growth.
Buying Your Way In
For multinational companies that can’t alter
their costs or processes quickly enough to
compete with local players, and for Chinese
firms that lack the production scale, R&D
mechanisms, and customer-facing capabilities
to compete with foreign players, there is still a
breakthrough option for entering the middle
market—mergers and acquisitions.
China’s entry into the World Trade Organization in 2001 fueled a surge in M&A activity.
Now, however, foreign acquirers are facing
tougher approval processes. China’s public
commitment to open markets remains strong,
but several high-profile deals have gotten stuck
at the provincial or ministerial level, owing to
increasing public concerns about selling out
to foreign firms. For instance, in its bid to
buy Xugong Group Construction Machinery,
China’s largest construction machinery manufacturer and distributor, the U.S. private equity
firm Carlyle Group met with unexpected resis-
harvard business review • september 2007
tance from the government and ended up
twice reducing its stake, ultimately to 45%. In
rejecting successive Carlyle bids, officials in
Beijing insisted the nation’s construction
equipment industry should be controlled by
“domestic hands.”
As the Carlyle Group learned, gaining regulatory and political approval for M&As in
China is a major undertaking. Foreign companies seeking such approval may need to draft
(and redraft) a compelling business case for
the acquisition, one that cites up front the
benefits for local companies and authorities.
Like Carlyle, they must be willing to adjust
(and readjust) the structure, terms, and conditions of a deal to gain government support.
They may also need to engage in heavy-duty
relationship building, investing the time and
resources required to woo critical players in
the deal.
As is always the case with M&As the world
over, it’s all about fit: There should be cost and
distribution synergies between the multinational and its target and little chance that the
local company’s products will cannibalize the
multinational’s premium brands. Successful acquirers in China—multinationals and Chinese
firms alike—use a clear strategic rationale to
select the right target. They overinvest in the
due diligence process. They take a systematic
approach to postmerger integration.
That was the game plan behind Gillette’s
2003 acquisition of Nanfu, then China’s leading battery manufacturer. Gillette’s Duracell
division throughout the 1990s was losing market share in China to lower-priced competitors.
By 2002, Duracell’s share of the Chinese domestic battery market was 6.5%. By contrast,
Nanfu controlled more than half the market.
After careful analysis, Gillette’s management
team recognized that its Duracell unit was at a
cost disadvantage compared with its rivals and
concluded it would be difficult to broaden the
brand’s market penetration. Facing such odds,
Gillette decided to buy into the good-enough
market, acquiring a majority stake in Nanfu.
But Gillette was extremely careful to protect
both Duracell’s and Nanfu’s brands in their respective segments. Gillette continues to sell
premium batteries in China under the Duracell
brand and has maintained Nanfu as the leading national brand for the mass market. The
dual branding, cost synergies, sales growth,
broadened product portfolio, economies of
page 9
124
The Battle for China’s Good-Enough Market
Many Chinese
companies believe they
must forge ahead and
buy established Western
brands and distribution
systems whether or not
they have the experience
and management tools to
handle them.
scale, and distribution to more than 3 million
retail outlets in China have paid off for
Gillette, which has seen significant increases in
its operating margins in China.
Buying into the good-enough segment also
worked for consumer-goods giants Danone,
L’Oréal, and Anheuser-Busch—companies
that saw the vast potential in China but
couldn’t get their costs low enough to compete. For instance, in 2004, Anheuser-Busch
outbid its competitor SABMiller to acquire
Harbin, the fourth-largest brewer in China.
That acquisition allowed Anheuser-Busch to
reach the masses while preventing Harbin
from swimming upstream. The next year, it
increased its stake in Tsingtao Brewery, from
9.9% to 27%. Both moves enabled the global
brewer to rapidly increase its share among
China’s drinkers of less-than-premium beer.
Chinese companies are also wrapping their
arms around acquisition strategies, attempting to establish their presence in the middle
market by purchasing brands, talent, and
other resources from target companies in
Europe and North America. To date, they’ve
met with mixed results. On the one hand,
Lenovo’s acquisition of IBM’s PC division
turned the Chinese computer maker into the
world’s third-largest PC company. On the
other hand, the acquisition experiences of
TCL, a major Chinese consumer electronics
manufacturer, have been less successful.
TCL built a strong position in the Chinese
market by producing and distributing basic
cathode-ray tube TVs at astonishingly low
prices. It also engaged in contract and privatelabel manufacturing for the U.S. and European markets. But TCL realized it would
need a strong brand to rise up from the low
end of the China market and that growing
organically in a mature industry like TV manufacturing would be prohibitively expensive.
So TCL acquired French firm Thomson, which
owned a number of well-known brands, including RCA. Unfortunately, Thomson also
owned some high-cost and unproductive
manufacturing facilities in France. TCL has
struggled since acquiring Thomson, as the
market for TVs has shifted from cathode-ray
to plasma and LCD technologies. In 2006,
the company lost $351 million from operations. Many Chinese companies believe that
in order to play in the global arena, they
must simply forge ahead, buying established
harvard business review • september 2007
Western brands and distribution systems—
whether or not they have the experience and
management tools to handle such acquisitions. But, as TCL’s story suggests, executing
such a plan is hardly cut-and-dried.
•••
In the 1960s and 1970s, the mantra for many
organizations was “Capture U.S. market
share, capture the world.” Today, China—and
its middle market in particular—has become
the object of multinationals’ ardent pursuit.
The enormous market potential of the country’s population, the formidable growth of
the economy, and China’s established position in low-cost sourcing and manufacturing
are providing competitive advantages for
many companies—benefits these organizations are then leveraging both inside and
outside the nation.
Local Chinese companies know their futures depend on entering the good-enough
space and attacking global leaders (and their
premium positioning) by offering low-cost
products of reasonable quality that they can
eventually take to the world. Multinationals
are beginning to recognize that ceding the
middle space to Chinese firms may breed
competitors that will ultimately challenge
them on a global scale. Ironically, Chinese
companies that have already gone global are
on the defensive as well. A recent Forbes Asia
article reported that as Haier has attacked international markets and won share abroad,
both local companies and multinationals
have been nibbling away at its share of
China’s middle market—which fell from 29%
in 2004 to 25% last year.
The stakes are high. All the more reason,
then, for companies that have stumbled in
China in the past to redouble their efforts.
Danone’s high product costs thwarted its
early attempts to sell dairy products in
China’s middle market. But that obstacle
was removed when the firm reengaged in
the fight, lowering its costs by buying a
local dairy.
Likewise, Caterpillar hasn’t diverted its
focus away from China and the importance
of the good-enough space. The company
plans to triple its sales by 2010, opening
more manufacturing plants and dealerships
and forming more joint ventures with local
companies. “Operational and sales success
in China is critical for the company’s long-
page 10
125
The Battle for China’s Good-Enough Market
term growth and profitability,” said Rich
Lavin, vice president of Caterpillar’s Asia
Pacific Operations Division, in November
2006. Shortly thereafter, the company moved
its divisional headquarters—from Tokyo
to Beijing.
Reprint R0709E
To order, see the next page
or call 800-988-0886 or 617-783-7500
or go to www.hbrreprints.org
harvard business review • september 2007
page 11
126
The Battle for China’s Good-Enough Market
Further Reading
HBR ARTICLE COLLECTION
Winning in the World’s Emerging
Markets
by Tarun Khanna, Krishna G. Palepu,
Jayant Sinha, Ming Zeng, and
Peter J. Williamson
Harvard Business Review
October 2006
Product no. 1455
This collection provides additional ideas for
entering China’s good-enough space. In
“Strategies That Fit Emerging Markets,” Tarun
Khanna, Krishna G. Palepu, and Jayant Sinha
recommend assessing market institutions in
the parts of China where you’re considering
introducing new offerings. For instance, how
strong are transportation infrastructures?
Then, decide whether altering your business
model will help you compensate for any weak
institutions. Dell Computer, for example, decided to sell its products in China through
local distributors and systems integrators after
discovering that Chinese consumers don’t
buy over the Internet.
To Order
For Harvard Business Review reprints and
subscriptions, call 800-988-0886 or
617-783-7500. Go to www.hbrreprints.org
For customized and quantity orders of
Harvard Business Review article reprints,
call 617-783-7626, or e-mai
customizations@hbsp.harvard.edu
specialized, high-value products (such as refrigerated containers). 3) Competitive networks
comprise hundreds of local entrepreneurial
companies that achieve scale and reduce
cost by using the same suppliers, labor, and
distribution channels. 4) Technology upstarts
are state-owned research institutions that
commercialize their technologies and spawn
new firms.
In “Emerging Giants: Building World-Class
Companies in Developing Countries,”
Khanna and Palepu advise analyzing Chinese
upstarts’ best practices to better position your
firm to compete in China’s good-enough space.
For example, Chinese enterprises exploit their
knowledge of local consumers. They leverage
familiarity with local talent and capital markets
to cost-effectively serve customers at home
and abroad. And some treat lack of market institutions as business opportunities, providing
new banking, insurance, and product-rating
services that turn into big companies.
In “The Hidden Dragons,” Ming Zeng and
Peter J. Williamson suggest that, to compete
against new rivals from China, you need to
gain familiarity with four types of local contenders: 1) National champions identify and
serve market segments that global giants
have missed. 2) Dedicated exporters develop
expertise with crucial technologies to export
page 12
127
Sunday, October 2
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Sunday, 2 October
08:30
Finance Session 8 – Professor Callahan
Topic: Quiz (covering material through Finance session 6)
Required Readings and Discussion Questions: None – study for the quiz.
To do before this session: Study for the quiz.
09:45
Break
10:00
Finance Session 9 – Professor Callahan
Topic: Raising Long-Term Capital
This session provides an opportunity to discuss how firms raise external
capital, what types of capital are acquired and why, and the comparative
costs and benefits of strategic capital acquisition. Topics will include
concepts of market efficiency and the process and costs of underwriting
common stock and debt.
Required Readings: RWJ Chapters 14 and 20
Discussion Questions:
1. Do you believe markets are efficient? Why or why not?
2. What do your views on market efficiency imply about firms’ financial
policies?
3. Think about whatever questions you have regarding the institutional
details or mechanics of stock or bond issuance.
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
11:15
Break
11:30
Marketing Session 9 – Professor Nunes
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
128
Page: 1 / 2
Global Executive MBA in Shanghai
Please review the results of your practice Markstrat Decision. We will
discuss and I will answer questions.
Topic: Brand Identity & Brand Image
Required Readings: Ward, Scott, Larry Light and Jonathan Goldstine: “What
High-Tech Managers Need to Know about Brands”, HBR Reprint 994II
12:45
Lunch
13:45
Marketing Session 10 – Professor Nunes
Topic: Pepsi Case
Required Readings: Handed out in Class
Assignment: Make Decision 1
15:00
Break
15:15
G7 and G8 Joint Session
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
129
Page: 2 / 2
What High-Tech Managers
Need to Know About Brands
by Scott Ward, Larry Light, and Jonathan Goldstine
Reprint 99411
130
JULY– AUGUST 1999
Reprint Number
Jim Collins
TURNING GOALS INTO RESULTS: THE POWER
OF CATALYTIC MECHANISMS
99401
Scot t Ward, L arry Light, and
Jonathan Goldstine
WHAT HIGH-TECH MANAGERS NEED
TO KNOW ABOUT BRANDS
99411
Peter Frost and
sandr a Robinson
THE TOXIC HANDLER: ORGANIZATIONAL HERO –
AND CASUALTY
99406
Cl audio FernA ndez-ArAoz
HIRING WITHOUT FIRING
99403
Tarun Khanna and
Krishna Palepu
THE RIGHT WAY TO RESTRUCTURE CONGLOMERATES
IN EMERGING MARKETS
99407
Robert G. Eccles, Kersten L . Lanes,
and Thom a s C. Wilson
ARE YOU PAYING TOO MUCH FOR THAT ACQUISITION?
99402
Forest L . Reinhardt
BRINGING THE ENVIRONMENT DOWN TO EARTH
99408
R ay Friedm an
THE CASE OF THE RELIGIOUS NETWORK GROUP
D onald N. Sull
WHY GOOD COMPANIES GO BAD
HBR CASE STUDY
99405
thinking about…
99410
so cial enterprise
Larry Fondation, Peter Tufano,
and Patricia Walker
COLLABORATING WITH CONGREGATIONS: OPPORTUNITIES
FOR FINANCIAL SERVICES IN THE INNER CITY
99404
perspectives
Introduction by
Regina Fa zio M aruca
RETAILING: CONFRONTING THE CHALLENGES THAT FACE
BRICKS-AND-MORTAR STORES
Peter Sealey
HOW E-COMMERCE WILL TRUMP BRAND MANAGEMENT
99412
bo oks in review
131
99409
What High-Tech Managers
Need to Know About Brands
Brands are not just names slapped on products by
the marketing department; they embody the value
those products have for your customers. That may be
more true for high-tech products than it is for soap.
by Scott Ward, Larry Light, and Jonathan Goldstine
M
any managers in high-tech companies believe that
market success depends primarily on the price-performance
ratio. At the same time, however, most would acknowledge
that the bulk of their offerings are fast becoming commodities (if they
are not already). Products and services are highly similar – printers
print and computers compute. And if one manufacturer boasts of more
“feeds and speeds” today, competitors will catch up tomorrow. Price
and performance are just the ante to get into the game.
art work by michael klein
Copyright © 1999 by the President and Fellows of Harvard College. All rights reserved.
132
b r a n d s f o r h i g h - t e c h m a n ag e r s
What, then, can make the difference between a
successful high-tech venture and an unsuccessful
one? One critical factor is brand management. The
problem is that many of the people leading hightech companies – managers who have grown up
through the technical side of the business – do not
truly understand what good brand management involves and what it can do for their companies. Most
think of marketing as merely selling, and branding
as an advertising campaign or a slogan – necessary
evils that are costly, difficult to assess, and antithetical to a business model built on delivering the
highest performance at the lowest price. The idea of
developing and maintaining a strong brand in the
fullest sense – conceiving of a promise of value for
customers and then ensuring that the promise is
kept through the way the product is developed, produced, sold, serviced, and advertised – simply does
not resonate as it should.
What’s needed is a sea change in managerial attitudes from a product-centric to a promise-centric
business model. Successful brand management
helps attract and keep customers. It can also be a
strong foundation from which to launch new products, improve relationships with channel partners,
foster good communication among employees
within and across business functions, and help a
company better focus its resources.
Changing long-held beliefs isn’t easy. But our experience suggests that the first step toward doing so
is to recognize the barriers that may be impeding
the change. In that spirit, we’ll first address what
we feel are the two most prevalent misconceptions
about brands in high-tech markets. Then we’ll elaborate on our definition of a brand and related concepts that are particularly important in high-tech
businesses. Finally, we’ll suggest some specific
steps senior managers can take to move their companies from a product-centric to a promise-centric,
brand-driven business.
Two Misconceptions About Branding
Most technically trained managers reflexively reject the idea that their businesses could be focused
around brand management, and they base that judgment on these two assumptions:
n Brands and brand images are relevant only when
purchase decisions are “irrational” or “emotional,”
which is fine for detergents, automobiles, and fashions. But they have nothing to do with markets
populated by highly sophisticated and experienced
customers, nothing to do with purchase decisions
based on benchmarking studies and objective performance data.
Brand management is best left to the marketing or
sales departments; it’s not central to the technical
direction of the company. A brand is just a logo,
trademark, slogan, or ad campaign, and marketing
handles those things.
Let’s tackle the first assumption first. It is true
that most of our knowledge about brand strategies
comes from the accumulated experience of consumer-packaged-goods companies like Procter &
Gamble, Nabisco, and Nestlé. Such companies invented brand management – and a wealth of enduring and highly profitable brands. But just because a
concept evolved in the consumer goods markets is
no reason to reject it in business-to-business markets in general or in high-technology markets in
particular.
A successful brand in any business commands
enduring premium profits because substantial
numbers of customers view it as a promise of receiving a certain type and level of value. Real value
is at the heart of strong relationships with customers. And value need not be defined narrowly in
terms of features and performance. Value in hightech markets, as in other markets, is a much more
complex concept.
“Price” and “performance,” for example, may
mean different things to different people. Does
price mean initial cost or life cycle cost? Does performance mean microprocessor speed, the product’s ability to keep pace with a company’s growth,
or a vendor’s ability to deal effectively with a multiplatform environment? Does performance include
presale and postsale service and support?
Most customers’ evaluations of price and performance include multiple definitions and dimensions,
and the trade-offs individuals make in their buying
decisions reflect different definitions of value and
different needs. Through strong brands, high-tech
companies can make it clear exactly which aspects
of their offerings’ price and performance benefit
their customers.
We acknowledge that an individual’s decision to
buy a certain brand of soap may be more emotional
or irrational than his or her decision to buy a particular phone or than a company’s decision to go with
a certain line of fax machines, at least as those
terms are popularly defined and understood. Consumers may buy a certain brand of cosmetic or
clothing to express a particular lifestyle, something
that many people think shouldn’t apply to hightech purchases.
But we argue that in markets such as personal
computers, the line between a high-tech product
and a consumer product is becoming increasingly
fuzzy. And parallel kinds of motives exist even in
n
86
harvard business review
133
July–August 1999
b r a n d s f o r h i g h - t e c h m a n ag e r s
High-Technology Brands:
Misconceptions Versus Realities
MISCONCEPTIONS
REALITIES
Technology products are bought on the basis
of the price-performance ratio, period.
Price and performance are important, but other factors
may also be highly influential. Additionally, the people
involved in purchase decisions may weigh various
performance factors differently.
High switching costs associated with a large
installed base are the key to profitability.
True for a while, but customers don’t like to feel
trapped, and competitors are dedicated to offering
seamless transitions, along with better performance
and functionality.
Brand management is used when product
differentiation is difficult or impossible.
By then it’s too late. A brand’s promise of value is the
core element of differentiation, not an alternative to it.
Branding is something the marketing department
does, and as such, it means advertising, trade
shows, and sales literature. The results are hard
to measure.
The promise of value must be reflected in every aspect
of the complete product offering in tangible and
measurable ways. Even psychological rewards among
brand users, like trust, can be measured and related
to business performance.
business-to-business high-tech purchases. Consider
the following:
n Many hard-nosed M.B.A. students purchase a relatively costly Hewlett-Packard calculator instead
of a less expensive brand that performs just as well
to signal their proficiency with technology to
themselves and to others.
n The vast majority of people who use personal
computers would not recognize the microprocessor
in their PCs if the machine were disassembled and
the parts arrayed in front of them. However, Intel’s
success in establishing a strong brand has arguably
influenced consumers’ purchasing decisions, effectively shifting some brand equity its way from PC
manufacturers.
n When managers of information technology weigh
the merits of a vendor’s bid, they may be thinking
not only about how well the vendor’s products and
services will complement their own company’s
business and technological environment but also
about how the purchase decision will look to others.
That is, IT managers may be just as concerned with
the image of the product – and with how the purchase decision will reflect on them – as they are
with the technical strengths or weaknesses of the
vendor’s offering.
harvard business review
How many people do you know once claimed that
they would be the last Apple buyer? Even at its lowest point, Apple still enjoyed considerable brand
loyalty from a devoted group of followers for whom
that brand was synonymous with the Apple’s simplicity and its utility for their particular needs. And
witness the enormous popularity of the iMac, the
result of Steve Jobs’s thus-far-successful efforts to
refocus the company’s marketing and productdevelopment strategies.
In short, a rational decision means one that maximizes value, but value can be based on objective or
subjective data.
As for the second assumption, consider this: It’s
true that creating logos and advertising campaigns
are a part of marketing. But those activities will
fall short of their intended goals if they are devel-
n
Scott Ward is a professor of marketing at the University
of Pennsylvania’s Wharton School in Philadelphia.
Larry Light is president of Arcature, a brand-strategy
consulting company in Stamford, Connecticut.
Jonathan Goldstine is program director for AIX strategy
in the Server Group of IBM in Somers, New York.
87
July–August 1999
134
b r a n d s f o r h i g h - t e c h m a n ag e r s
oped and implemented without some sort of unifying foundation. The foundation for all marketing
activities should be creating and nurturing a promise of value to customers. A promise of value (the
brand) – and delivery on that promise – is critical if
a company is going to differentiate itself from its
competitors and stake a solid claim in its intended
market. If all functions in an organization are not
helping to create and nurture a single promise of
value, customers will be, at best, confused and, at
worst, angry.
It’s true that building and positioning a promise
of value for a high-tech product or service often
begins with a prototype core offering that may be
understood and purchased initially by only a tiny
cadre of technophiles, so the importance of the
promise of value may not be widely apparent right
away. But for that product to achieve some degree of
broader market acceptance, the product will likely
need to attract not only a broader base of customers
but also a network of ancillary products and services.
For example, a server’s prowess in accommodating
ever-increasing workloads, its scalability, is of little
value if it lacks supporting applications written by
independent software vendors.
A company cannot expect to build that broader
acceptance simply by making a promise of value
in advertising. It must
first know what promise
to make and to whom.
That requires the ability
to assess the potential of
relevant technologies
and to anticipate its customers’ current and future needs – even before
customers can articulate
those needs. It must also
ensure that the promise
of value is understood
and fulfilled as the company manages complex
networks of value-adding
partners, ranging from
consulting firms to systems integrators, from
independent software vendors to resellers.
Gateway Computer Systems is a good example
of a company that understands the importance of
marketing as a companywide function. Gateway
differentiates itself from other mail-order PC companies through folksy advertisements in otherwise
terse and dense catalogs. But the company knows
that its marketing message – in essence, a promise
of friendly service – must be backed up by efficient
Any high-tech
company
attempting to
build a brand
must pay for the
mistakes other
companies have
made.
help lines and effective order and service fulfillment to ensure that its promise of value is consistently kept.
If these arguments aren’t enough to dispel the
misconceptions, consider one more thing: some
high-tech companies enjoy considerable brand
equity, or financial value, that can be attributed to
a brand name. Essentially, a brand’s value is derived
by multiplying its annual net after-tax profits, adjusted to exclude the earnings expected for an
equivalent generic product, by a discount rate that
reflects the brand’s strength as defined by such factors as the brand’s ability to influence the market,
the stability of its customer franchise, the ability to
sustain demand in the face of technological change,
and the strength of supporting communications.
Calculated according to that methodology, the
world’s most valuable brands include IBM, Microsoft, Intel, and Hewlett-Packard.1
Fine-Tuning the Definition of a Brand
Brands are often confused with logos or trademarks.
A trademark is a distinguishing name, sign, symbol,
or design, or some combination of them, that identifies the goods or services of one seller. A brand is
a distinctive identity that differentiates a relevant,
enduring, and credible promise of value associated
with a product, service, or organization and indicates the source of that promise. Emphasis can be
placed at the corporate level or at the level of a subbrand. For high-tech brands, the emphasis is often
at the corporate level because specific offerings are
generally configured for specific clients, so identifying a subbrand isn’t feasible. IBM does not have subbrands for its e-business offerings, for example, since
its solutions are unique to each of its servers and include software, service, and support provided by
business partners. By contrast, Lotus shares brand
equity with its subbrand Domino. Just as a trademark must be seen and recognized by customers in
order to perform its identification function, so too
must a brand’s promise of value be tangible and predictably manifested in a company’s business behavior and, ultimately, in its products and services.
Brands can be built around many different promises of value. We’ve said that performance is the
price of admission in high-tech markets; nevertheless, for some high-tech companies, the promise of
value can center around the related idea that they
offer cutting edge products and services. Lucent
Technologies’ current brand-building effort is a
case in point. Lucent’s message is, in effect, “We
know competitors will match us, but look to us to
be the first with the newest technology.” By con-
88
harvard business review
135
July–August 1999
b r a n d s f o r h i g h - t e c h m a n ag e r s
trast, IBM is not often first to market with the latest
technology; instead, its promise of value is built on
its long tradition of superior service and support.
The president of one of its major European customers told one IBM senior executive, “As long as
IBM is in the general price range of competitors,
we’ll always buy IBM for the service and support
you deliver.”
Of course, the promise of value must be relevant
to the people or businesses a company wants to
have as its customers; no business can promise
everything for everyone. But it is possible to tailor
a promise to meet the desires of a specific subgroup
of customers, as long as that promise can be kept
and as long as it doesn’t confuse anyone. That’s why
a company must consider its own capabilities and
its target customer segments before developing its
brand message. Consider how IBM tweaks its brand
message for application development tools. Traditionally, target customers for the tools were IT
managers in large organizations. Those managers
valued service, support, and reliability. However, a
new breed of customer has emerged in recent years.
These customers are young, sophisticated, highly
competent, and comfortable with technologies.
They care less about IBM’s tradition of service and
support than they do about being able to work with
so-called bleeding edge technologies. Some even
look forward to the hassles of working with new
technologies for the intellectual and technical challenges they pose. Accordingly, IBM’s promise of
value for this group emphasizes technical prowess,
layered on top of its traditional message of solid
service and support.
EMC, a manufacturer of data storage systems,
knows the importance of choosing a particular type
of customer before defining the promise of value.
Attempting to avoid the price wars that resulted
from marketing to technically sophisticated buyers
in IT departments, EMC instead tailored its promise for senior managers who could dictate purchasing decisions. In a series of advertisements in publications like the Wall Street Journal, EMC raised
relevant issues for senior executives, such as the
role of data storage in securing competitive advantage. EMC realized that in IT departments, purchasing decisions are often narrowly based on price,
but for senior executives, any decision reflects
broader considerations. As luck would have it,
EMC’s decision to create its promise of value for
senior executives, and to market to them, may have
bought the company some time in the mid- to late
1980s, when the company faced critical product
quality issues. It is hard to imagine any credible
promise of value that EMC could have made to IT
harvard business review
professionals and other technical customers who
were directly affected by the company’s quality
problems at that time. Under Michael C. Ruettgers,
who became EMC’s CEO in 1992, the quality problems have been solved, and profits have climbed
from $30 million in that year to an expected $1 billion in 1999.
Managers in high-tech businesses face many of
the same problems as managers in other businesses
in formulating distinctive and credible promises
of value. But in high-technology markets, offering
a distinctive value proposition is especially difficult since, with lightning speed, competitors
match one another ’s
clearly defined performance characteristics
(modem speed, microprocessor capacity, and
the like). What’s more,
many buyers have been
burned by both new and
established companies
that promised much
but delivered little – and
often delivered late, to boot. There are almost as
many business failures as there are start-ups in the
high-tech arena in a given year; any high-tech company attempting to build a brand must pay for the
mistakes other high-tech companies have made
along the way. There isn’t a great deal of trust between customers and companies in high-tech markets simply because there are many and frequent
new entrants, and most have not had the time or
the wisdom to build a promise of value that goes beyond price-performance.
Building trust is a worthy goal, however. The
credibility of a company’s promise of value results
from persistence and consistency. It would be difficult, for example, for a new entrant, or even an established one, to claim that it could fulfill phone or
Internet orders for personal computers better than
Dell does. And Apple’s devotion to simple, easy-touse computer systems continually shores up its distinctive promise of value. On the other hand, consider that Netscape had little competition for quite
some time for its Internet browser. Microsoft’s inroads into that company’s once-dominant position
in the browser market may be due to Netscape’s
failure to articulate clearly a distinctive and credible promise of value as much as to Microsoft’s
bundling of its browser into Windows software.
Lacking any loyalty to Netscape, many consumers
elected to “switch rather than fight.” Netscape
The credibility
of a company’s
promise of value
results from
persistence and
consistency.
89
July–August 1999
136
b r a n d s f o r h i g h - t e c h m a n ag e r s
might have preempted Microsoft by giving customers tangible reasons to stick with its browser.
Even after Microsoft’s entry, Netscape could have
questioned customers who defected and those who
stuck with Netscape to determine where its implicit promise of value was being fulfilled and where
it was not. With that knowledge, Netscape might
have emphasized the salient dimensions of its
value proposition in its advertising and modified
certain technical aspects of its offering along those
same lines as well.
Powerful brands make promises that are enduring. Making a promise is serious business, and, as in
personal life, making too
many promises, or changing them frequently, raises
uncertainty in the people
to whom the promises are
made. Making and keeping
a promise, and keeping it
consistently, can be a powerful source of competitive
advantage. Studies of the
U.S. consumer-packagedgoods industry show that
many brands that were
market leaders half a century ago are still market
leaders today, despite inroads from cheaper privatelabel brands and generic products. While being the
market leader half a century ago does not guarantee
leadership, or even survival, today (Remember
Schlitz, Woolworth’s, and PanAm?), there’s a higher
probability that truly strong brands will continue
to be leaders because they make and keep a promise
of value over successive generations of technologies.
No doubt IBM’s reputation for solid products and for
great service and support has helped its Server
Group catch up to a radically changed environment
for buying large computers. IBM realizes that it can
no longer be profitable if it just sells “the box,” so
the Server Group is aggressively pursuing new partners to provide a wider array of software applications, technical features, service, and support.
Time will tell if IBM’s past leadership in the mainframe market will continue in the new, and very
different, server market. True, newer technologies
can eclipse older ones; electromechanical calculating machines never had a chance against electronic
ones. But an enduring promise of value can buy
time for a brand in the face of new technology or, as
we suggested earlier in the discussion of EMC’s
products, in the face of serious lapses in product
quality.
Monopoly
power built on
overwhelming
market share is
not, in the end,
a promise of
value.
Consider the devotion of scientists and engineers
to Hewlett-Packard. If a competitor introduces a
superior product, HP’s brand equity gives the company a chance to at least match its competitor.
Tide’s brand equity allowed Procter & Gamble to
plan an orderly progression of products to respond
to competitors’ scented offerings, liquid detergents,
and proliferation of package sizes. The Novell
brand –and its promise of leadership in superior and
secure networking software – has bought that company time in the face of competitive inroads from
Microsoft NT and Netscape Calendar. Recently,
Novell unbundled its powerful Novell Directory
Services (NDS) from its flagship NetWare operating
system software and continued to fulfill its leadership promise with a new version, Scalable Directory
Services, or SCADS.
Managers in high-tech companies may fail to
think about the different promises their products
could fulfill because their instincts and training tell
them that good, cost-effective high-tech products
and services will succeed as long as they perform
as promised, satisfy customers’ needs, and garner a
large installed base that will withstand any upstart
brand. But such product-centric thinking is highly
dangerous because customers do not necessarily experience a company’s promise of value just because
its products enjoy wide acceptance. (Remember
WordPerfect?) There is a tendency among managers
to believe that a product’s success is ensured once
a large installed base is secured, since switching
costs pose a barrier to buying from other suppliers.
Witness the practice by high-tech vendors of giving
away software and hardware products in order to
penetrate a market rapidly. However, sheer presence and satisfactory performance do not ensure
that users will become loyal customers – that is to
say, customers who are eager to buy products, not
just customers who feel compelled to buy them. No
customers like to feel trapped by the pervasiveness
of a company’s hardware or software. Sooner or
later, customers will defect. Monopoly power built
on overwhelming market share is not, in the end, a
promise of value. People will buy because they feel
they have to buy, but they will not become loyal
customers.
Building a Powerful Brand
Powerful high-tech brands build equity through a
process we illustrate in our brand pyramid, which
is based on materials developed by Larry Light. (See
the exhibit “How High-Tech Brands Build Equity.”)
The pyramid’s bottom level represents the core
product – the tangible, verifiable product character-
90
harvard business review
137
July–August 1999
b r a n d s f o r h i g h - t e c h m a n ag e r s
istics. Many high-tech managers are most comfortable in this space, and, unfortunately, it is where
many high-tech products reside. Increasingly, however, high-tech purchases involve not just technologists but also business managers and end users,
who are far more interested in what a technology
product does for them than in how it works. As
high-tech managers come to understand this, many
start changing the way they speak of their offerings.
Instead of selling “products,” they sell “solutions”
or “benefits.” Such a shift in thinking and in language marks a step in the right direction; in fact, it
represents the second level of the pyramid. But it is
not enough. The first two levels of the pyramid still
embody the elements of product competition, not
those of brand competition. Competitors can continually match and leapfrog over one another by
offering better and more features and by identifying the benefits of their products for customers. For
instance, two manufacturers make Unix-based
servers that not only perform similarly but also produce the same benefits for customers –the ability to
run data-mining applications, executive information systems, and the like.
The third level of the pyramid is where a company can truly differentiate itself from competitors
by providing emotional rewards for its business.
How do customers feel when experiencing the
functional benefits of the offering? How do customers feel when experiencing the benefits of the
brand? Do they feel confident? Productive? Innovative? Caring? Responsible? Successful? It is exceedingly difficult for many high-tech managers to acknowledge, much less embrace, the idea that
emotions can be so important to a company’s success. But goods and services that reside in that third
level are indeed developed and positioned as a way
of fulfilling a promise of value to selected customers, not simply as technologies in search of a
market. In this regard, Apple comes to mind and
also IBM’s venerable Customer Information Control System (or CICS). Eckhard Pfeiffer, the former
CEO of Compaq, illustrated the important differences between the bottom and the top of the pyramid this way: “Consumers don’t go shopping for a
24-valve, six-cylinder, 200-horsepower, fuel-injected
engine. They shop for a Taurus, a Lexus, a BMW, a
Jeep Cherokee, a Hummer, whatever. They shop for
well-known, trusted brands.”2
While some may question the validity of that assertion, given Compaq’s disappointing recent earnings and Pfeiffer’s subsequent departure from the
company, the fact remains that Compaq did enjoy
considerable success in the PC market for many
How High-Tech Brands Build Equity
To build a strong high-tech brand, managers need to answer the following questions:
LEVEL 5
What does “value” mean for
the typical loyal customer?
LEVEL 4
LEVEL 3
LEVEL 2
LEVEL 1
What is the essential nature
and character of the brand?
What psychological rewards or emotional
benefits do customers receive by using this
brand’s products? How does the customer feel?
What benefits to the customer or solutions
result from the brand’s features?
What are the tangible, verifiable, objective, measurable characteristics of
products, services, ingredients, or components that carry this brand name?
harvard business review
91
July–August 1999
138
b r a n d s f o r h i g h - t e c h m a n ag e r s
years. Declining profit margins in the business as a
whole, fueled by such strategies as companies actually giving PCs away to encourage end users to get
on the Internet, have left the PC market barren of
strong brands. Compaq is currently pinning its
hopes on translating its promise of value for higherend technology applications. The company’s ads
boast of its entrance into the enterprise-computing
arena, pointing out that 17 of the 20 largest stock
exchanges in the world run on its systems and
claiming that Compaq “out integrates” the top IT
integrators. Time will tell if Compaq can leverage
its promise of value in this very different market.
The top two levels of the pyramid illustrate the
concept that powerful brands attract and hold customers with their particular promises of value. At
the top level of the pyramid is the personality of the
brand. It’s the characteristics the brand would have
if it had human qualities: friendly, warm, caring,
confident, decisive, aggressive, or the like. The next
level of the pyramid describes the deeper values
that the brand reflects. We are particularly interested
in reflecting values of the target customer that will
create and reinforce brand loyalty. By “values,” we
mean such things as conservative values, family
values, achievement-oriented values, and so on.
Taken together, these two levels of the pyramid define the relevant and differentiating character of
the brand. In the end, powerful brands enable customers to fill in the blank
in the following sentence
with ease: “Oh yes, [brand
name]; that’s the company
that
.”
Focusing a company
around brand management – getting from the
bottom two basic levels of
the pyramid to levels
three, four, and five – does
not mean tearing up the
organization chart, forming yet another set of
teams to explore a new initiative, or instituting a gaggle of new processes.
Business-planning processes and topics are the
same in a promise-centric company as they are in
a product-centric organization. A brand plan is a
business plan. The fundamental difference between
a product-centric and a brand-centric company lies
in the attitudes of the people throughout the organization – not just in the marketing department – in
their understanding of what it means to shift from
Focusing
a company
around brand
management
does not mean
tearing up the
organization
chart.
selling products or services to selling a promise
of value.
How can senior managers help the process along?
By doing the following:
Figure out what the company’s current promise
of value seems to be. What do customers think of
the company and its offerings? Do they see offerings as brands or merely as technology products?
How do customers feel about the total experience
of buying and using the company’s products? We
are not suggesting that taking this step is as simple
as sending out a survey that asks, “What emotional
rewards do you get from using this product?”
Rather, it is a matter of probing and listening. Consider, for example, the experience of one manufacturer of Unix-based servers. In a very competitive
marketplace – where nearly equivalent boxes are
produced by HP, IBM, Silicon Graphics, and Sun –
managers at one of these companies conducted a
series of focus groups with server buyers and users
around the world. They found that those people had
many and complex perceptions of the competitors,
the offerings, and the promises they felt the competitors made. In other words, they had a lot to say.
The most intriguing data were about the kinds of
promises of value the customers wanted but felt
some vendors did not or could not deliver. Armed
with the data, the company formulated several alternative promises of value and tested them out in
a second round of research.
Once the promise is clear, think about how that
promise relates to the company’s strengths and employees’ impressions of what the company is offering. Do the internal and external impressions
match? If not, how do they differ?
Refine the promise of value. This step requires
managers to make three crucial decisions. First,
they must understand how the potential customer
markets are segmented and choose the segments
they wish to serve. Second, they must determine
what type of promises are feasible to offer. Third,
they must set a branding policy – that is, they must
decide whether and how much the equity of any
one subbrand will be shared with other subbrands
and with the parent brand. Older, established hightech companies like IBM can generally rest a great
deal of equity at the parent-brand level. But at times
the parent may wish to allow a subbrand to build an
identity of its own. Such branding decisions must
be handled with care. The trick is to enhance the
promise of value for selected customer segments
without jeopardizing or diluting the core promise.
In some cases, of course, the company will want
to maintain a strong connection between its parent
brand and a given subbrand. GroupWise, NetWare,
92
harvard business review
139
July–August 1999
b r a n d s f o r h i g h - t e c h m a n ag e r s
ManageWise, and BorderManager, for example, are
all Novell subbrands, but it is not clear how much
customers associate those products with the parent
company or whether Novell could be building
stronger brand equity if it made the connection between the subbrands and the parent more explicit.
Determine what the company’s various business
functions must do to make good on the promise
of value. It may help at this point to try to distill
the promise of value to its essence. For example,
if the promise boils down to superior service and
support, then those capabilities must be fully developed, maintained, and monitored. Resources
should be directed not only at creating a team of
fully qualified service technicians for postsale service but also at developing Web-based service capabilities. Service and support should also be built
into technology offerings in the first place.
To make sure it could keep its promise, for instance, the server vendor hammered out its value
proposition – which centered around providing tailored solutions to help buyers achieve sustained
competitive advantage – with managers and employees from every business function within the
company. Those people suggested instituting specific objectives and measures to make the promise
clear and to make it work. For example, to deliver
on the promise of providing tailored solutions, the
company’s manufacturing operation needed to recruit and closely collaborate with the independent
software vendors that would tailor applications to
each buyer’s needs. To deliver on the promise to help
a buyer achieve competitive advantage, the company needed to go beyond selling boxes to understanding a customer’s business.
In some instances, the changes were minor – but
meaningful for customers. For example, the company instituted changes in its order fulfillment
processes. Previously, customers were told only
when their computer would be shipped. But customers in Asia wanted to know when their computers would arrive, so the company changed its procedures to take into account the delivery time to
offshore destinations.
Delivering on a promise of service and support
need not be costly – in fact, it can cut costs and produce profits. Customers, channel partners, and
technology vendors can all win. For example, consider Support Pack, which Hewlett-Packard offers
to customers buying its low-end, high-volume line
of printers, PCs, plotters, fax machines, and the
like. Support Pack is a service package in a box designed for computer superstores and other resellers
that distribute HP’s products. Previously, a customer would be offered the opportunity to purchase
harvard business review
a postsale service contract immediately after purchasing an HP product. But channel partners are
much better at selling physical products than services, so HP set out to make after-sale service profitable, both for the company and for its channel
partners. Support Pack is an actual box, displayed
next to HP products.
It’s designed to be easy
for customers to buy
and channel partners to
sell and earn significant
margins on: Support
Pack sells for 10% to
20% of the price of the
corresponding hardware
product. It contains a
“Dear Customer” letter with instructions
for registering as a Support Pack customer, a
mail-in registration
card, information about
service terms and conditions, and options for
obtaining service. With
this service product,
HP cuts down on incoming service calls, dealers get a product they
can sell at a good profit, and customers are reassured that they will get service and support if they
need them.
Institute measures of brand performance. Begin
to develop the algorithms that relate those measures to business performance. For example, it’s important to track how customers perceive the brand
relative to its competition over time. It’s also important to monitor the health of key dimensions of
a brand’s promise of value. Managers should try to
modify research on customer satisfaction levels
to determine if satisfaction is leading to loyalty or
if satisfaction levels are simply high or low for all
players in the market. The company should try to
gather information that reveals which customers
are true brand loyalists and why others defect.
These data can be extremely useful as managers try
to decide where to place resources and how the
business might more extensively change its activities to fulfill its promises.
Business
planning is the
same in a
promise-centric
company as it is
in a productcentric company.
A brand plan is a
business plan.
The Rewards of Brand Management
As we’ve suggested throughout, high-tech businesses that actively manage their brands stand to
gain a great deal. For example, as we’ve discussed,
allocating resources becomes a more direct process
93
July–August 1999
140
b r a n d s f o r h i g h - t e c h m a n ag e r s
What Lies Between Your Brand and Your Customers?
One of the difficulties in maintaining a powerful
brand in high-tech markets stems from the involvement of numerous partners in the production, distribution, and support of complete technology products.
Many high-tech products contain components from
several suppliers and are sold through a variety of
channels. For example, a server such as IBM’s AS/400
contains components manufactured by several suppliers, and it may be bundled for sale with software written by independent software vendors and sold by systems integrators, computer resellers, or some other
value-adding entity that makes and maintains contact
with customers. Can IBM make, deliver, and control a
relevant, distinctive, and enduring promise of value to
the end user through so many filters?
Companies in many industries use numerous
sources for their product components. However, the
brand name indicates where the buck stops. If a buyer
of a Ford Mustang has a problem with wheel bearings
made in Mexico or engine mounts manufactured in
eastern Europe, the problem must be resolved by Ford,
through one of its dealers, or ultimately in Dearborn.
Because of the relatively great need for service and
support associated with many high-tech products, the
issue of who, exactly, is the seller is critical. To use a
well-known example, the “Intel Inside” microprocessor may fail, but the computer company that installed
the part is responsible to the customer. The computer
manufacturer may deal with Intel, but that’s of no
concern to the consumer. There are variations on this
scenario, but the theme is the same – customers
don’t want to buy a basket of hardware and software;
they want to buy a complete offering, and they want
reassurance from somebody. That somebody is the
creator and owner of the brand.
Many companies also use multiple routes to distribute products to customers. They’re responding to the
growth in distribution channels to serve segments of
customers who want and get alternate ways to buy
products and the need all companies have to reduce
the high costs associated with direct sales forces. The
complexity of high-tech products and their market
when a company is focused on filling one value
proposition across all business functions. In fact,
brand management can become a powerful unifying force throughout the company. High-tech companies usually contain various camps of people
from very different backgrounds. Product developers with strictly technical backgrounds may have
little exposure to marketing, production, or other
business functions. Conversely, M.B.A.’s in sales
applications requires distributors to do much more
than simply resell products. They may install, upgrade, and service new products. They may train customers to use the new products. They may integrate
new products with existing systems. The term “valueadded reseller” is apt. Because of VARs’ importance,
managers in high-tech businesses may begin to feel
that they are customers. Moreover, buyers may believe that their most important contact is with the
VAR, not with the manufacturer. But that state of
affairs is a recipe for dilution of a powerful brand.
Managers of high-tech brands must understand that
the promise of value is made to the end user, not to the
reseller. The promise of value is inherently a “pull”
marketing concept, in which demand is driven more
by customers who pull the product through reseller
channels, than it is a “push” concept, in which the
company offers resellers high margins and other incentives to push the goods out to end users. Pull marketing is common in the consumer-packaged-goods
arena, where manufacturers advertise heavily to end
users. Cosmetics and detergents are among the most
intensively advertised consumer packaged goods, and
they enjoy considerable levels of customer loyalty.
Push marketing is traditionally more associated with
business-to-business products in that a higher proportion of the money spent on marketing is devoted to
direct sales forces and to distributors, which push the
product for the manufacturer to end users. We argue
that successful brands – in both consumer goods and
technology products – are more likely to result from
generating consumer desire –the desire to pull –as a result of a brand’s promise of value, than from attempting to push products through resellers, which are likely to be much more interested in high margins than in
reinforcing a manufacturer’s promise of value. Hightech managers must try to ensure that VARs understand the brand’s promise of value and precisely what
they must do to fulfill and reinforce that promise.
Companies should consider instituting performance
measurements to monitor their VARs’ success in delivering on the promise of value.
and marketing may have scant understanding of
their industry’s core technologies and sufficient,
but not extraordinary, knowledge of their company’s own technology products. Brand management concepts provide a common denominator
and a common language to bridge these different
thought worlds.
Externally, strong brand management helps breed
loyal customers – customers the company under-
94
harvard business review
141
July–August 1999
b r a n d s f o r h i g h - t e c h m a n ag e r s
stands and knows how to work with. Similarly,
brand management can improve the relationships a
company has with its suppliers and distributors. It’s
fashionable to refer to those relationships as “partnerships,” characterized by cooperation and communication, but some are really naked power struggles. (See the insert “What Lies Between Your Brand
and Your Customers?”) If a company’s promise of
value to customers is well understood internally,
then its managers can evaluate and select partners
based on their ability and willingness to support the
promise. What’s more, a well-articulated promise
forms a solid basis for clearly and consistently
defining the roles of, and allocating tasks between,
a company and its partners.
Finally, high-tech companies that manage their
brands effectively are well positioned to increase
their profits. In general, marketing high-tech products and services costs less than marketing consumer packaged goods. High-tech markets are often
more segmented, focused, and granular. For business-to-business customers, advertising is a less
important ingredient of the promotion mix than
more cost-effective ways of approaching customers
such as trade shows, the Internet, and user groups –
all powerful vehicles for reaching highly targeted
customers. Moreover, targeting customers means
that expenditures for product development and promotion can be tightly focused and efficiently allocated. On the revenue side, strong technology brands
harvard business review
can increase customers’ willingness to buy related
products and services and pay price premiums.
Launching new products may also be less costly for
powerful brands because of their loyal customer
base and the increased willingness of potential customers to deal with a well-known brand.
Our experience indicates that many managers in
high-tech businesses find brand management concepts interesting but are loathe to move from a
product-centric to a brand-centric business model.
The “it doesn’t apply here” argument is often based
on what are believed to be unique characteristics of
high-tech products and markets, especially the
volatility of the industry caused by swiftly changing technology and high levels of uncertainty among
buyers. We argue, however, that it is precisely those
volatile conditions that make the brand concept
especially pertinent. When things change quickly,
and when buyers face great uncertainty, they want
to deal with a company they perceive has a vision of
their needs and interests that goes beyond price and
performance. They want to deal with a company
they feel they can trust.
1. For more information on measuring brand equity, see the case note
“Brand Valuation Methodology: A Simple Example,” Harvard Business
School Publishing 596092.
2. “Mission Impossible: Winning the Computer Advertising Wars,” Upside, September 1996.
Reprint 99411
To place an order, call 1-800-988-0886.
95
July–August 1999
142
CASE STUDIES AND
HARVARD BUSINESS REVIEW
ARTICLE REPRINTS
Many readers have asked for an easy way
to order case studies and article reprints or to
obtain permission to copy. In response, we
have established a Customer Service Team
to grant permission, send rush copies in paper
form, deliver files in Acrobat (PDF) format
electronically (Harvard Business Review
articles only), or customize collections.
Please contact the Customer Service Team:
Phone: 617-496-1449
United States and Canada: 800-668-6780
(8 A.M. to 6 P.M. weekdays; voice mail
after hours)
Fax: 617-496-1029 (24 hours, 7 days a week)
E-mail: custserv@hbsp.harvard.edu
(24 hours, 7 days a week)
Web Site: http://www.hbsp.harvard.edu
Prices (minimum order, $10):
Harvard Business Review Reprints
(Discounts apply to multiple copies of the
same article.)
1–9 copies
10–49
50–79
80–99
100–499
Electronic
HARVARD BUSINESS REVIEW
SUBSCRIPTION SERVICE
United States and Canada
Phone: 800-274-3214
Rates per year: United States, $95;
Canada, U.S.$105
International and Mexico
Phone: 44-1858-435324
Fax: 44-1858-468969
Rates per year: international, U.S.$145;
Mexico, U.S.$105
Orders, inquiries, and address changes:
Harvard Business Review
Tower House, Sovereign Park
Lathkill Street, Market Harborough
Leicestershire le16 9ef
England
International customer service e-mail
address: harvard@subscription.co.uk
Payments accepted: Visa, MasterCard,
American Express; checks at current
exchange rate payable to
Harvard Business Review.
Bills and other receipts may be issued.
CATALOGS
$5.50 each
$5.00
$4.50
$4.00
$3.50
$5.50 each
Harvard Business School Publishing
Media Catalog
This 32-page, full-color catalog features more
than 40 management development video and
interactive CD-ROM programs.
Harvard Business School Case Studies
$5.50 each
For quantity estimates or quotes on
customized products, call
Frank Tamoshunas at 617-495-6198.
Fax: 617-496-8866
PERMISSIONS
For information on permission to quote
or translate Harvard Business School
Publishing material, contact:
Customer Service Department
Harvard Business School
Publishing Corporation
60 Harvard Way
Boston, MA 02163
Phone: 617-496-1449
United States and Canada: 800-668-6780
Fax: 617-495-6985
E-mail: custserv@hbsp.harvard.edu
143
Harvard Business School Press
This latest full-color catalog features books
for the fast-paced business world where you
live and work.
Harvard Business School Publishing
Catalog of Best-Selling Teaching Materials
This collection of teaching materials
contains those items most requested by
our customers.
Harvard Business School Publishing
Catalog of New Teaching Materials
Designed for individuals looking for the
latest materials in case method teaching.
Monday, October 3
Global Executive MBA in Shanghai
GEMBA VIII – Theme 3A
General Manager as Integrator
29 Sept – 3 Oct 2011, Shanghai Jiao Tong University, Shanghai
Professors Ty Callahan (Finance and Business Economics) and Joe Nunes (Marketing)
All assigned readings and cases for all the week's sessions need to be read and studied
before the first session at SJTU. Focus on reading and being prepared to discuss the
cases!
Monday, 3 October
08:30
Marketing Session 11 – Professor Nunes
Topic: Mattel Case
Required Readings:
-
Mattel and the Toy Recalls (B) Ivey 908M11
-
Donaldson, Thomas “Values in Tension: Ethics Away from Home” HBR
Reprint 96502
Discussion Questions:
1. What went wrong with Mattel’s recall strategy?
2. Who are Mattel’s stakeholders? Who did Mattel cater to in the recall?
3. What values did Mattel exhibit during the recall? How did they affect
Mattel?
4. What should Mattel do right now (at the point in the case) and in the
future?
09:45
Break
10:00
Integrative Session with Professors Nunes and Callahan
Topic: TBD
11:15
Break
11:30
Finance Session 10 – Professor Callahan
Topic: Returning Value to Shareholders (aka Payout Policy)
In recent years, about 40% of firms’ income has been paid out to
shareholders via dividends. But it is also true that many firms pay no
dividends at all. What do we know about dividend policy? This session
will explore various methods of returning value to shareholders including
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
144
Page: 1 / 2
Global Executive MBA in Shanghai
cash dividends, stock repurchases, and extraordinary dividends. Topics
will include dividend policy and personal taxes, factors favoring a high
dividend policy, and some empirical regularities about US corporate
payout policy.
Required Readings: RWJ Chapter 19
Discussion Questions:
1. What are the typical methods used by firms to return cash to
shareholders?
2. What role do taxes play in corporate payout policy?
3. What other factors influence corporate payout policy?
To do before this session:
-
Review readings
-
Prepare answers to the discussion questions
12:45
Lunch
13:45
Finance Session 11 – Professor Callahan
Topic: Returning Value to Shareholders – Applied
In this session we will use the General Motors Corporation (D) case to
review payout policy theory and look at various tradeoffs inherent in
choosing a policy. We examine the interaction between payout policy,
investment policy, and capital structure policy.
Required Reading: The General Motors Corporation (D) case
Discussion Questions:
1. What options are Mr. Finnegan and his staff evaluating?
2. What are the pros and cons of each option?
3. What do you recommend General Motors to do?
To do before this session: Prepare the case for discussion
15:00
Break
15:15
Summary and Evaluations – Professors Callahan and Nunes
GEMBA VIII, Theme 3A, 29 Sept to 3 Oct, 2011, daily syllabus
145
Page: 2 / 2
S
w
908M11
MATTEL AND THE TOY RECALLS (B)1
Professors Hari Bapuji and Paul Beamish wrote this case solely to provide material for class discussion. The authors do not intend
to illustrate either effective or ineffective handling of a managerial situation. The authors may have disguised certain names and
other identifying information to protect confidentiality.
Ivey Management Services prohibits any form of reproduction, storage or transmittal without its written permission. Reproduction of
this material is not covered under authorization by any reproduction rights organization. To order copies or request permission to
reproduce materials, contact Ivey Publishing, Ivey Management Services, c/o Richard Ivey School of Business, The University of
Western Ontario, London, Ontario, Canada, N6A 3K7; phone (519) 661-3208; fax (519) 661-3882; e-mail cases@ivey.uwo.ca.
Copyright © 2008, Ivey Management Services
Version: (A) 2008-02-21
China is now issuing two types of toys: leaded and un-leaded.
– Jay Leno, U.S. Talk Show Host
On August 14, 2007, the U.S. Consumer Product Safety Commission (CPSC) in cooperation with Mattel
announced five different recalls of Mattel’s toys. See Exhibit 1 for excerpts of the CPSC recall notices. On
the same day, Mattel issued a press release (see Exhibit 2) and held a press conference. Bob Eckert, CEO
of Mattel, made a five minute briefing and answered the questions posed by the reporters. Following are
some excerpts from Eckert’s address.2
As you know today Mattel announced in cooperation with the U.S. Consumer Product
Safety Commission voluntary recalls on two issues, a product recalled for impermissible
levels of lead and an expansion of the November 2006 magnet recall. We’ve already put
measures in place to address these issues, and I will talk about those in a moment.
Obviously we don’t wanna have recalls, but in acting responsibly we won’t hesitate to take
action to correct issues to assure the safety of our products and the safety of children. I
want to underscore that Mattel has extremely rigorous testing and quality procedures in
place and we will continue to be vigilant in enforcing quality and safety. First Mattel has
voluntarily recalled one toy from the Cars die cast vehicle line, manufactured between
May 2007 and July 2007 containing impermissible levels of lead. The recall of the ‘Sarge
Toy’ results from Mattel’s on going testing procedures. The Cars toy was produced by
Early Light Industrial Company, one of Mattel’s contract manufacturing facilities in
China, which subcontracted the painting of parts of the toy to another vendor named Hong
Li Da, also in China. While the painting subcontractor was required to utilize certified
paint supplied directly from Early Light, he had instead violated Mattel standards and
utilized paint from a non-authorized third party supplier. To address this issue we have
1
This case has been written on the basis of published sources. Consequently, the interpretations and perspectives
presented are not necessarily those of Mattel and other organizations represented in this case or any of their employees.
2
Transcript of Eckert’s address to reporters.
146
Page 2
9B08M011
immediately implemented a strengthened three point check system…. (details of the
system)
Additionally Mattel is voluntarily recalling certain toys with magnets manufactured
between January 2002 and January 31st, 2007 that may release small powerful magnets.
The recall expands upon Mattel’s voluntary recall of 8 toys in November 2006 and is
based on a thorough internal review of all of our brands that have toys with magnets and
analyzed the ways in which magnets may come loose. Since January, 2007 all magnets
used in our toys have been locked into the toy with sturdy material holding in the edges
around the exposed face of the magnet or completely covering the magnet. We now
believe it is prudent to recall our older toys with magnets that do not meet our latest
retention system requirements. This means we are recalling 72 toys that were distributed
in prior years. The safety of children is our main concern and we’re confident that our new
requirements work based on our continued testing and consumer experience. The risk of
magnets are swallowed is serious and we believe that all our toys with magnets should
have the safety benefit of our new standards.
The news of the recall spread like wildfire all over the world. The media coverage it received was
unprecedented, with TV channels running the story through the day. Several analysts pointed to the
previous recalls of Chinese-made goods and demanded that the U.S. and Chinese governments must act.
The recall of Mattel toys was quickly followed by several other recalls of Chinese-made goods. About 40
different products, most made in China, were recalled for excess lead. Mattel announced three more recalls
on September 4 in which an additional 773,900 toys made in China were recalled for excess lead.
The spate of recalls severely eroded consumer confidence. In a poll conducted by Reuters/Zogby,3 the
majority of people (close to 80 per cent) reported that they were apprehensive about buying goods made in
China. Nearly two-thirds (63 per cent) of the respondents reported that they were likely to participate in a
boycott of Chinese goods until the Chinese government improved the regulations governing the safety of
the goods exported to the United States. Several other opinion polls conducted by news agencies and
market research firms revealed similar sentiments.
The governments in the West quickly responded to the crisis of confidence. At a summit of North
American political leaders in Canada, the heads of governments of Canada, the United States, and Mexico
decided to crack down on unsafe goods, particularly those designed for children. Using the Mattel recalls
case, EU Consumer Commissioner Meglena Kuneva, initiated an extensive review of the strengths and
weaknesses of the consumer product safety mechanisms in Europe. The review had involved extensive
work with national surveillance authorities, the Chinese authorities, the U.S. authorities, the European toy
industry, retailers, as well as consultations with the European Parliament. The government of Brazil
decided to halt the import of toys by Mattel until the lead issue was resolved.
The U.S. Senate as well as the House of Commons held hearings on the safety of imported products and
Bob Eckert was summoned to testify in both the hearings. In those hearings, Eckert asserted:4 “a few
vendors, either deliberately or out of carelessness, circumvented our long-established safety standards and
procedures.”
3
http://www.signonsandiego.com/news/nation/20070919-0400-usa-foodsafety-poll.html
Testimony of Robert Eckert, CEO, Mattel, to the Sub-committee on Commerce, Trade, and Consumer Protection of the
Committee on Energy and Commerce. September 19, 2007.
4
147
Page 3
9B08M011
The recalls catapulted consumer product safety to the center of debate. Questions were raised about
whether the CPSC had enough resources to ensure product safety. Consumer advocates and some
politicians pointed to the steady budget and staffing cuts the CPSC faced; with 420 employees in 2007, the
CPSC was half its size of the 1980s. Many wondered if that number was adequate to monitor 15,000
consumer products in a market valued at US$614 billion. It was pointed out that CPSC had only one
employee devoted to testing the safety of toys. Also, only 15 CPSC inspectors were available to check all
the U.S. ports where import shipments were received.5
Given the limited resources of the CPSC, it was easy for unscrupulous companies to “play truant.” And, the
CPSC “lacked the teeth” to check it. For example, the maximum penalty the CPSC could impose on
companies for violations was $1.8 million. Imposing penalties was never easy because the burden of proof
rested with the CPSC. Additionally, the CPSC could not even make public its concerns or investigations
about companies. It was required by law to take prior approval of the companies whose names were being
divulged.6
The role of Mattel in ensuring product safety was also under scrutiny. Some observers pointed that Mattel
had not informed the CPSC within the stipulated time. In the past, Mattel was fined twice by the CPSC for
not informing the latter about product hazards in a timely manner. If the alleged delay by managers was
indeed true, then it was possible that shareholders might sue the directors and senior executives of the
company for the delays and exposing the company to risk.
In an interview with the Wall Street Journal, Eckert felt that the CPSC requirement of immediately
reporting the incidents was unreasonable and that Mattel had the freedom to investigate the incidents
before providing the information to the CPSC. When asked by the media about the time Mattel took to
recall, Eckert said that the company asked the CPSC to initiate a fast-track recall and they acted as fast as
they could. Some observers criticized Mattel for being unapologetic about the recalls. Efforts by several
reporters to reach Mattel after August 14 were in vain.
Mattel’s customers were livid about the recalls and wondered if the company had any control and
monitoring systems at all because they were recalling toys sold over the previous three years. Some
wondered if Mattel had any quality systems to test the manufactured toys for safety. As the recalls were
announced, parents found it difficult to empty the toy baskets of their children without “breaking their
hearts.” Some families filed class-action lawsuits, asking Mattel to pay for tests to determine if children
were exposed to lead. More lawsuits were expected to follow.
The recalls made the licensors of brands to Mattel, such as Disney and Sesame Workshop, very nervous.
They feared the erosion of their brand value because of the lead paint issue. Disney announced independent
testing of the toys it made with Disney brand names. One of the largest toy retailers, Toys’R’Us also began
to conduct its own lead testing of toys on its shelves.7
The continued attention to the issue of recall and particularly Mattel’s role began to affect the image of all
toys sold in the United States. The toys made by every company were being scrutinized and consumers
were looking more carefully at the toys to find out where they were made. Many consumers rushed in
search of toys made in the United States or other developed countries, but they were hard to find. Not to be
discouraged, some enthusiasts set up websites to inform shoppers about where to buy American toys
5
Stephen Labaton. Bigger budget? No, responds safety agency. New York Times, Oct. 30, 2007.
Felcher M. 2001. It's No Accident: How Corporations Sell Dangerous Baby Products. Common Courage Press: Monroe,
ME
7
http://www.reuters.com/article/domesticNews/idUSN1040588720070910
6
148
Page 4
9B08M011
(www.howtobuyamerican.com) and others set up businesses that sold toys not made in China, aptly named
NMC Toys (Not Made in China Toys www.nmctoys.com). A few companies, such as Little Tykes, which
manufactured some of their toys in the United States began to prominently display Made in USA labels on
their toys.
Some analysts argued that the suppliers in China and elsewhere were compromising on safety to meet the
ever increasing pressure of the Western toy companies to supply toys and other products at a cheaper cost,
even in the face of increasing raw material and wage costs. This resulted in a double-squeeze for the toy
suppliers. Some consumer advocates asserted that companies like Mattel which brought the toys into the
United States had the primary responsibility for ensuring the safety of products — that no matter where in
the global supply chain, the problem might have occurred.
The suppliers in China faced pressures from large toy companies, who in turn faced pressures from large
retailers, to cut down costs. Additionally, the economic growth enjoyed by China resulted in rising wages,
and a general increase in the cost of doing business. Within China, toy-making is clustered in Guangdong
province. The Nominal Wage Rate Index (NWRI) in Guangdong increased to 545 in 2003 from a base of
100 in 1991. The increase was much higher than the national average of 450 in 2003, and was fifth largest
within China. The average Consumer Price Index (CPI) in China rose from 100 in 1992 to 202 in 2004.
The rise in CPI was less stark for Guangdong province, reaching 189.8
The suppliers in China faced another problem: the rising value of the yuan when compared to other Asian
currencies. For example, since 1997 the Chinese yuan has appreciated nearly four fold against the
Indonesian rupiah, doubled in value against the Philippine peso, and increased in value by at least 1.5 times
against the South Korean won, the Malaysian ringgit and the Thai baht. As a result, these destinations were
becoming increasingly attractive for manufacturing and the advantage of operating in China was eroding.9
As a result of the increased wages, cost of living and the value of the yuan, the pattern of economic activity
in China underwent a rapid change. Industrial activity had shifted to higher-value industries which could
absorb the rising costs. Exhibit 3 presents the changes in industry concentration by region in China. From
being dominant in only one region in 1990, electronic equipment was the most dominant industrial activity
in four regions in 2006.10 The shift in industrial activity was best captured in the words of the Mattel CEO:
Wage rates are going up in southern China, and it’s harder for us to find employees in
southern China. You know, next to a toy factory 20 years ago, there was empty land.
Today next to every toy factory, I think you can look to your left and see a cell phone
plant or some sort of electronics plant. You might be able to look to your right and see an
auto manufacturer.
The effect of recalls began to take a toll on the already besieged toy suppliers in China. On August 11,
2007, Cheng Shu-hung, who directly managed the operations of Lee Der, committed suicide. He was 48,
single, and lived in a 250 square foot room in one of Lee Der’s offices. He was considered to be kind to the
factory workers and was credited with the better working conditions that prevailed in the three factories of
Lee Der. Shop floor salaries for a 10-hour, six-day week in Lee Der factories ranged between US$120 and
180 a month, higher than the local average of $130 a month for seven-day week schedules that often ran
8
Delios A, Beamish P, Zhao X. 2008. The evolution of Japanese investment in China: From toys to textiles to business
process outsourcing. Asia Pacific Business Review (forthcoming).
9
Ibid.
10
Ibid.
149
Page 5
9B08M011
14-hours a day. Also, employees received overtime pay when the shift exceeded 10 hours. One of the last
things Cheng Shu-hung did was to sell his factories and pay wages to his employees.11
Following the recalls, Chinese employees in Lee Der and other factories became jobless. The conditions of
workers became an issue of discussion. Some observers wondered what the effect of lead was on the
employees who painted lead on the toys, and thus ingested it, every day of the week.
The recalls began to severely erode “Brand China” and the Chinese government quickly set up a taskforce
under the leadership of Chinese Vice Premier Wu Yi to ensure product safety. This taskforce intensified
the inspection of Chinese plants and suspended or revoked the export licenses of hundreds of companies.
Some suppliers named in the recalls were jailed.
Faced with intense pressure from all quarters, the Chinese authorities asserted that the majority of products
made in China were safe and that Western companies were unduly blaming China. Several suppliers who
worked with big companies and were forced to close factories or lay off workers asserted that Mattel and
other large companies were making them scapegoats.
In what appeared to be a counter-offensive, China rejected North American imports such as frozen pig
kidneys imported from the United States and frozen pork spareribs from Canada. These products were
found to contain residues of ractopamine, forbidden for use as veterinary medicine in China.12 Also, China
rejected shipments of U.S.-made orange pulp and dried apricots containing high levels of bacteria and
preservatives.13
In an effort aimed at enhancing product safety, the CPSC and its Chinese-counterpart AQSIQ met in
Washington on September 11-12, 2007. This meeting culminated in agreement to ban the use of lead in
toys made in China. At the meeting, in his address, the AQSIQ chief asserted that the West was blaming
China for the problems created by its toy companies. In support of his assertion, he mentioned a recent
Canadian study which found that the majority of toy recalls in the U.S. were due to design flaws.
According to a report in the New York Times on September 12, 2007, two Canadian business school
researchers, after analyzing the toy recalls in the United States over the previous 20 years, found that 76 per
cent of the recalls were due to design flaws such as sharp edges, easily detachable small parts, and long
strings. In contrast, only 10 per cent were due to manufacturing flaws such as using poor material, incorrect
assembly, and use of unacceptable material like lead paint. The researchers argued that China should not be
blamed for most of the recalls, when a vast majority of the problems were because of the designs made in
the corporate headquarters of toy companies.
Mattel had considerable interests in China. Five of its factories were located in China and a very large
number of factories made toys for Mattel, directly or indirectly. The Chinese news agencies began to report
that Chinese suppliers were being made a scapegoat by Mattel, despite the fact that 90 per cent of the toys
recalled on August 14 were due to magnets detaching, which was a design problem for which Mattel was
responsible. The loss of reputation for China as a result of the recalls was huge and Mattel seemed like the
floodgate that had opened it.
11
http://www.ckgsb.edu.cn:8080/article/600/3051.aspx
http://www.cbc.ca/consumer/story/2007/09/17/china-exports.html
13
http://www.cbc.ca/consumer/story/2007/06/26/china-trade.html
12
150
Page 6
9B08M011
Exhibit 1
CPSC RECALL NOTICES
FOR IMMEDIATE RELEASE
August 14, 2007
Release #07-273
Firm's Recall Hotline: (888) 597-6597
CPSC Recall Hotline: (800) 638-2772
CPSC Media Contact: (301) 504-7908
Additional Reports of Magnets Detaching from Polly Pocket Play Sets Prompts Expanded Recall by Mattel
WASHINGTON, D.C. - The U.S. Consumer Product Safety Commission, in cooperation with the firm named below,
today announced a voluntary recall of the following consumer product. Consumers should stop using recalled
products immediately unless otherwise instructed.
Name of Products: Various Polly Pocket dolls and accessories with magnets
Units: About 7.3 million play sets (about 2.4 million play sets were recalled on November 21, 2006)
Importer: Mattel Inc., of El Segundo, Calif.
Hazard: Small magnets inside the dolls and accessories can come loose. The magnets can be found by young
children and swallowed or aspirated. If more than one magnet is swallowed, the magnets can attract each other and
cause intestinal perforation or blockage, which can be fatal.
Incidents/Injuries: Since the previous recall announcement, Mattel has received more than 400 additional reports of
magnets coming loose. CPSC was aware in the first recall announcement of 170 reports of the magnets coming out
of the recalled toys. There had been three reports of serious injuries to children who swallowed more than one
magnet. All three suffered intestinal perforations that required surgery.
Description: The recalled Polly Pocket play sets contain plastic dolls and accessories that have small magnets. The
magnets measure 1/8 inch in diameter and are embedded in the hands and feet of some dolls, and in the plastic
clothing, hairpieces and other accessories to help the pieces attach to the doll or to the doll’s house. The model
number is printed on the bottom of the largest pieces on some of the play sets. Products manufactured after
November 1, 2006 and are currently on store shelves are not included in this recall. Contact Mattel if you cannot find
a model number on your product to determine if it is part of the recall.
Sold at: Toy stores and various other retailers from May 2003 through November 2006 for between $15 and $30.
Manufactured in: China
Remedy: Consumers should immediately take these recalled toys away from children and contact Mattel to receive a
voucher for a replacement toy of the customer’s choice, up to the value of the returned product.
~~~~~~~~~~~~~~
Mattel Recalls Doggie Day Care™ Magnetic Toys Due to Magnets Coming Loose
Name of Product: Doggie Day Care™ play sets
Units: About 1 million
Importer: Mattel Inc., of El Segundo, Calif.
Hazard: Small magnets inside the toys can fall out. Magnets found by young children can be swallowed or aspirated.
If more than one magnet is swallowed, the magnets can attract each other and cause intestinal perforation or
blockage, which can be fatal.
Incidents/Injuries: The firm has received two reports of magnets coming loose. No injuries have been reported.
Description: The recalled Doggie Day Care play sets have various figures and accessories that contain small
magnets.
Sold at: Toy stores and various other retailers nationwide from July 2004 to August 2007 for between $4 and $20.
Manufactured in: China
Remedy: Consumers should immediately take the recalled toys away from children and contact Mattel to receive a
free replacement toy.
~~~~~~~~~~~~~~~~~~~~~
Mattel Recalls Barbie and Tanner™ Magnetic Toys Due to Magnets Coming Loose
Name of Product: Barbie and Tanner™ play sets
Units: About 683,000
Importer: Mattel Inc., of El Segundo, Calif.
Hazard: A small magnet inside the “scooper” accessory can come loose. Magnets found by young children can be
swallowed or aspirated. If more than one magnet is swallowed, the magnets can attract each other and cause
intestinal perforation or blockage, which can be fatal.
Incidents/Injuries: The firm has received three reports of magnets coming loose. No injuries have been reported.
151
Page 7
9B08M011
Exhibit 1 (continued)
Description: The recall involves Barbie and Tanner™ play sets -- model numbers J9472 and J9560. The toys
include a “scooper” accessory with a magnetic end. Recalled scoopers have a visible, silver-colored, disc-shaped
magnet on the end of the scooper. Scoopers with a white material covering the magnet and products manufactured
after January 31, 2007 are not recalled.
Sold at: Toy stores and various other retailers nationwide May 2006 to August 2007 for about $16.
Manufactured in: China
Remedy: Consumers should immediately take the recalled toys away from children and contact Mattel to receive a
free replacement toy.
~~~~~~~~~~~~~~~~~~~~~
Mattel Recalls “Sarge” Die Cast Toy Cars Due To Violation of Lead Safety Standard
Name of Product: “Sarge” die cast toy cars
Units: About 253,000
Importer: Mattel Inc., of El Segundo, Calif.
Hazard: Surface paints on the toys could contain lead levels in excess of federal standards. Lead is toxic if ingested
by young children and can cause adverse health effects.
Incidents/Injuries: None reported.
Description: The recall involves die cast “Sarge” 2 ½ inch toy cars. The toy looks like a military jeep and measures
about 2 ½ inches long by 1 inch high by 1 inch wide. The recalled toy has the markings “7EA” and “China” on the
bottom. The “Sarge” toy car is sold alone or in a package of two, and may have the product number M1253 (for single
cars) and K5925 (for cars sold as a set) printed on the packaging. The cars marked “Thailand” are not included in this
recall.
Sold at: Retail stores nationwide from May 2007 through August 2007 for between $7 and $20 (depending on
whether they were sold individually or in sets).
Manufactured in: China
Remedy: Consumers should immediately take the recalled toys away from children and contact Mattel. Consumers
will need to return the product to receive a replacement toy.
~~~~~~~~~~~~~~~~~~~~~~
Mattel Recalls Batman™ and One Piece™ Magnetic Action Figure Sets Due To Magnets Coming Loose
Name of Product: Batman™ and One Piece™ magnetic action figure sets
Units: About 345,000
Importer: Mattel Inc., of El Segundo, Calif.
Hazard: Small, powerful magnets inside the accessories of the toy figures can fall out and be swallowed or aspirated
by young children. If more than one magnet is swallowed, they can attract inside the body and cause intestinal
perforation, infection or blockage which can be fatal.
Incidents/Injuries: The firm is aware of 21 incidents where a magnet fell out of the toy figure, including a case of a 3year-old boy who was found with a magnet in his mouth. The boy did not swallow the magnet and no injuries have
been reported to Mattel and CPSC.
Description: The recalled Batman™ toys include:
• The Batman™ Magna Battle Armor™ Batman™ figure with model number J1944,
• The Batman™ Magna Fight Wing™ Batman™ figure with model number J1946,
• The Batman™ Secret ID™ figure with model number J5114, and
• The Batman™ Flying Fox™ figure with model number J5115. The seven inch tall action figures include the
Batman logo on the front and include magnetic accessories. The model number is located on the lower right
corner of the tag which is sewn to the figure.
The recalled One Piece™ toy is:
• One Piece™ Triple Slash Zolo Roronoa™ figure with model number J4142. The 5 ½ inch tall action figure
has green hair, black pants, and has magnets in his hands which connect to magnets on various swords that
the figure can hold. The model number is printed on the back of the action figure’s left leg.
Sold at: Discount department stores and toy stores nationwide from June 2006 through June 2007 for about $11.
Manufactured in: China
Remedy: Consumers should immediately stop using the toy and contact Mattel for instructions on how to return it to
receive a free replacement toy.
152
Page 8
9B08M011
Exhibit 2
MATTEL PRESS RELEASE - MATTEL ANNOUNCES EXPANDED RECALL OF TOYS
One product recalled for impermissible levels of lead
November 2006 magnet recall expanded
EL SEGUNDO, Calif., August 14, 2007 – Mattel, Inc. announced today that the company has voluntarily recalled one toy
from the “CARS” die-cast vehicle line (“Sarge” character), manufactured between May 2007 and July 2007, containing
impermissible levels of lead. The recalled vehicles include 436,000 total toys, including 253,000 in the U.S. and 183,000
outside of the U.S.
The recall of the Sarge toy results from Mattel’s increased investigation and ongoing testing procedures following the recall
of select Fisher-Price toys on August 1, 2007. The toy was produced by Early Light Industrial Co., Ltd (Early Light), one of
Mattel’s contract manufacturing facilities in China, which subcontracted the painting of parts of the toy to another vendor,
Hong Li Da (HLD), also in China. While the painting subcontractor, HLD, was required to utilize paint supplied directly from
Early Light, it instead violated Mattel’s standards and utilized paint from a non-authorized third-party supplier.
“We have immediately implemented a strengthened three-point check system: First, we’re requiring that only paint from
certified suppliers be used and requiring every single batch of paint at every single vendor to be tested. If it doesn’t pass, it
doesn’t get used. Second, we are tightening controls throughout the production process at vendor facilities and increasing
unannounced random inspections. Third, we’re testing every production run of finished toys to ensure compliance before
they reach our customers. We’ve met with vendors to ensure they understand our tightened procedures and our absolute
requirement of strict adherence to them,” said Jim Walter, senior vice president of Worldwide Quality Assurance, Mattel.
Additionally, Mattel announced the voluntary recall of magnetic toys manufactured between January 2002 and January 31,
2007, including certain dolls, figures, play sets and accessories that may release small, powerful magnets. The recall
expands upon Mattel’s voluntary recall of eight toys in November 2006 and is based on a thorough internal review of all
Mattel’s brands. Mattel is recalling 18.2 million magnetic toys globally (9.5 million in the U.S.); however, the majority of the
toys are no longer at retail. Beginning in January 2007, Mattel implemented enhanced magnet retention systems in its toys
across all brands.
“Since our November 2006 magnet-related recall, we have implemented more robust magnet retention systems and more
rigorous testing. We are exercising caution and have expanded the list of recalled magnetic toys due to potential safety risks
associated with toys that might have loose magnets,” said Walter.
“The safety of children is our primary concern, and we are deeply apologetic to everyone affected,” said Robert A. Eckert,
chairman and chief executive officer, Mattel. “Mattel has rigorous procedures, and we will continue to be vigilant and
unforgiving in enforcing quality and safety. We don’t want to have recalls, but we don’t hesitate to take quick and effective
action to correct issues as soon as we’ve identified them to ensure the safety of our products and the safety of children.”
Issues Safety Alert to Consumers
Mattel is working in cooperation with the U.S. Consumer Product Safety Commission and other regulatory agencies
worldwide. Mattel is also working with retailers worldwide to identify and remove affected products from retail shelves.
Details of the recall are as follows:
Mattel voluntarily recalled 63 magnetic toys sold at retail prior to January 2007. Magnetic toys recalled within the U.S.
include 44 Polly Pocket™ toys, 11 Doggie Day Care® toys, 4 Batman™ toys, 1 One Piece™ toy, and the accessory part of
2 Barbie® toys. For additional information regarding the magnetic toy recall, contact Mattel at (888) 597-6597, or visit the
company’s Web site at www.service.mattel.com.
The Sarge toy from the “CARS” die-cast vehicle line was manufactured between May 2007 and August 2007. For additional
information regarding the Sarge toy recall, contact Mattel at (800) 916-4997, or visit the company’s Web site at
www.service.mattel.com.
A full list of products is published on the company’s Web site at www.mattel.com, as well as by the Consumer Products
Safety Commission. Consumers should immediately take these products away from children and contact Mattel to arrange
return and to receive a voucher for a replacement toy of the consumer’s choice, up to the value of the returned product.
Source: Mattel Website
153
Page 9
9B08M011
Exhibit 3
INDUSTRY CONCENTRATION BY REGION IN CHINA
(1990 - 2006)
Region
Industry (1990)
Conc.
Industry (2000)
Conc.
Industry (2006)
Conc.
North
Industrial Machinery
16%
Industrial Machinery
17%
Electronic Equipment
17%
Northeast
Apparel
20%
Industrial Machinery
17%
Electronic Equipment
19%
East
Apparel
27%
Apparel
16%
Electronic Equipment
19%
Mid-South
Electronic Equipment
26%
Electronic Equipment
28%
Electronic Equipment
30%
Southwest
Food Products
50%
Transportation Equipment
49%
Transportation Equipment
47%
Northwest
Food Products
25%
Industrial Machinery
41%
Industrial Machinery
32%
Note: North = Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia
Northeast = Liaoning, Jilin, Heilongjiang
East = Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Shandong
Mid-South = Henan, Hubei, Hunan, Guangdong, Guangxi, Hainan
Southwest = Chongqing, Sichuan, Guizhou, Yunnan, Tibet
Northwest = Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang
Source: Delios A, Beamish P, Zhao X. 2008. The evolution of Japanese investment in China: From toys to textiles to
business process outsourcing. Asia Pacific Business Review (forthcoming).
154
Values in Tension:
Ethics Away from Home
by Thomas Donaldson
Harvard Business Review
Reprint 96502
155
Harvard Business Review
HBR Subscriptions
Harvard Business Review
U.S. and Canada
Subscription Service
P.O. Box 52623
Boulder, CO 80322-2623
Telephone: (800) 274-3214
Fax: (617) 496-8145
Outside U.S. and Canada
Tower House
Sovereign Park
Lathkill Street
Market Harborough
Leicestershire LE16 9EF
Telephone: 44-85-846-8888
Fax: 44-85-843-4958
American Express, MasterCard, VISA accepted. Billing available.
HBR Article Reprints
HBR Index and Other Catalogs
HBS Cases
HBS Press Books
Harvard Business School Publishing
Customer Service - Box 230-5
60 Harvard Way
Boston, MA 02163
Telephone: U.S. and Canada (800) 545-7685
Outside U.S. and Canada: (617) 495-6117 or 495-6192
Fax: (617) 495-6985
Internet address: custserv@cchbspub.harvard.edu
HBS Management Productions
Videos
Harvard Business School Management Productions videos are
produced by award winning documentary filmmakers. You’ll find
them lively, engaging, and informative.
HBR Custom Reprints
Please inquire about HBR’s custom service and quantity discounts.
We will print your company’s logo on the cover of reprints, collections, or books in black and white, two color, or four color. The process is easy, cost effective, and quick.
Telephone: (617) 495-6198 or Fax: (617) 496-8866
Permissions
For permission to quote or reprint on a one-time basis:
Telephone: (800) 545-7685 or Fax: (617) 495-6985
For permission to re-publish please write or call:
Permissions Editor
Harvard Business School Publishing
Box 230-5
60 Harvard Way
Boston, MA 02163
Telephone: (617) 495-6849
156
HarvardBusinessReview
SEPTEMBER-OCTOBER 1996
Reprint Number
JAMES C. COLLINS AND JERRY I. PORRAS
BUILDING YOUR COMPANY’S VISION
96501
DAVID A. THOMAS
AND ROBIN J. ELY
MAKING DIFFERENCES MATTER: A NEW PARADIGM
FOR MANAGING DIVERSITY
96510
ANN MAJCHRZAK
AND QIANWEI WANG
BREAKING THE FUNCTIONAL MIND-SET IN
PROCESS ORGANIZATIONS
96505
N. CRAIG SMITH, ROBERT J. THOMAS,
AND JOHN A. QUELCH
A STRATEGIC APPROACH TO MANAGING PRODUCT RECALLS
96506
RICHARD B. FREEMAN
TOWARD AN APARTHEID ECONOMY?
96503
WITH COMMENTARIES BY: ROBERT B. REICH, JOSH S. WESTON,
JOHN SWEENEY, WILLIAM J. MCDONOUGH, AND JOHN MUELLER
GEORGE STALK, JR., DAVID K. PECAUT,
AND BENJAMIN BURNETT
BREAKING COMPROMISES, BREAKAWAY GROWTH
96507
JOHN STRAHINICH
HBR CASE STUDY
THE PITFALLS OF PARENTING MATURE COMPANIES
96508
BARBARA E. TAYLOR, RICHARD P. CHAIT,
AND THOMAS P. HOLLAND
SOCIAL ENTERPRISE
THE NEW WORK OF THE NONPROFIT BOARD
96509
THOMAS DONALDSON
WORLD VIEW
VALUES IN TENSION: ETHICS AWAY FROM HOME
96502
JAMES P. WOMACK
AND DANIEL T. JONES
IDEAS AT WORK
BEYOND TOYOTA: HOW TO ROOT OUT WASTE AND
PURSUE PERFECTION
MARC LEVINSON
BOOKS IN REVIEW
CAPITALISM WITH A SAFETY NET?
157
96511
96504
W O R L D
V I E W
When is different just different, and
when is different wrong?
with people from the home country,
whose standards should prevail?
Even the best-informed, bestintentioned executives must rethink their assumptions about business practice in
foreign settings.
What works in a
company’s home
country can fail in
a country with different standards of
ethical conduct. Such difficulties are
unavoidable for businesspeople who
live and work abroad.
But how can managers resolve the
problems? What are the principles
that can help them work through
the maze of cultural differences and
establish codes of conduct for globally ethical business practice? How
can companies answer the toughest
question in global business ethics:
What happens when a host country’s
ethical standards seem lower than
the home country’s?
ple in Denmark or Singapore who
refuse to offer or accept bribes. Likewise, if Belgians fail to find insider
trading morally repugnant, who
cares? Not enforcing insider-trading
laws is no more or less ethical than
enforcing such laws.
The cultural relativist’s creed –
When in Rome, do as the Romans
do – is tempting, especially when
failing to do as the locals do means
forfeiting business opportunities.
The inadequacy of cultural relativism, however, becomes apparent
when the practices in question are
more damaging than petty bribery or
insider trading.
In the late 1980s, some European
tanneries and pharmaceutical companies were looking for cheap wastedumping sites. They approached virtually every country on Africa’s west
coast from Morocco to the Congo.
Values in Tension:
by Thomas Donaldson
When we leave home and cross
our nation’s boundaries, moral clarity often blurs. Without a backdrop
of shared attitudes, and without
familiar laws and judicial procedures
that define standards of ethical conduct, certainty is elusive. Should a
company invest in a foreign country
where civil and political rights are
violated? Should a company go along
with a host country’s discriminatory
employment practices? If companies
in developed countries shift facilities to developing nations that lack
strict environmental and health regulations, or if those companies
choose to fill management and other
top-level positions in a host nation
Competing Answers
One answer is as old as philosophical discourse. According to cultural
relativism, no culture’s ethics are
better than any other’s; therefore
there are no international rights and
wrongs. If the people of Indonesia
tolerate the bribery of their public
officials, so what? Their attitude is
no better or worse than that of peo-
Copyright © 1996 by the President and Fellows of Harvard College. All rights reserved.
158
Nigeria agreed to take highly toxic
polychlorinated biphenyls. Unprotected local workers, wearing thongs
and shorts, unloaded barrels of PCBs
and placed them near a residential
area. Neither the residents nor the
workers knew that the barrels contained toxic waste.
We may denounce governments
that permit such abuses, but many
countries are unable to police transnational corporations adequately
even if they want to. And in many
countries, the combination of ineffective enforcement and inadequate
regulations leads to behavior by unscrupulous companies that is clearly
wrong. A few years ago, for example,
a group of investors became interested in restoring the SS United
States, once a luxurious ocean liner.
Before the actual restoration could
begin, the ship had to be stripped of
its asbestos lining. A bid from a U.S.
company, based on U.S. standards
for asbestos removal, priced the job
DRAWINGS BY MICHAEL REAGAN
W O R L D
at more than $100 million. A company in the Ukranian city of Sevastopol offered to do the work for less
than $2 million. In October 1993,
the ship was towed to Sevastopol.
A cultural relativist would have
no problem with that outcome, but I
do. A country has the right to establish its own health and safety regulations, but in the case described
above, the standards and the terms
of the contract could not possibly
have protected workers in Sevastopol from known health risks. Even
if the contract met Ukranian standards, ethical businesspeople must
object. Cultural relativism is morally blind. There are fundamental values that cross cultures, and companies must uphold them. (For an
economic argument against cultural
relativism, see the insert “The Culture and Ethics of Software Piracy.”)
V I E W
they had used with U.S. managers:
the participants were asked to discuss a case in which a manager
makes sexually explicit remarks to
a new female employee over drinks
in a bar. The instructors failed to
consider how the exercise would
work in a culture with strict conventions governing relationships between men and women. As a result,
the training sessions were ludicrous.
They baffled and offended the Saudi
participants, and the message to
avoid coercion and sexual discrimination was lost.
The theory behind ethical imperialism is absolutism, which is based
on three problematic principles. Absolutists believe that there is a single
list of truths, that they can be expressed only with one set of concepts, and that they call for exactly
the same behavior around the world.
Ethics Away from Home
At the other end of the spectrum
from cultural relativism is ethical
imperialism, which directs people to
do everywhere exactly as they do at
home. Again, an understandably appealing approach but one that is
clearly inadequate. Consider the
large U.S. computer-products company that in 1993 introduced a
course on sexual harassment in its
Saudi Arabian facility. Under the
banner of global consistency, instructors used the same approach to
train Saudi Arabian managers that
The first claim clashes with many
people’s belief that different cultural
traditions must be respected. In
some cultures, loyalty to a community – family, organization, or society – is the foundation of all ethical
behavior. The Japanese, for example,
define business ethics in terms of
Thomas Donaldson is a professor at
the Wharton School of the University of Pennsylvania in Philadelphia,
where he teaches business ethics.
He wrote The Ethics of International
Business (Oxford University Press,
1989) and is the coauthor, with
Thomas W. Dunfee, of Business
Ethics as Social Contracts, to be
published by the Harvard Business
School Press in the fall of 1997.
HARVARD BUSINESS REVIEW
September-October 1996
159
loyalty to their companies, their
business networks, and their nation.
Americans place a higher value on
liberty than on loyalty; the U.S. tradition of rights emphasizes equality,
fairness, and individual freedom. It
is hard to conclude that truth lies on
one side or the other, but an absolutist would have us select just one.
The second problem with absolutism is the presumption that people must express moral truth using
only one set of concepts. For instance, some absolutists insist that
the language of basic rights provide
the framework for any discussion of
W O R L D
V I E W
The Culture and Ethics of
Software Piracy
Before jumping on the cultural
relativism bandwagon, stop and
consider the potential economic
consequences of a when-in-Rome
attitude toward business ethics.
Take a look at the current statistics on software piracy: In the
United States, pirated software is
estimated to be 35% of the total
software market, and industry
losses are estimated at $2.3 billion per year. The piracy rate is
57% in Germany and 80% in
Italy and Japan; the rates in most
Asian countries are estimated to
be nearly 100%.
There are similar laws against
software piracy in those countries. What, then, accounts for
the differences? Although a country’s level of economic development plays a large part, culture,
including ethical attitudes, may
be a more crucial factor. The 1995
annual report of the Software
Publishers Association connects
software piracy directly to culture and attitude. It describes
Italy and Hong Kong as having
“‘first world’ per capita incomes,
along with ‘third world’ rates of
piracy.” When asked whether one
should use software without paying for it, most people, including
people in Italy and Hong Kong,
ethics. That means, though, that entire cultural traditions must be ignored. The notion of a right evolved
with the rise of democracy in postRenaissance Europe and the United
States, but the term is not found in
either Confucian or Buddhist traditions. We all learn ethics in the context of our particular cultures, and
the power in the principles is deeply
tied to the way in which they are
expressed. Internationally accepted
lists of moral principles, such as the
United Nations’ Universal Declaration of Human Rights, draw on
say no. But people in some countries regard the practice as less
unethical than people in other
countries do. Confucian culture,
for example, stresses that individuals should share what they create with society. That may be, in
part, what prompts the Chinese
and other Asians to view the concept of intellectual property as a
means for the West to monopolize its technological superiority.
What happens if ethical attitudes around the world permit
large-scale software piracy? Software companies won’t want to invest as much in developing new
products, because they cannot expect any return on their investment in certain parts of the
world. When ethics fail to support technological creativity,
there are consequences that go
beyond statistics – jobs are lost
and livelihoods jeopardized.
Companies must do more than
lobby foreign governments for
tougher enforcement of piracy
laws. They must cooperate with
other companies and with local
organizations to help citizens understand the consequences of
piracy and to encourage the evolution of a different ethic toward
the practice.
many cultural and religious traditions. As philosopher Michael Walzer
has noted, “There is no Esperanto
of global ethics.”
The third problem with absolutism is the belief in a global standard of ethical behavior. Context
must shape ethical practice. Very
low wages, for example, may be considered unethical in rich, advanced
countries, but developing nations
may be acting ethically if they encourage investment and improve living standards by accepting low
wages. Likewise, when people are
6
malnourished or starving, a government may be wise to use more fertilizer in order to improve crop yields,
even though that means settling for
relatively high levels of thermal water pollution.
When cultures have different standards of ethical behavior – and different ways of handling unethical behavior – a company that takes an
absolutist approach may find itself
making a disastrous mistake. When
a manager at a large U.S. specialtyproducts company in China caught
an employee stealing, she followed
the company’s practice and turned
the employee over to the provincial
authorities, who executed him. Managers cannot operate in another culture without being aware of that
culture’s attitudes toward ethics.
If companies can neither adopt a
host country’s ethics nor extend the
home country’s standards, what is
the answer? Even the traditional litmus test – What would people think
of your actions if they were written
up on the front page of the newspaper? – is an unreliable guide, for
there is no international consensus
on standards of business conduct.
Balancing the Extremes:
Three Guiding Principles
Companies must help managers
distinguish between practices that
are merely different and those that
are wrong. For relativists, nothing is
sacred and nothing is wrong. For absolutists, many things that are different are wrong. Neither extreme
illuminates the real world of business decision making. The answer
lies somewhere in between.
When it comes to shaping ethical
behavior, companies must be guided
by three principles.
䡺 Respect for core human values,
which determine the absolute moral
threshold for all business activities.
䡺 Respect for local traditions.
䡺 The belief that context matters
when deciding what is right and
what is wrong.
Consider those principles in action. In Japan, people doing business
together often exchange gifts – sometimes expensive ones – in keeping
with long-standing Japanese tradition. When U.S. and European com-
HARVARD BUSINESS REVIEW
160
September-October 1996
W O R L D
panies started doing a lot of business
in Japan, many Western businesspeople thought that the practice of
gift giving might be wrong rather
than simply different. To them, accepting a gift felt like accepting a
bribe. As Western companies have
become more familiar with Japanese
traditions, however, most have come
to tolerate the practice and to set
different limits on gift giving in Japan than they do elsewhere.
Respecting differences is a crucial
ethical practice. Research shows
that management ethics differ
among cultures; respecting those
differences means recognizing that
some cultures have obvious weaknesses – as well as hidden strengths.
Managers in Hong Kong, for example, have a higher tolerance for some
forms of bribery than their Western
counterparts, but they have a much
lower tolerance for the failure to acknowledge a subordinate’s work. In
some parts of the Far East, stealing
credit from a subordinate is nearly
an unpardonable sin.
People often equate respect for
local traditions with cultural rela-
V I E W
tivism. That is incorrect. Some practices are clearly wrong. Union Carbide’s tragic experience in Bhopal,
India, provides one example. The
company’s executives seriously underestimated how much on-site
management involvement was
needed at the Bhopal plant to compensate for the country’s poor infrastructure and regulatory capabilities. In the after math of the
disastrous gas leak, the lesson is
clear: companies using sophisticated
technology in a developing country
must evaluate that country’s ability
to oversee its safe use. Since the incident at Bhopal, Union Carbide has
become a leader in advising companies on using hazardous technologies safely in developing countries.
Some activities are wrong no matter where they take place. But some
practices that are unethical in one
setting may be acceptable in another. For instance, the chemical EDB, a
soil fungicide, is banned for use in
the United States. In hot climates,
however, it quickly becomes harmless through exposure to intense solar radiation and high soil tempera-
What Do These Values Have
in Common?
Non-Western
Western
Kyosei (Japanese):
Living and working together for
the common good.
Individual liberty
Dharma (Hindu):
The fulfillment of inherited duty.
Egalitarianism
Santutthi (Buddhist):
The importance of limited
desires.
Political participation
Zakat (Muslim):
The duty to give alms to the
Muslim poor.
Human rights
HARVARD BUSINESS REVIEW
tures. As long as the chemical is
monitored, companies may be able
to use EDB ethically in certain parts
of the world.
Defining the Ethical
Threshold: Core Values
Few ethical questions are easy for
managers to answer. But there are
some hard truths that must guide
managers’ actions, a set of what I
call core human values, which define minimum ethical standards for
all companies. 1 The right to good
health and the right to economic advancement and an improved standard of living are two core human
values. Another is what Westerners
call the Golden Rule, which is recognizable in every major religious and
ethical tradition around the world.
In Book 15 of his Analects, for instance, Confucius counsels people
to maintain reciprocity, or not to do
to others what they do not want
done to themselves.
Although no single list would satisfy every scholar, I believe it is possible to articulate three core values
that incorporate the work of scores
of theologians and philosophers
around the world. To be broadly relevant, these values must include elements found in both Western and
non-Western cultural and religious
traditions. Consider the examples of
values in the insert “What Do These
Values Have in Common?”
At first glance, the values expressed in the two lists seem quite
different. Nonetheless, in the spirit
of what philosopher John Rawls
calls overlapping consensus, one can
see that the seemingly divergent values converge at key points. Despite
important differences between
Western and non-Western cultural
and religious traditions, both express shared attitudes about what it
means to be human. First, individuals must not treat others simply as
tools; in other words, they must recognize a person’s value as a human
being. Next, individuals and communities must treat people in ways
that respect people’s basic rights. Finally, members of a community
must work together to support and
improve the institutions on which
the community depends. I call those
7
September-October 1996
161
W O R L D
three values respect for human dignity, respect for basic rights, and
good citizenship.
Those values must be the starting
point for all companies as they formulate and evaluate standards of
ethical conduct at home and abroad.
But they are only a starting point.
Companies need much more specific guidelines, and the first step to
developing those is to translate the
core human values into core values
for business. What does it mean, for
example, for a company to respect
human dignity? How can a company
be a good citizen?
I believe that companies can respect human dignity by creating and
sustaining a corporate culture in
which employees, customers, and
suppliers are treated not as means to
an end but as people whose intrinsic
value must be acknowledged, and by
producing safe products and services
in a safe workplace. Companies can
respect basic rights by acting in ways
that support and protect the individual rights of employees, customers,
and surrounding communities, and
by avoiding relationships that vio-
V I E W
late human beings’ rights to health,
education, safety, and an adequate
standard of living. And companies
can be good citizens by supporting
essential social institutions, such as
the economic system and the education system, and by working with
host governments and other organizations to protect the
environment.
The core values establish a moral compass for business practice. They can help
companies identify
practices that are acceptable and those
that are intolerable –
even if the practices are compatible
with a host country’s norms and
laws. Dumping pollutants near people’s homes and accepting inadequate standards for handling hazardous materials are two examples
of actions that violate core values.
Similarly, if employing children
prevents them from receiving a basic
education, the practice is intolerable. Lying about product specifications in the act of selling may not
affect human lives directly, but it
too is intolerable because it violates
the trust that is needed to sustain
a corporate culture in which customers are respected.
Sometimes it is not a company’s
actions but those of a supplier or
customer that pose problems. Take
Many companies don’t do
anything with their codes
of conduct; they simply
paste them on the wall.
the case of the Tan family, a large
supplier for Levi Strauss. The Tans
were allegedly forcing 1,200 Chinese
and Filipino women to work 74
hours per week in guarded compounds on the Mariana Islands. In
1992, after repeated warnings to the
Tans, Levi Strauss broke off business
relations with them.
Creating an Ethical
Corporate Culture
The core values for business that
I have enumerated can help companies begin to exercise ethical judgment and think about how to operate ethically in foreign cultures, but
they are not specific enough to guide
managers through actual ethical
dilemmas. Levi Strauss relied on a
written code of conduct when figuring out how to deal with the Tan
family. The company’s Global Sourcing and Operating Guidelines, formerly called the Business Partner
Terms of Engagement, state that
Levi Strauss will “seek to identify
and utilize business partners who aspire as individuals and in the conduct of all their businesses to a set of
ethical standards not incompatible
with our own.” Whenever intolerable business situations arise, managers should be guided by precise
statements that spell out the behavior and operating practices that the
company demands.
Ninety percent of all Fortune 500
companies have codes of conduct,
and 70% have statements of vision
and values. In Europe and the Far
East, the percentages are lower but
8
HARVARD BUSINESS REVIEW
162
September-October 1996
W O R L D
are increasing rapidly. Does that
mean that most companies have
what they need? Hardly. Even
though most large U.S. companies
have both statements of values and
codes of conduct, many might be
better off if they didn’t. Too many
companies don’t do anything with
the documents; they simply paste
them on the wall to impress employees, customers, suppliers, and the
public. As a result, the senior managers who drafted the statements
lose credibility by proclaiming values and not living up to them. Companies such as Johnson & Johnson,
Levi Strauss, Motorola, Texas Instruments, and Lockheed Martin,
however, do a great deal to make the
words meaningful. Johnson & Johnson, for example, has become well
known for its Credo Challenge sessions, in which managers discuss
ethics in the context of their current
business problems and are invited to
criticize the company’s credo and
make suggestions for changes. The
participants’ ideas are passed on to
the company’s senior managers.
Lockheed Martin has created an innovative site on the World Wide
Web and on its local network that
gives employees, customers, and
suppliers access to the company’s
ethical code and the chance to voice
complaints.
Codes of conduct must provide
clear direction about ethical behavior when the temptation to behave
unethically is strongest. The pronouncement in a code of conduct
that bribery is unacceptable is useless unless accompanied by guide-
The company’s code of conduct,
however, is explicit about actual
business practice. With respect to
bribery, for example, the code states
that the “funds and assets of Motorola shall not be used, directly or
indirectly, for illegal payments of
any kind.” It is unambiguous about
what sort of payment is illegal: “the
payment of a bribe to a public official or the kickback of funds to an
employee of a customer....” The
code goes on to prescribe specific
procedures for handling
commissions to intermediaries, issuing sales invoices, and disclosing
confidential information
in a sales transaction – all
situations in which employees might have an
opportunity to accept or
offer bribes.
Codes of conduct must be explicit
to be useful, but they must also
leave room for a manager to use his
or her judgment in situations requiring cultural sensitivity. Hostcountry employees shouldn’t be
forced to adopt all home-country
Many activities are neither
good nor bad but exist in
moral free space.
lines for gift giving, payments to get
goods through customs, and “requests” from intermediaries who are
hired to ask for bribes.
Motorola’s values are stated very
simply as “How we will always act:
[with] constant respect for people
[and] uncompromising integrity.”
HARVARD BUSINESS REVIEW
V I E W
values and renounce their own.
Again, Motorola’s code is exemplary. First, it gives clear direction:
“Employees of Motorola will respect
the laws, customs, and traditions of
each country in which they operate,
but will, at the same time, engage in
no course of conduct which, even if
legal, customary, and accepted in
any such country, could be deemed
to be in violation of the accepted
business ethics of Motorola or the
laws of the United States relating to
business ethics.” After laying down
such absolutes, Motorola’s code
then makes clear when individual
judgment will be necessary. For example, employees may sometimes
accept certain kinds of small gifts
“in rare circumstances, where the
refusal to accept a gift” would injure Motorola’s “legitimate business
interests.” Under certain circumstances, such gifts “may be accepted
so long as the gift inures to the benefit of Motorola” and not “to the
benefit of the Motorola employee.”
Striking the appropriate balance
between providing clear direction
and leaving room for individual
9
September-October 1996
163
W O R L D
judgment makes crafting corporate
values statements and ethics codes
one of the hardest tasks that executives confront. The words are only a
start. A company’s leaders need to
refer often to their organization’s
credo and code and must themselves
be credible, committed, and consistent. If senior managers act as
though ethics don’t matter, the rest
of the company’s employees won’t
think they do, either.
Conflicts of Development
and Conflicts of Tradition
Managers living and working
abroad who are not prepared to grapple with moral ambiguity and tension should pack their bags and
come home. The view that all business practices can be categorized as
either ethical or unethical is too
simple. As Einstein is reported to
have said, “Things should be as simple as possible – but no simpler.”
Many business practices that are
considered unethical in one setting
may be ethical in another. Such activities are neither black nor white
but exist in what Thomas Dunfee
and I have called moral free space.2
In this gray zone, there are no tight
prescriptions for a company’s behavior. Managers must chart their own
courses – as long as they do not violate core human values.
Consider the following example.
Some successful Indian companies
offer employees the opportunity for
one of their children to gain a job
with the company once the child has
completed a certain level in school.
The companies honor this commitment even when other applicants
are more qualified than an employee’s child. The perk is extremely
valuable in a country where jobs are
hard to find, and it reflects the Indian culture’s belief that the West has
gone too far in allowing economic
opportunities to break up families.
Not surprisingly, the perk is among
the most cherished by employees,
but in most Western countries, it
would be branded unacceptable
nepotism. In the United States, for
example, the ethical principle of
equal opportunity holds that jobs
should go to the applicants with the
best qualifications. If a U.S. com-
V I E W
pany made such promises to its employees, it would violate regulations
established by the Equal Employment Opportunity Commission.
Given this difference in ethical attitudes, how should U.S. managers
react to Indian nepotism? Should
they condemn the Indian companies, refusing to accept them as partners or suppliers until they agree to
clean up their act?
Despite the obvious tension be-
tween nepotism and principles of
equal opportunity, I cannot condemn the practice for Indians. In a
country, such as India, that emphasizes clan and family relationships
and has catastrophic levels of unemployment, the practice must be
viewed in moral free space. The decision to allow a special perk for employees and their children is not necessarily wrong – at least for members
of that country.
The Problem with Bribery
Bribery is widespread and insidious. Managers in transnational companies routinely confront
bribery even though most countries have laws against it. The fact
is that officials in many developing countries wink at the practice, and the salaries of local bureaucrats are so low that many
consider bribes a form of remuneration. The U.S. Foreign Corrupt Practices Act defines allowable limits on petty bribery in
the form of routine payments
required to move goods through
customs. But demands for bribes
often exceed those limits, and
there is seldom a good solution.
Bribery disrupts distribution
channels when goods languish on
docks until local handlers are
paid off, and it destroys incentives to compete on quality and
cost when purchasing decisions
are based on who pays what under
the table. Refusing to acquiesce is
often tantamount to giving business to unscrupulous companies.
I believe that even routine
bribery is intolerable. Bribery undermines market efficiency and
predictability, thus ultimately
denying people their right to a
minimal standard of living. Some
degree of ethical commitment –
some sense that everyone will
play by the rules – is necessary for
a sound economy. Without an
ability to predict outcomes, who
would be willing to invest?
10
There was a U.S. company
whose shipping crates were regularly pilfered by handlers on the
docks of Rio de Janeiro. The handlers would take about 10% of
the contents of the crates, but the
company was never sure which
10% it would be. In a partial solution, the company began sending
two crates – the first with 90% of
the merchandise, the second with
10%. The handlers learned to
take the second crate and leave
the first untouched. From the
company’s perspective, at least
knowing which goods it would
lose was an improvement.
Bribery does more than destroy
predictability; it undermines essential social and economic systems. That truth is not lost on
businesspeople in countries
where the practice is woven into
the social fabric. CEOs in India
admit that their companies engage constantly in bribery, and
they say that they have considerable disgust for the practice. They
blame government policies in
part, but Indian executives also
know that their country’s business practices perpetuate corrupt
behavior. Anyone walking the
streets of Calcutta, where it is
clear that even a dramatic redistribution of wealth would still
leave most of India’s inhabitants
in dire poverty, comes face-toface with the devastating effects
of corruption.
HARVARD BUSINESS REVIEW
164
September-October 1996
W O R L D
How can managers discover the
limits of moral free space? That is,
how can they learn to distinguish
a value in tension with their own
from one that is intolerable? Helping
managers develop good ethical judgment requires companies to be clear
about their core values and codes of
conduct. But even the most explicit
set of guidelines cannot always provide answers. That is especially true
in the thorniest ethical dilemmas,
in which the host country’s ethical
standards not only are different but
also seem lower than the home
following question: Would the practice be acceptable at home if my
country were in a similar stage of
economic development? Consider
the difference between wage and
safety standards in the United States
and in Angola, where citizens accept
lower standards on both counts. If a
U.S. oil company is hiring Angolans
to work on an offshore Angolan oil
rig, can the company pay them lower
wages than it pays U.S. workers in
the Gulf of Mexico? Reasonable people have to answer yes if the alternative for Angola is the loss of both the
foreign investment and
the jobs.
Consider, too, differences in regulatory environments. In the 1980s,
the government of India
fought hard to be able to
import Ciba-Geigy’s
Entero Vioform, a drug
known to be enormously
effective in fighting
dysentery but one that had been
banned in the United States because
some users experienced side effects.
Although dysentery was not a big
problem in the United States, in India, poor public sanitation was contributing to epidemic levels of the
disease. Was it unethical to make
the drug available in India after it
had been banned in the United
States? On the contrary, rational
people should consider it unethical
not to do so. Apply our test: Would
the United States, at an earlier stage
of development, have used this drug
despite its side effects? The answer
is clearly yes.
But there are many instances
when the answer to similar questions is no. Sometimes a host country’s standards are inadequate at any
level of economic development. If a
country’s pollution standards are so
low that working on an oil rig would
considerably increase a person’s risk
of developing cancer, foreign oil
companies must refuse to do business there. Likewise, if the dangerous side effects of a drug treatment
outweigh its benefits, managers
should not accept health standards
that ignore the risks.
When relative economic conditions do not drive tensions, there is
If a company declared all
gift giving unethical, it
wouldn’t be able to do
business in Japan.
country’s. Managers must recognize
that when countries have different
ethical standards, there are two
types of conflict that commonly
arise. Each type requires its own line
of reasoning.
In the first type of conflict, which
I call a conflict of relative development, ethical standards conflict because of the countries’ different levels of economic development. As
mentioned before, developing countries may accept wage rates that
seem inhumane to more advanced
countries in order to attract investment. As economic conditions in a
developing country improve, the
incidence of that sort of conflict usually decreases. The second type of
conflict is a conflict of cultural tradition. For example, Saudi Arabia,
unlike most other countries, does
not allow women to serve as corporate managers. Instead, women may
work in only a few professions, such
as education and health care. The
prohibition stems from strongly
held religious and cultural beliefs;
any increase in the country’s level of
economic development, which is
already quite high, is not likely to
change the rules.
To resolve a conflict of relative development, a manager must ask the
HARVARD BUSINESS REVIEW
V I E W
a more objective test for resolving
ethical problems. Managers should
deem a practice permissible only if
they can answer no to both of the
following questions: Is it possible
to conduct business successfully in
the host country without undertaking the practice? and Is the practice a
violation of a core human value?
Japanese gift giving is a perfect example of a conflict of cultural tradition. Most experienced businesspeople, Japanese and non-Japanese alike,
would agree that doing business in
Japan would be virtually impossible
without adopting the practice. Does
gift giving violate a core human
value? I cannot identify one that it
violates. As a result, gift giving may
be permissible for foreign companies
in Japan even if it conflicts with ethical attitudes at home. In fact, that
conclusion is widely accepted, even
by companies such as Texas Instruments and IBM, which are outspoken against bribery.
Does it follow that all nonmonetary gifts are acceptable or that
bribes are generally acceptable in
countries where they are common?
Not at all. (See the insert “The Problem with Bribery.”) What makes the
routine practice of gift giving acceptable in Japan are the limits in its
scope and intention. When gift giving moves outside those limits, it
soon collides with core human values. For example, when Carl Kotchian, president of Lockheed in the
1970s, carried suitcases full of cash
to Japanese politicians, he went beyond the nor ms established by
Japanese tradition. That incident
galvanized opinion in the United
States Congress and helped lead to
passage of the Foreign Corrupt Practices Act. Likewise, Roh Tae Woo
went beyond the norms established
by Korean cultural tradition when
he accepted $635.4 million in bribes
as president of the Republic of Korea
between 1988 and 1993.
Guidelines for Ethical
Leadership
Learning to spot intolerable practices and to exercise good judgment
when ethical conflicts arise requires
practice. Creating a company culture that rewards ethical behavior is
11
September-October 1996
165
W O R L D
essential. The following guidelines
for developing a global ethical perspective among managers can help.
Treat corporate values and formal
standards of conduct as absolutes.
Whatever ethical standards a company chooses, it cannot waver on its
principles either at home or abroad.
Consider what has become part of
company lore at Motorola. Around
1950, a senior executive was negotiating with officials of a South American government on a $10 million
sale that would have increased the
company’s annual net profits by
nearly 25%. As the negotiations
neared completion, however, the executive walked away from the deal
because the officials were asking for
$1 million for “fees.” CEO Robert
Galvin not only supported the executive’s decision but also made it
clear that Motorola would neither
accept the sale on any terms nor do
business with those government
officials again. Retold over the
decades, this story demonstrating
Galvin’s resolve has helped cement
a culture of ethics for thousands of
employees at Motorola.
Design and implement conditions
of engagement for suppliers and customers. Will your company do business with any customer or supplier?
What if a customer or supplier uses
child labor? What if it has strong
links with organized crime? What if
it pressures your company to break
a host country’s laws? Such issues
are best not left for spur-of-themoment decisions. Some companies
have realized that. Sears, for instance, has developed a policy of not
contracting production to companies that use prison labor or infringe
on workers’ rights to health and
safety. And BankAmerica has specified as a condition for many of its
loans to developing countries that
environmental standards and human
rights must be observed.
Allow foreign business units to
help formulate ethical standards and
interpret ethical issues. The French
pharmaceutical company Rhône-
V I E W
Poulenc Rorer has allowed foreign
subsidiaries to augment lists of corporate ethical principles with their
own suggestions. Texas Instruments
has paid special attention to issues
of international business ethics by
creating the Global Business Practices Council, which is made up of
managers from countries in which
the company operates. With the overarching intent to create a “global
ethics strategy, locally deployed,”
the council’s mandate is to provide
ethics education and create local
processes that will help managers
in the company’s foreign business
units resolve ethical conflicts.
In host countries, support efforts
to decrease institutional corruption.
Individual managers will not be able
to wipe out corruption in a host
country, no matter how many bribes
they turn down. When a host country’s tax system, import and export
procedures, and procurement practices favor unethical players, companies must take action.
Many companies have begun to
participate in reforming host-country institutions. General Electric, for
example, has taken a strong stand in
India, using the media to make repeated condemnations of bribery in
business and government. General
Electric and others have found, however, that a single company usually
cannot drive out entrenched corruption. Transparency International, an
organization based in Germany, has
been effective in helping coalitions
of companies, government officials,
and others work to reform briberyridden bureaucracies in Russia, Bangladesh, and elsewhere.
Exercise moral imagination. Using moral imagination means resolving tensions responsibly and creatively. Coca-Cola, for instance, has
consistently turned down requests
for bribes from Egyptian officials but
has managed to gain political support and public trust by sponsoring a
project to plant fruit trees. And take
the example of Levi Strauss, which
discovered in the early 1990s that
12
two of its suppliers in Bangladesh
were employing children under the
age of 14–a practice that violated the
company’s principles but was tolerated in Bangladesh. Forcing the suppliers to fire the children would not
have ensured that the children received an education, and it would
have caused serious hardship for the
families depending on the children’s
wages. In a creative arrangement,
the suppliers agreed to pay the children’s regular wages while they attended school and to offer each child
a job at age 14. Levi Strauss, in turn,
agreed to pay the children’s tuition
and provide books and uniforms.
That arrangement allowed Levi
Strauss to uphold its principles and
provide long-term benefits to its
host country.
Many people think of values as
soft; to some they are usually unspoken. A South Seas island society
uses the word mokita, which means,
“the truth that everybody knows but
nobody speaks.” However difficult
they are to articulate, values affect
how we all behave. In a global business environment, values in tension
are the rule rather than the exception. Without a company’s commitment, statements of values and
codes of ethics end up as empty platitudes that provide managers with
no foundation for behaving ethically. Employees need and deserve
more, and responsible members of
the global business community can
set examples for others to follow.
The dark consequences of incidents
such as Union Carbide’s disaster in
Bhopal remind us how high the
stakes can be.
1. In other writings, Thomas W. Dunfee and
I have used the term hypernorm instead of
core human value.
2. Thomas Donaldson and Thomas W. Dunfee,
“Toward a Unified Conception of Business
Ethics: Integrative Social Contracts Theory,”
Academy of Management Review, April 1994;
and “Integrative Social Contracts Theory:
A Communitarian Conception of Economic
Ethics,” Economics and Philosophy, spring 1995.
Reprint 96502
To place an order, call 1-800-545-7685.
HARVARD BUSINESS REVIEW
166
September-October 1996
Harvard Business School
9-299-009
August 11, 1998
The General Motors Corporation (D)
1993-1996
In December 1996, John D. Finnegan and his staff were preparing for an upcoming meeting of
the Board of Directors of the General Motors Corporation (GM) that was scheduled for
January 27, 1997. Finnegan, the Treasurer of GM, was responsible for developing a plan to address a
cash surplus identified in GM’s most recent five year plan. The Board would consider whether GM
should hold excess cash, increase the firm’s dividend, or repurchase shares. If the firm were to
proceed with a stock buy-back, Finnegan and his group would need to recommend how to proceed
with the details of its execution and whether to incorporate the use of puts or an Accelerated Share
Repurchase structure. These decisions would be made in light of GM’s financial policies, its recent
history, and its projections of future needs.
Rebuilding General Motors’ Balance Sheet
GM’s new management team inherited a difficult position when they took over at the end of
1992. Having maintained solvency through the most difficult part of a downturn in the auto industry
and having raised sufficient capital to begin the company’s rebuilding process, the new management
next provided an internal mandate to “clean up” the company’s balance sheet. Specifically, the
Treasury staff was given until the end of 1994 to accomplish the following objectives:
•
Determine cash balances necessary to survive the next automotive downturn;
•
Determine the debt-to-total capital ratio that moves the corporation to its optimal
WACC, work with the rating agencies on communicating a plan necessary to
improve the credit rating, and make the changes to the capital structure that are
necessary to meet the target;
•
Determine a sustainable dividend level and dividend yield and regain investor
confidence through an achievable and well communicated dividend policy; and
Charles M. Williams Fellow Markus F. Mullarkey and William J. Wildern, IV, MBA 1997 prepared this case under the
supervision of Professor Peter Tufano as the basis for class discussion rather than to illustrate either effective or ineffective
handling of an administrative situation. It is part of a four-part General Motors Corporation case series including HBS
Cases #299-006 through #299-009.
Copyright © 1998 by the President and Fellows of Harvard College. To order copies or request permission to
reproduce materials, call 1-800-545-7685 or write Harvard Business School Publishing, Boston, MA 02163. No
part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in
any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the
permission of Harvard Business School.
1
167
299-009
The General Motors Corporation (D)
•
Fully fund the pension plan.
The key financial targets that the Group recommended are described in the General Motors
Corporation (B) case.
The return to profitability
Between 1993 and 1996, U.S. automotive sales began to rebound from the slump in 1990-1992.
By 1996, retail sales of passenger cars and light trucks were expected to approach 17 million units, up
13% from 1992. While GM’s market share fell from 33.1% in 1992 to 31.0% in 1996, total GM units
sold in the U.S. rose 6% in this period, and revenues rose 24%. Coupled with an improvement in
reduction in cost of goods sold, this generated a turnaround in pre-tax income from negative $3.3
billion in 1992 to a profit of $6.7 billion in 1996 (see Exhibit 1).
The improvement in GM’s auto business significantly relaxed the cash difficulties the
company had been facing, generating an estimated operating cash flow of $18.7 billion in 1996, over
three times its operating cash flow of $5.9 billion in 1991. This allowed the company to re-establish its
investment program, and between 1994 and 1996, the company dedicated $27.2 billion in cash flow to
investing activities, 28% above 1992 spending levels. In addition, stronger results allowed GM to
resume contributions to its corporate pension plans, reducing GM’s underfunded pension liability
from $22.3 billion to $4.6 billion between 1993 and 1996.1 As a result of its improved performance,
GM’s stock price improved over the 1992-1996 time period, despite a slight recession in 1994 that
affected all auto manufacturers (see Exhibit 2).
Increasing the dividend
As a result of the improved performance, in the second quarter of 1995, GM’s Board of
Directors voted to increase the company’s quarterly dividend from $0.20 per quarter to $0.30 per
quarter. This was done primarily to keep GM’s dividend yield on par with those of the other major
U.S. automakers, both of whom had been increasing their dividends recently and were expected to do
so again in the second quarter of 1995 (see Exhibit 3). In addition, executives recognized that, “a lot
of work had been done and people wanted to send a signal to the market that GM was on track
again.”2 The dividend was increased to $.40 per quarter nine months later in the first quarter of 1996,
still well below the $0.75 per quarter GM paid prior to the dividend cut in 1991. 3 These dividend
increases were intended as a signal of GM’s confidence in its future earnings potential.
1 See Harvard University Kennedy School of Government Cases #C108-97-1385.0 and #C108-97-1386.0,
Protecting Pension Benefits: The Pension Benefit Guarantee Corporation Meets General Motors: Parts A and B, for more
information on the history and details of GM’s pension actions during this period.
2 However, despite an initial 1.0% increase in GM’s stock price over the two days following the dividend
announcement, GM’s share price fell 1.4% over the three week period following the dividend increase (see
Exhibit 4). In the comparable two-day and three-week periods, the S&P 500 index rose 1.7% and 3.8%
respectively.
3 The immediate market reaction to the 1996 dividend announcement was strong. GM’s share price increased
2.3% in the two days following the announcement, while the S&P 500 index was up 0.7%. In addition, over
three weeks following the announcement GM’s share price increased further, gaining a total of 7.6% (see Exhibit
4).
2
168
The General Motors Corporation (D)
299-009
Cleaning up the balance sheet
As a result of GM’s improving operating performance, the finance staff was able to undertake
a number of programs in an effort to reconfigure the company’s balance sheet, to reduce leverage, to
upgrade the company’s rating, and to simplify the structure of its equity claims.
Paying down debt By 1993, GM’s total debt-to-total capital (book) ratio had exceeded 58.3%. In
addition, throughout the 1990-1992 period, GM’s EBIT-to-interest expense ratio was negative,
reflecting the company’s operating losses, and even its ratio of EBITDA to interest expense dropped
below 1.00 in 1991. Between 1992 and 1994, however, GM began decreasing its debt outstanding in
an effort to reduce the company’s debt-to-total capital ratios. As a result, by the end of 1996, the
company was expected to meet its leverage targets, with a debt-to-total capital ratio of 22% and an
EBIT-to-interest coverage ratio of 6.2.
Converting the PERCS In 1991, GM had issued 17.8 million shares of Preference Equity
Redemption Cumulative Stock, which paid $.83 per quarter. In essence the PERCS were a form of
common stock with capped appreciation. June 1994, the PERCS were converted into GM common
stock on a one-for-one basis, reducing its preferred dividend payments and lowering overall dividend
payments. GM had the option to pay $53.79 per share in cash between May 1 and July 1, 1994 to force
conversion prior to the mandatory one-for-one stock conversion date of July 1, 1994. With common
stock trading between $50 and $55 per share in the three months prior to mandatory conversion, the
company announced that it would convert all outstanding PERCS on a one-for-one basis into GM
common as of the closing price on June 17, 1994. The stock closed at $53-7/8 and all outstanding
PERCS were converted. The conversion was covered by issuing 17.8 million new shares (increasing
the total number of outstanding GM common shares by that number).
Tendering the Preference Stocks GM also sought to eliminate a substantial amount of its
outstanding preference stock. This stock, which had been issued in the early 1990’s, when GM was
under financial stress, carried an average dividend coupon of 8.9%. This was very inefficient in
comparison to debt financing because the dividend coupon (unlike an interest coupon on a debt
security) was not tax deductible. As a result, GM sought to tender for the shares, thereby reducing
the after-tax coupon it paid, cleaning-up the company’s balance sheet, and improving the efficiency of
its financing. In May, 1995, GM successfully tendered 57% of the Series B, D, and G preferred stock
with an outlay of $1.3 billion in cash.
GM also considered the use of a number of other innovative structures to maximize shareholder
value. One such structure, Trust Originated Preferred Shares (TOPrS), was being considered in late1996 for implementation in 1997. In essence, the structure replaced the preference stock that
remained after the May 1995 tender offer with subordinated debentures through the use of an
innovative trust structure. While the resulting structure was economically quite similar to the original
structure, it was far more efficient financially because of the significant tax advantage it created as
GM deducted interest payments on the debentures for tax purposes. However, with respect to the
economics of the transaction and the accounting by GM, TOPrS would retain many attractive features
of equity.
GM’s 1996 Balance Sheet Review
Thus, by the end of 1996, the company was largely restored to financial health as measured
by the policy targets it had set in the early 1990s. The company was projected by the end of the year
to achieve its leverage target (20%-25% debt-to-total capital on a book basis) and to have fully funded
its pension liability when measured on an economic basis. At the same time, the company was
working with rating agencies to restore its debt rating to a credible single-A level, and was expecting
3
169
299-009
The General Motors Corporation (D)
action in that regard by the middle of 1997. Lastly, for the first time since the financial crisis, the
company was projecting a substantial cash surplus over the target cash balances to which it had
committed itself at the time of its crisis in 1992 ($13 billion), thereby prompting a review of the
company’s future cash needs.
The first step in determining the appropriate cash balances for the company going forward
was to update the company’s 1993 analysis of its likely cash needs over the coming 5 year business
plan. GM’s capital planning group initiated the process by putting together financial projections
which incorporated not only the company’s budgets and strategic plans, but also the market
conditions the company was expected to face. The casewriters’ estimates of these projections
indicated that GM could be expected to generate approximately $4.9 billion in excess cash (beyond
the target level) by 1997 and $15 billion in cash over the next five years (see Exhibit 5).4
Realizing that the company would post cash balances well above its $13 billion cash target by
the end of 1996, many investors had begun to speculate openly about the company’s response. In
October 1996 one analyst forecasted, “a large dividend hike [and] a share buyback program.”
Similarly, another analyst stated that, “in addition to dividend increases, we would also attach a high
probability to a share repurchase authorization being passed within the next 24 months.”
As a result, the Treasury staff was asked to provide a recommendation about the disposition
of the company’s cash balances to be discussed in early 1997 with the President’s Council composed
of the seven top executives of the corporation.5 If approved, the recommendation would be discussed
at the January 27, 1997 meeting of the firm’s Board of Directors. In the meantime, the Treasury staff
would work with the company’s Chief Financial Officer, J. Michael Losh, to generate and evaluate
alternatives. Given GM’s recent history, Finnegan was sure the matter would receive an unusually
high level of attention and scrutiny. As one executive noted, “you have to remember, a lot of these
people had the unpleasant experience of living through that tumultuous period when they had to
defend their actions day-in and day-out. Now that GM was out of trouble, many of them saw a
chance to put forth a policy that enabled them to fulfill their past promises.” The finance team knew
the importance of the task, and the visibility it would receive, both inside and outside GM.
GM’s Alternatives
Broadly speaking, there were four main alternatives GM needed to consider moving forward:
spending the money on investments beyond the company’s current five year investment plan,
holding the money as an additional cash buffer against future downturns, reducing the firm’s
leverage, or returning cash to investors in the form of dividends (either a one-time special dividend or
a common stock dividend increase) or a share repurchase program.
4 Estimates of GM’s future cash balances are made by the casewriters and based on historical trends and
economic and automotive market estimates. They are for illustrative purposes only and do not necessarily
represent the company’s actual expectations of its future cash position.
5 In January 1997, the President’s Council consisted of: John F. Smith, Jr., Chairman, CEO, and President;
J. Michael Losh, CFO; G. Richard Wagoner, President North American Operations; Louis R. Hughes, President
International Operations; Harry J. Pearce, Vice Chairman; C. Michael Armstrong, CEO Hughes Electronics
Corporation; and J. T. Battenberg, III, President of Delphi Automotive Systems.
4
170
The General Motors Corporation (D)
299-009
Incremental spending
One potential use of the surplus cash balance was for GM to find incremental spending
opportunities that would enhance shareholder value, or to accelerate its currently planned investment
program. A number of alternatives were considered such as the following:
Investments Executives at GM had analyzed the potential for incremental growth or globalization
initiatives, and appreciated the importance of maintaining the flexibility to take advantage of strategic
opportunities that might arise. As a result, the Treasury staff felt that the allocation of some funds to
cover such contingencies was an attractive proposition. These opportunities might require as much
as $500 million in investment in any given year, though it was too late in 1996 to begin any initiatives
in that calendar year.
Pension Contributions Despite having virtually eliminated its large underfunded pension
liability, GM could still make incremental contributions to its pension programs as a buffer against
potential poor investment returns, future increased labor settlements, or changes in the discount rate
at which future obligations would be calculated. However, the company did not foresee a significant
risk of falling below the regulatory constraints on their pension funding given the contributions that
were already planned.
Holding the Cash
GM could hold onto the cash and boost its cash balances. Given the large impact an
economic downturn could have on GM’s profitability and cash flows (see Exhibit 6), there were some
on the Board of Directors that felt that erring on the conservative side was appropriate. They were
also concerned by the projected slowdown in automotive sector performance that was expected to
occur during 1997, and were aware of the difficulties a number of firms in the automotive industry
had faced in periods where cash became constrained.
Finnegan realized, however, that conservatism was not without its costs. Because cash
balances earned considerably lower returns than capital invested in a profitable business, Finnegan
did not want GM to be more conservative than was realistically necessary to ensure the company
could continue making profitable investments as opportunities arose. Beyond that point, however,
more cash would be a drain on GM’s return on equity. Other costs of large cash balances involved
increasing the likelihood of shareholder criticism and activism against GM management.
In evaluating this alternative, Finnegan would rely on the 1993 analysis by the Planning
Group that had been recently revised. The analysis, which examined the company’s possible usage of
cash under potentially severe economic scenarios, recommended a cash balance of $13 billion would
provide ample insurance against a worst-case scenario.
Reducing debt
Excess cash can be thought of as negative leverage, which gives rise to taxable interest income
and additional earnings. Rather than hold more cash, GM could reduce its leverage by repurchasing
outstanding debt in the marketplace. Given recent declines in interest rates and the recent
improvements in GM’s credit quality, however, the debt would have to be repurchased at prices
above par, giving rise to negative earnings charges as the company took an immediate
extinguishment of debt charge (the full loss would be recognized immediately, not amortized over
future periods). However, the loss would also generate a positive tax impact by accelerating the
recognition of the write-down for tax purposes. The alternative of reducing the debt balance would
have to be evaluated within the context of the company’s stated total debt-to-total capital targets,
which specified a 20-25% debt-to-total capital ratio and an A rating or higher.
5
171
299-009
The General Motors Corporation (D)
Returning the cash to shareholders
Rather than spending the cash, holding on to it, or using it to reduce leverage, GM could
return the cash to investors. This was a contentious issue among some key Board of Directors
members. For example, some felt that investment advisors would consider returning cash to
shareholders as an indication that the company was “out of good ideas,” and that would be a poor
reflection on its management as well as its Board of Directors. However, as one very well known
investor had remarked to GM executives, “As a mature company in a cyclical industry, you shouldn’t
be getting capital from the market, you should be giving it back.” This critique resonated with many
senior executives and Board members. Cash could be returned to shareholders via dividends or
share repurchases, each of which is treated in more detail below:
Dividends
In analyzing potential dividend increases, the company referred to the dividend policy which
had been set following the dividend cuts in the early 1990s.6 The policy indicated that GM’s dividend
decisions should be guided by an analysis of and comparison to competitors’ dividend yields, an
ability to maintain a reasonable average payout ratio throughout the business cycle, and an ability to
sustain dividend payments over the long-term.
Analysis of Chrysler and Ford’s dividends indicated that GM’s dividend yield (its annual
dividend as a percentage of common stock price) was somewhat lower than its competitors (see
Exhibit 3). Whereas Ford and Chrysler currently delivered 4.5% dividend yields, GM’s current
quarterly dividend of $0.40 per share resulted in a dividend yield of only 3.0%.7 Many felt this was a
problem as it would place GM at a disadvantage vis-à-vis other auto companies with respect to their
ability to attract yield-oriented investors who were interested in holding auto stocks. In addition, this
yield was much lower than GM’s historical dividend yield which had ranged from 5.5% to 8.0%
between 1987 and 1990. The Treasury staff had also performed an analysis of GM’s dividend payout
ratio.
Lastly, the Treasurer’s Office took the cash forecast prepared by the Capital Planning group
and estimated the level of dividend that could be sustained over the following 5 years. Casewriters’
estimates indicate that an annual dividend increase of $0.30 per quarter, while feasible in good times,
would leave the company with little flexibility. A large economic downturn might force GM to
decide among a dividend cut, a cut in planned investment, or an increase in external financing. This
is precisely the sort of situation GM was seeking to avoid, and the reason GM had created its stated
financial policies after the financial crisis in the early 1990’s.
One way to mitigate the risk of future dividend cuts while still announcing a large dividend
was to simply announce a “special dividend.” If labeled in this way, investors would understand that
the dividend was not expected to be repeated on a quarterly basis, and would not be disappointed if
the dividend could not be maintained. In fact, this was exactly how GM used to manage its dividend
payout in the 1970’s, when second and fourth quarter special dividends came to be an expected part
of GM stock ownership. In years when the company was producing strong results, the special
dividend would be large, whereas in slower years, it would shrink or disappear altogether. While this
seemed to avert the risk of announcing a strong dividend, there was some question as to whether it
had any impact on the company’s stock price. As CFO J. Michael Losh put it, “We were doing then
what we now know is exactly the wrong thing. When you use special dividends, you get no credit
from investors for the dividend, because they know they can’t count on it in the future.”
6 See HBS Case #299-007, The General Motors Corporation (B) for more details on GM’s dividend policy decisions.
7 The average dividend yield of stocks in the S&P 500 index at the end of 1996 was approximately 2.0%.
6
172
The General Motors Corporation (D)
299-009
Share repurchase programs
Another option to address the cash surplus was a share repurchase program. This had a
number of potential benefits as compared with a dividend increase or special dividend. For example,
it was considered more tax efficient from the standpoint of taxable investors because the maximum
federal income tax rate on dividend income was higher than that on capital gains income (39.6% to
the highest taxed investor versus a 28% long-term capital gains tax8). Moreover, a share repurchase
program would have a positive EPS benefit compared to a dividend increase due to an unchanged
net income being applied to a reduced number of shares outstanding upon completion of the
program. While it was unclear whether this EPS boost would be perceived by the market as
advantageous, some reasoned that if the P/E multiple for GM remained in the then current 8 to 9
range for an automotive company, GM’s stock price would naturally rise due to increasing EPS.
Additionally, if it were possible to consistently and predictably boost EPS through the use of
repurchases, the stock might even eventually enjoy a higher P/E multiple as a result.
There were also potential signaling benefits of announcing a stock buyback program.
Research from academics and Wall Street analysts showed that companies that announced share
repurchases had been shown to enjoy an average stock price boost anywhere from 2%-15% around
the time of the announcement (see Exhibit 7). Some studies documented the persistence of these
excess returns for longer periods of time after the announcement.9
Implementing a repurchase
If GM decided to pursue a repurchase program, there were three primary ways to execute it:
through an open market purchase, a fixed-price tender offer, or a Dutch auction style tender offer.
Exhibit 8 provides a description of each method.
Open market repurchases Of the three primary forms of share repurchases, open market
repurchase programs were the most widely used, accounting for roughly 90% of repurchase
programs initiated between 1989 and 1996.10 In an open market repurchase, a company simply
repurchased shares through brokers on the open market. The overall size of the repurchase program
would be announced to shareholders, though specific repurchases would not need to be reported
individually. The primary advantage of such a program was its flexibility with respect to timing,
purchase price, and number of shares purchased. In addition, since GM could speed up or slow
down the rate at which it repurchased shares over time, it could maintain flexibility to match its
seasonal cash flows. In addition, opportunistic purchases of large blocks at attractive pricing could
speed up completion of the program while minimizing its cost. Lastly, no pre-set premium was
anticipated on the shares, and skillful broker execution could actually lead to an average repurchase
price somewhat below market prices (reportedly as much as 5%).
The major drawbacks to the program were the risk of price appreciation of GM’s shares (which would
increase the average repurchase price) and the restriction on the company’s ability to execute the
program quickly. For example, SEC Rule 10(b)-18 limited the shares of its own stock a company
8 US tax rates in effect in December 1996.
9 Samuel Stewert, Jr., for example, documents a 5-20% excess annual return over the five years subsequent to a
repurchase program (see “Should a Corporation Repurchase its Own Stock?,” The Journal Of Finance, June 1976),
while Ikenbrery, Lakonishok, and Vermaelen document a 12.6% average excess returns over the four years
following repurchase announcements (see “Market Underreaction to Open Market Share Repurchases,” Working
Paper No. 106, Rice University, July 1994).
10 Reported by Chris Innes, Peter B. Blanton, Nomo Nomo-Ongolo, and Cindy Sieden, “Stock Buybacks –
Strategy and Tactics,” Salomon Brothers, February 1997 using data from Securities Data Company.
7
173
299-009
The General Motors Corporation (D)
could repurchase to approximately 25% of the company’s daily trading volume.11 As a result, open
market repurchases usually took longer to complete than other means of repurchasing shares.
Fixed-price tender offer Fixed-price tender offers were less common than open market
repurchases, accounting for 6.5% of all programs between 1989 and 1996. Under a fixed-price tender
offer, GM would announce the proposed tender offer including the number of shares it sought to
repurchase and the fixed price at which it offered to purchase them. It would then file with the SEC
and mail tender documents to shareholders.
This alternative had a number of advantages including the rapid timing (within seven to nine weeks)
and known costs of the program. It also provided the ability for management to describe the
rationale for the program in the offering documentation mailed to shareholders. The drawbacks,
however, included its inflexibility and the higher costs of the program as compared with other
alternatives. For example, once announced, GM would not be able to change the price of the tender
offer as the market price of its shares moved. Nor could the number of shares which would be
repurchased be altered. Also, the tender was typically carried out at a significant premium (often as
large as 15% above the then prevailing market price of the shares).
Dutch-auction style tender offer Dutch-auction style tender offers accounted for 3.8% of all
announced share repurchase programs. Under this type of program, a company would announce the
number of shares it sought to repurchase and an expiration date for the offer, but instead of
announcing a tender price, it would announce a minimum and maximum repurchase price.
Tendering shareholders would then offer to sell their shares at prices within the range announced by
the company. At the expiration of the offer, the price required to repurchase all of the shares sought
by the company would be set and shareholders tendering at that price or lower would sell their
shares at the tender price. Shareholders who tendered at higher prices would not be bought out. The
primary advantage of this alternative as compared with a fixed-price tender is that it decreased the
risk of significant over- or under-pricing of the repurchase (because the eventual tender price was
effectively set by the market and flexible over the timeframe of the offer). The primary drawback was
that the complexity of the process might reduce the likelihood of small shareholders tendering shares,
and therefore increase the risk that the offer would go only partially subscribed.
In addition to determining the method of repurchase, GM would also have to consider what
size program to attempt. GM had approached a number of the company’s bankers, who had advised
the company that a share repurchase needed to target at least 5% of the company’s outstanding
shares (about 37 million shares in GM’s case) for it to be considered a credible signal of the
undervalued status of the company’s stock. This was supported by research findings which indicated
that open market share repurchase programs accounting for 5% or more of a company’s outstanding
shares generated a 2.2% three day excess return, while those accounting for less than 5% only
generated a 0.6% excess return.12 Beyond 5%, there was disagreement whether incremental size
would have an impact. Tender offers (fixed-price or dutch-auction), however, were usually only
undertaken in instances where the company was repurchasing 10% or more of its outstanding
shares.13
11 Rule 10(b)-18 was intended to help prevent a corporation from artificially boosting its stock price through
market intervention. Basically, a firm can only be represented by one broker on any given treading day, and
that broker can not participate in the first trade of the day, the last half hour of the day, or on any bid where the
last bid was lower than the current bid (an uptick).
12 See Chris Innes, Peter B. Blanton, Nomo Nomo-Ongolo, and Cindy Sieden, “Stock Buybacks – Strategy and
Tactics,” Salomon Brothers, February 1997.
13 Similar to the above study with respect to open market repurchase plans, a study quoted in the publication
mentioned above indicated that fixed price tender offer programs accounting for 10% or more of a company’s
outstanding shares generated a 8.2% three day excess return, while those accounting for less than 10% only
8
174
The General Motors Corporation (D)
299-009
Alternative repurchase structures
In conjunction with the three share repurchase methods outlined above, there were a number
of other ways GM could complete a share repurchase program and generate tax and/or accounting
benefits not available under traditional repurchase methods. Exhibit 8 provides a description of each
method.
Put writing While put options were not considered a substitute for a share repurchase program, a
number of banks had proposed that GM consider writing puts to supplement its repurchase program.
A put option gives the holder the right to sell stock at a fixed exercise price. As the writer (or seller)
of the puts, GM would give others the right to sell GM stock to the company at a fixed exercise price.
GM could write put options equivalent to a portion of the number of shares it intended to repurchase
and receive the premium for the sale of these options. Over time, GM would carry out open market
repurchase program as outlined above (at currently prevailing market prices). Upon expiration of the
put options, if the options were out of the money (i.e., GM stock rose in price above the exercise
price), they would not be exercised, and GM would continue with the open-market repurchase. If,
however, the options were in the money, they would be exercised, and the put holders would deliver
the shares to GM at the exercise price of the puts, and GM would buy fewer shares in the open
market. In either case, GM would still receive the up-front, tax-free premium from the put options it
sold (Exhibit 9 and 10 provide financial markets data as of the time of the case).
Bankers suggested that the potential advantages of this program would be the overall reduction in
cost for repurchasing the stock which resulted from the put premium GM received, the tax-free
nature of the put premium, and the signal that purchasing puts provides (implying that GM expects
its stock price to rise). The drawbacks include the placement of a floor on GM’s repurchase price, the
increase in short positions in GM’s stock that result from the put buyer’s efforts to dynamically hedge
their put positions, and the potential additional disclosure problems related to the limitation that GM
cannot be in possession of any material inside information when entering into the put contract.
In conjunction with writing puts on its stock, GM could also buy call options and create a collar.
With a cashless collar, GM could use the premium from the written puts to fund the purchase of the
call options, with no net inflow or outflow at the initiation of the contract.14 By using a collar
position, GM would effectively lock in a price range within which they could repurchase stock.
Accelerated share repurchases A second derivative-based repurchase enhancement was what
was referred to as an Accelerated Share Repurchase program (ASR). This type of program was
intended to allow a company to purchase and retire a targeted number of shares immediately with
the final price of the shares determined by the market price of the shares over a fixed period of time
in the future.
To execute such a program GM would enter into an agreement with an investment bank under which
the bank would borrow the company’s shares from investors and sell them to GM immediately at the
current market price. The bank would then buyback the shares in the market over a pre-specified
period of time at the then-current market prices and return the borrowed shares to the lenders. At the
end of the period, the bank would then charge or pay GM the difference between the average
generated a 0.8% excess return. Dutch-auction style programs appeared less size sensitive, with programs
accounting for 10% or more of a company’s outstanding shares generating a 7.4% three day excess return, while
those accounting for less than 10% only generated a 6.2% excess return.
14 Additionally, it is possible to synthetically create a forward contract, or a commitment to repurchase the
shares at a fixed price at a certain point in the future, by selling puts and buying an equal amount of calls with
the same expiration date and strike price.
9
175
299-009
The General Motors Corporation (D)
purchase price of the shares and the initial price paid by GM, so that GM’s ultimate price for the
shares would closely match the average market price of the shares over the period.
Bankers stated that a primary benefit of this transaction would be that GM would receive the EPS
accounting benefit of an immediate share repurchase without the pressure of being forced to
complete the repurchase immediately. ASRs disconnect the timing of the accounting impact of a
share repurchase from the market implementation, and outsource the execution of the repurchase
program to the bank. However, the ASR constituted an irrevocable commitment to purchase the
stock; if the stock price appreciated rapidly, the company would not be able to terminate or
decelerate the program as it could a traditional open market purchase program. Lastly, an ASR
program would result in an increase in reported short interest as the investment bank would have to
borrow shares at the beginning of the program.
The Decisions
The range of decisions available to GM was quite large. Finnegan realized that it would be
critical not only to consider the simple dollars and cents of each choice, but to consider the impact
each might have on the perceptions of investors and analysts. In particular, Finnegan was acutely
aware of the benefits of consistency and stability; analysts and investors looked for patterns and
trends in a company’s actions. For example, if the company had declared dividend increases every 3
to 5 quarters for a number of years, not declaring one could prove to be as powerful a signal as cutting
a dividend was otherwise. Finnegan was eager to build what he considered a “consistent track
record” with the investment community.
Because the Board of Directors would have the ultimate decision to approve the
recommendations, understanding the group’s dynamics would be important as well. A number of
Directors had not been present during the financial crisis of the early 1990’s and the Board was
divided. Some felt that erring on the conservative side by holding the excess cash was appropriate,
while others questioned the need for the planned $13 billion in cash holdings. One thing they had in
common, however, was an appreciation of the visibility their decisions would have both inside and
outside the company.
10
176
The General Motors Corporation (D)
299-009
GM financial results 1992-1996 —data in $ billions unless otherwise noted
Exhibit 1
Financial Results
1992
1993
1994
1995
1996
U.S. market size - millions of units
15.0
16.0
17.3
16.5
17.0
U.S. GM unit sales - millions of units
5.0
5.2
5.6
5.3
5.3
33.1%
32.5%
32.3%
32.3%
31.0%
U.S. share of market - %
Worldwide revenues
132.2
138.2
148.5
160.3
164.1
Cost of goods sold - % of revenues
79.6%
77.0%
76.5%
75.7%
75.5%
Research & development spending
5.9
6.0
7.0
8.2
8.9
Pre-tax income
(3.3)
2.6
7.1
8.3
6.7
Net income (before extraordinary items)
(2.6)
2.5
4.9
6.0
5.0
Net income (after extraordinary items)
(23.5)
2.5
4.9
6.9
5.0
Total cash flow from operations
Capital expenditures
Total cash flow from investing activities
Net increase/(decrease) in debt
1
9.8
14.7
11.1
16.5
18.7
6.6
6.5
7.2
10.1
9.9
1.8
0.4
(16.2)
(22.1)
(17.7)
0.4
(0.4)
(0.2)
(1.9)
1.1
(21.1)
(0.6)
7.2
10.5
0.1
1.4
1.1
1.1
1.3
1.5
(6.9)
(12.5)
2.0
5.3
2.6
8.0
10.5
11.0
10.2
17.0
1992
1993
1994
1995
1996
Long-term debt/equity
110.1%
111.1%
47.4%
17.6%
22.2%
Total debt/total capital
55.6%
58.3%
36.0%
21.7%
22.0%
EBIT/interest expense
-1.4
1.8
6.1
19.3
6.2
BBB+
BBB+
A-
Net increase/(decrease) in equity
1
Cash dividend payments
Total cash flow from financing activities
Year-end cash balance
Leverage Measures
2
1
Senior bond rating
ABBB+
Source: GM financial statements
1
Data reported with GMAC accounted for on an equity basis.
2 Includes cash and marketable securities
1
Exhibit 2 Monthly stock and index prices scaled relative to GM’s price as of December, 1992
$70
$70
$60
$60
$50
$50
$40
$40
$30
$30
GM
$20
$20
Ford
Chr y sler
$10
$10
S&P 500 index
1
Dec-96
Sep-96
Jun-96
Mar-96
Dec-95
Sep-95
Jun-95
Mar-95
Dec-94
Jun-94
Sep-94
Mar-94
Dec-93
Sep-93
Jun-93
Mar-93
$0
Dec-92
$0
As of December 1996, GM’s common stock price was $52.59 per share.
11
177
299-009
The General Motors Corporation (D)
Quarterly dividends of the major U.S. automakers
Exhibit 3a
Recesssion
Rebuilding
$0.80
GM
$0.70
Ford
$0.60
Chry sler
$0.50
$0.40
$0.30
$0.20
$0.10
Exhibit 3b
1996:1
1995:2
1994:3
1993:4
1993:1
1992:2
1991:3
1990:4
1990:1
1989:2
1988:3
1987:4
1987:1
$0.00
Dividend yields of the major U.S. automakers
Recesssion
14.0%
Rebuilding
GM
12.0%
Ford
Chr y sler
10.0%
8.0%
6.0%
4.0%
2.0%
Exhibit 4
Event
4/28/95
2/2/96
Dividend
increase
3 weeks following announcement
Change in
GM Price
Change in
S&P 500
Excess
1
Returns
Change in
GM Price
Change in
S&P 500
Excess
1
Returns
1.0%
1.7%
-0.7%
-1.4%
3.8%
-5.2%
7.6%
3.5%
+4.1%
Dividend
2.3%
0.7%
+1.6%
increase
Source: Calculated by casewriters using data from Datastream
1
1996:1
Impact of dividend increases on GM stock price
2 days following announcement
Date
1995:2
1994:3
1993:4
1993:1
1992:2
1991:3
1990:4
1990:1
1989:2
1988:3
1987:4
1987:1
0.0%
Based on GM’s equity beta versus the S&P 500 index over the prior 250 day period.
12
178
The General Motors Corporation (D)
299-009
GM capital projections 1996-2000 (as of 1996)
Exhibit 5
$ Billions
1996
1997
1998-2000
17.0
18.9
28.0
13.0
13.0
13.0
4.0
5.9
15.0
4.0
1.9
9.1
0.0
(0.5)
(1.5)
(0.5)
(0.5)
(1.5)
0.0
0.0
0.0
Cash Forecast
Ending cash balance
Cash reserve required
2
Ending excess cash
Forecast Excess Cash Generation
1
3
Potential Uses of Excess Cash
Growth/Globalization Spending
Other contingencies
Additional Pension Funding
Adjusted Excess Cash Generation
3.5
0.9
6.1
Source: Casewriters’ estimates based on historical trends and economic and auto industry forecasts
1
2
3
Approximate year end balances in 1998 and 1999 were estimated to be $22 billion and $25 billion,
respectively.
$13 billion was the company’s previously-stated cash balance target, and was based on the amount
of cash GM was expected to need to weather a significant downturn in automotive sales.
Excess cash generation between 1998 and 2000 was estimated at approximately $3.0 billion per year.
Effects of cyclical auto sales on GM's profitability and cash flow
Exhibit 6
$20,000
20,000
18,000
16,000
$15,000
14,000
12,000
10,000
$ millions
8,000
6,000
$5,000
4,000
2,000
$0
Units (,000's)
$10,000
0
1989
1990
1991
1992
1993
1994
1995
1996
-2,000
-4,000
-$5,000
-6,000
-8,000
Unit sales
GM Pre-tax income
-$10,000
GM Operating Cash Flow s
-10,000
13
179
299-009
The General Motors Corporation (D)
Results of academic studies on the stock price impact of stock repurchases by U.S. firms
Exhibit 7
Sample Size
Time Period
Excess Return
Open Market Repurchases
Vermaelen
243
1970-1978
3.4%
1239
1980-1990
3.5%
Comment and Jarrell
1197
1985-1988
2.3%
Salomon Brothers, Inc.
1794
1991-1996
1.9%
Ikenberry, Lakonishok, and Vermaelen
1
Tender Offers
Vermaelen
131
1962-1977
14.1%
Dann
143
1962-1976
15.4%
Masulis
199
1963-1978
16.9%
Fixed-price
68
1984-1989
11.0%
Dutch-auction
64
1984-1989
7.9%
129
1989-1996
3.7%
77
1989-1996
6.7%
Comment and Jarrell
Salomon Brothers, Inc.
Fixed-price
Dutch-auction
Source:
1
Chris Innes, Peter B. Blanton, Nomo Nomo-Ongolo, and Cindy Sieden, “Stock
Buybacks – Strategy and Tactics,” Salomon Brothers, February 1997. These
figures represent the market reaction over a relatively short time interval (i.e., two
or three days) to announcements of repurchase programs .
The authors also document an additional average 12.6% excess return over the four years
following the announcement date of an open market repurchase program.
14
180
181
Source: GM documents.
• Company must issue a press
release announcing the size of the
repurchase program but does not
announce any individual
repurchases
• Not all shareholders have an
equal chance to participate.
Holders of large blocks are usually
targeted
• In practice virtually unlimited for
liquid stocks
• Daily limit: 25% of the average
daily trading volume for the last
four weeks plus any blocks of
5,000 shares or more
• Market price at time of repurchase
(unknown in advance)
• Skillful broker execution could
lead to average repurchase prices
below market prices (as much as
5%)
• Company is exposed to the risk of
a rising stock price
• Per share commission (roughly
$0.02 - $0.04/share)
Open Market Repurchase
• A tender offer document must be
distributed to all shareholders, and
notification must be made in the
press
• Offer must remain open for at
least 20 business days and for 10
business days if terms are
modified
• Generally done in conjunction with
open market activities, and a
company’s share repurchase
announcement is sufficient
disclosure
• If put repurchase obligation
exceeds 3% of shareholders’
equity, additional disclosure may
be required
• Purchaser of put warrants is a
large financial institution
• Purchaser of put warrants buys
stock in a manner similar to open
market repurchase program
• If the puts are exercised, net
repurchase price equals the strike
price less upfront put premium
• If the puts expire out of the
money, net repurchase price
equals the open market
repurchase price less upfront put
premium received
• Minimal legal fees, if any
• No execution costs
• Company receives tax-free cash
premium for selling put warrants
• Fixed Price: Fixed, predetermined price at a premium to
market (as high as 15%)
• Dutch Auction: A set price within
a specified range. Determined by
bid at a premium to market
(typically 5%-10%)
• Broker/Dealer fees
• Legal, printing, advertising,
depository and information agent
fees
• Roughly 1.0% total
• Fixed Price: Every shareholder
has an equal and pro rata chance
to participate
• Dutch Auction: Every shareholder
has an equal chance to participate
• In practice, a company may have
aggregate puts outstanding equal
to 5-10 days of trading volume
Sale of Put Options
• Not limited by regulation
• Typically at least 10% of
outstanding shares
Self-Tender
(Fixed-Price/Dutch Auction)
Overview of share repurchase program alternatives
SEC
Requirements
and Disclosure
Participation
Costs
Share
Repurchase
Price
Size of
Program
Exhibit 8
299-009
• Bank’s hedging activity is
generally targeted at non-block,
“retail size” trades
• Company may purchase
additional blocks of stock through
the bank during the ASR
• Generally, a company’s share
repurchase announcement is
sufficient disclosure
• If settlement amount at maturity is
likely to exceed 3% of
shareholders’ equity, additional
disclosure may be required
• Per share commission (roughly
$0.02 - $0.04/share)
• Final price is equal to the average
market price over a fixed period of
time
• Company is exposed to the risk of
a rising stock price
• Size limited by bank’s ability to
borrow shares at the
commencement of an ASR
Accelerated Share Repurchase
15
The General Motors Corporation (D) -15-
299-009
The General Motors Corporation (D)
Historical volatility of GM common equity returns — June 1990 - December 1996
1
45%
40%
35%
30%
25%
20%
15%
10%
5%
Source: Casewriter calculations based on data from Datastream
1
Based on the annualized standard deviation of daily returns over the prior 100 day period. As of
December 31, 1996, GM’s prior 100 day volatility was 21.4%
Exhibit 9b
Implied volatility of GM common equity returns — March 1994 - December 1996
40%
Annual Volatility
35%
30%
25%
20%
15%
10%
5%
11/29/96
9/30/96
7/31/96
5/31/96
3/29/96
1/31/96
11/30/95
9/29/95
7/31/95
5/31/95
3/31/95
1/31/95
11/30/94
9/30/94
7/29/94
5/31/94
3/31/94
0%
Source: Compiled from Bloomberg which reports implied volatilities from call options using either
the Black-Scholes or binomial model. As of December 31, 1996, implied volatility on call options was
22.6%.
16
182
12/31/96
8/13/96
3/26/96
11/7/95
6/20/95
1/31/95
9/13/94
4/26/94
12/7/93
7/20/93
3/2/93
10/13/92
5/26/92
1/7/92
8/20/91
4/2/91
11/13/90
0%
6/26/90
Annualized Standard Deviation of Daily
Returns
Exhibit 9a
The General Motors Corporation (D)
Exhibit 10
299-009
Term structure of U.S. Treasury STRIPS - December 1996
Maturity
Bond Equivalent Yield
3 month
5.31%
6 month
5.36%
1 year
5.63%
2 year
5.89%
3 year
6.04%
5 year
6.18%
10 year
6.49%
20 year
6.74%
30 year
6.69%
Source: Bloomberg
17
183