Competition,
Strategy, and
Business Performance
Anita M. McGahan
etween 1981 and 1997, the financial-market premium on publicly
traded U.S. corporations rose dramatically. At the same time, accounting profitability dropped steadily. Figures 1 and 2 show the trends. In
Figure 1, the financial-market premium is the ratio of the value of all
financial claims on a firm (including stocks and bonds) to the replacement value
of the firm’s assets.1 In Figure 2, accounting profitability is the ratio of after-tax
income to the book value of assets.
B
Figures 1 and 2 are surprising because financial-market premiums and
accounting profitability do not move together. If the two performance measures
followed similar trends, then the figures would suggest an alignment between
investor expectations and actual returns. Although the divergence in the trends
is surprising, there are several possible explanations. One is that the average
financial-market premium rose even as profitability fell because investors had
more money to invest. As the aggregate supply of funds increased, the financialmarket values of firms (relative to their asset bases) increased as well. In this
environment, the added supply of financial capital also may have encouraged
corporations to expand into markets with lower returns, and thus average
accounting profitability declined. A different potential explanation attributes the
increased financial-market premium to fundamental improvements in investor
expectations, perhaps because of breakthroughs in underlying technology. As
Special thanks to Michael E. Porter for discussions and for co-authorship of many of the articles in this
research stream. I am grateful to Carliss Baldwin, Barbara Feinberg, Benson P. Shapiro, the editor, two
anonymous referees, and seminar participants at the Harvard Business School for comments and
suggestions.Thanks to Rebecca Evans for research support.The Division of Research at the Harvard
Business School provided generous financial support for this project.
74
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
U.S. Corporations
25
20
15
10
5
0
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Ratio of Market Value
FIGURE 2. Accounting Probability of
U.S. Corporations
14
12
10
8
6
4
2
0
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Ironically, many executives
feel more pressure to improve
operating results even as financialmarket premiums increase. The
pressure may occur with increased
threat of takeover, more active
boards of directors, or increased
competitive threats. Can strategies
be improved constantly? Management consultants, business academics, investment analysts, and the
business press have all offered
executives a range of theories on
this question. Companies spend
hundreds of thousands and sometimes millions of dollars to get current advice from experts on the
newest approaches to improving
rates of return. They often advise
integrating core capabilities, subcontracting unessential activities,
divesting unrelated businesses, and
consolidating through merger with
rivals.
FIGURE 1. Financial-Market Premium of
Value of Returns
companies pursued the new projects, their accounting profitability
was temporarily depressed.
Yet despite all the advice,
Sources: Developed from data in the Compustat Basic File screened for
anomalies and for matching with the Compustat Business-Segment Reports.
executives with operating responsiThe screening approach is described in Anita M. McGahan,“The
bilities have not had much inforPerformance of U.S. Corporations: 1981-1994,” Harvard Business School
mation that can help them set goals
Working Paper, July 1998.
and frame strategic problems. Little
data is available for identifying how
the performance history of a business compares with trends in the economy. As
a consequence, it is difficult to know when a problem requires a radical solution.
When is a small operating problem an indication of deeper strategic issues, and
when is it temporary? How often should strategy be reformulated? How common are turnarounds? How long should a high performer expect to sustain its
profitability? Is it unusual for a company to show low profitability for a long
time? Operating executives often must rely on their experience rather than hard
data to make judgments about the magnitude and scope of strategic problems.
This article seeks to remedy this problem by providing broad information on the
characteristics of businesses with different kinds of performance histories. It
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
75
Competition, Strategy, and Business Performance
offers benchmarks to help executives form judgments about strategic challenges
and opportunities.
This study relies on recently collected data on the profitability of publicly
traded U.S. firms within specific businesses. The data has been compiled since
about 1980 by Compustat, when a new reporting requirement for businessspecific operating profit, sales, and assets went into effect. The dataset has several limitations. First, private companies are missing. Second, international
companies are excluded if they are not represented on U.S. exchanges. Third,
the industry definitions, given by the U.S. Standard Industry Classification (SIC)
System, are broad in some sectors and narrow in others. Fourth, the reported
information on a company’s businesses may not conform to its strategic business
units. Companies are required to report on businesses that constitute at least
10% of sales. Many firms often group together several similar operating units for
reporting. On average, each business reported by Compustat probably represents
two or three different operating units.2 For this reason, the reporting entities are
called “business segments” instead of “business units.”
Although the dataset suffers these problems, it has the substantial advantage of representing the first source of comprehensive information on the profitability by industry of both diversified and single-business public corporations.
Over the past few years, Professor Michael E. Porter and I have used the data
in several research studies to identify fundamental facts about business performance over time in a broad range of economic sectors. The studies were motivated by a sense that the new data provided unprecedented opportunities to
understand the relevance of industry, corporate-parent, and business-specific
influences on performance.3
Patterns in Business Performance
Strategic goals for performance are expressed in multiple ways: market
share, revenue growth, earnings per share, dividend growth, and operating
returns are all common measures. Is there a single correct measure of performance? Surely not. Despite the diverse possibilities, this study relies on a single
measure, the ratio of operating income to assets (i.e., “accounting profitability”).
This measure represents a return on invested capital and has the advantage of
capturing the broad operating qualities of a business. Revenue growth and market share are important drivers of operating income. The requirements to compensate equity and debt claimants are reflected in assets. The ratio of operating
income to assets is also one of the only available measures of performance at the
business level.
This study uses the financial-market premium as a supplementary
measure of performance.4 It reflects investor expectations about a company’s
prospects for generating value in the future. The financial-market premium is
greater than one when investors expect the company to generate more value
from its booked assets than if the assets were sold off. Whereas accounting
76
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
profitability reflects historic and current management influence, the financialmarket premium reflects investor expectations about future operating returns.
However, the financial-market premium has several disadvantages as a measure
of operating performance. First, it is only available at the corporate level and not
at the business level. Second, it can fluctuate with shifts in investor expectations
that are not fundamentally related to the operations of the business. For these
reasons, the classification scheme in this study relies on accounting profitability
rather than the financial-market premium.
The Compustat Business-Segment Reports, the primary source of information, were screened to eliminate records associated with missing or meaningless industry identifiers. Records were eliminated when assets or sales are less
than $10 million because they often do not represent operating businesses. Single-year appearances were excluded for the same reason. For the financial-market analyses, companies with less than $50 million in assets were eliminated
because their financial paper may not be traded frequently. Both of these
screens conform to precedents in the literature.5 Financial institutions were
also screened out of the dataset because their operating characteristics are quite
different than those of other types of businesses. Because business units are
aggregated in the reports submitted by firms, the dataset is not well suited for
analysis of entry and exit from the economy.
The screened dataset covers about 60% of the non-financial corporate
assets reported to the Internal Revenue Service over the 1981-1997 period. It
includes 13,547 business segments in 8,018 corporations and in 664 industries.
The average profitability across all businesses over the 1981-1997 period is
9.21%.6 About half of the observations are associated with diversified firms.
(Here, the term “diversified” is used to describe corporations that participate in
more than one business in the screened dataset.) The average business segment
posts $883 million in assets. On average, diversified corporations post $2,688
million in assets, and single-business firms post $791 in assets. (Thus, singlebusiness firms have smaller business segments on average than diversified
corporations.)
The first step in the analysis is the classification of businesses by their
performance over the 1981 to 1997 period.7 It would be too cumbersome to
represent all possible paths of profitability among businesses over 17 years,
especially since the average business is represented in the dataset for 10.5 years.
These complications are handled by classifying businesses based on their profitability ranking during the first four and last four years in which they are represented in the dataset. This approach has the virtue of simplicity. A comparison
with more complex classification methods suggests that the biases in the simple
classification scheme are not severe, and the relationships across categories in
profitability, size, relative growth, financial-market premiums, and financialmarket values are consistent.8
Figure 3 shows the categories. The left-hand-side of the matrix represents
average accounting profitability among businesses in the first four years that
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
77
Competition, Strategy, and Business Performance
High (25%)
Ending Performance
Medium (50%)
Low (25%)
High
(25%)
Sustained High
Performers
Declining High
Performers
Fallen High
Performers
Medium
(50%)
Rising Moderate
Performers
Steady Moderate
Performers
Declining Moderate
Performers
Turnarounds
Rising
Underperformers
Chronic
Underperformers
15.95%
3.89%
Low
(25%)
14.80%
2.88%
they appear in the dataset.9 The top of the figure represents average accounting
profitability among businesses in the last four years in which they appear.10 The
lines show the cutoff thresholds.11 The thresholds were selected so that the top
25% of businesses by rank in profitability would qualify as “high” performers,
the bottom 25% would qualify as “low” performers, with the rest qualifying as
“medium” performers. Each label in each box describes the trend in the relevant
range. For example, the upper left box, labeled “Sustained High Performers,”
represents businesses that both began and ended with high profitability.
Figure 4 shows that the number of businesses was not evenly distributed
by category. Only 74 (or 0.5%) of the 13,547 businesses are “Turnarounds,”
FIGURE 4. Number of Businesses by Category
Ending Performance
Beginning
Performance
Beginning
Performance
FIGURE 3. Performance Categories
78
Sustained High
Performers
2,628 (19.4%)
Declining High
Performers
599 (4.4%)
Fallen High
Performers
160 (1.2%)
Rising Moderate
Performers
685 (5.1%)
Steady Moderate
Performers
5,517 (40.7%)
Declining Moderate
Performers
572 (4.2%)
Turnarounds
74 (0.5%)
Rising Underperformers
658 (4.9%)
Chronic Underperformers
2,654 (19.6%)
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
TABLE 1. Sustainability
Beginning
Performance
Ending Performance
% with
same rank
% with
higher rank
% with
lower rank
High
77.6%
N/A
22.4%
Moderate
81.4%
10.2%
8.4%
Low
78.4%
21.6%
N/A
beginning with low performance and ending with high performance.12 “Fallen
High Performers,” businesses that dropped from high to low profitability, are also
relatively rare and account for just 1.2% of the total.13 The “Steady Moderate
Performer” category is the most populated with 40.7% of the businesses.14
Implicit in Figure 4 is information about the sustainability of high,
medium, and low profitability. Of the 25% of businesses that started with high
performance (top row), the figure shows that most of them (19.4%) sustained
their high performance over the period. Table 1, which is derived from Figure 4,
makes this explicit. It shows that 77.6% of the businesses that began with high
performance also ended with high performance. The rest ended with a lower
rank in profitability. Moderate performance was sustained at an even higher rate
of 81.4%.
Indeed, all types of performance—high, medium, and low—were quite
persistent. Surprisingly, low profitability was slightly more persistent than high
performance. Moderate performance was more persistent than any other type.15
A question arises regarding the rate of entry and exit from the various
categories. Were businesses that began with low performance more likely to exit
than those that began with high or medium performance? As indicated earlier,
the dataset is not well suited for analyzing questions of economic entry and
exit.16 Exit from the dataset occurs with bankruptcy, merger, privatization, and a
change in industry identifier. Entry into the dataset occurs with an initial public
offering, de novo entry by an established firm, and with a change in industry
identifier. During the 1980s and 1990s, corporate reorganization frequently led
to changes in identifiers. As a result, the average business segment posted 10.5
years of data between 1981 and 1997.
How much lower is the profitability of Chronic Underperformers than of
those in other categories? Table 2 shows an average ratio of operating income to
assets of –7 .2%.17 Of course, Sustained High Performers, with a 25.9% average,
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
79
Competition, Strategy, and Business Performance
TABLE 2. Average Profitability by Category
Beginning
Performance
Category
Sustained High
Performers
Aver- Average
Ending
age
Size in
Perfor- ProfitaAssets
mance
bility ($ mil.)
Median
Average
Average
Size in Financial- FinancialAssets
Market
Market
($ mil.) Premium
Value
High
High
25.9
459
92
1.64
974
High
Med.
15.2
853
143
1.31
1,303
Fallen High
Performers
High
Low
7.9
223
73
1.27
318
Rising Moderate
Performers
Med.
High
15.1
677
142
1.39
1,028
Steady Moderate
Performers
Med.
Med.
9.2
967
162
1.14
1,223
Declining Moderate
Performers
Med.
Low
2.3
364
75
1.01
427
Turnarounds
Low
High
10.7
253
77
1.32
474
Rising
Underperformers
Low
Med.
3.6
867
126
1.11
1,046
Chronic
Underperformers
Low
Low
-7.2
419
62
1.28
485
Declining High
Performers
show the highest average profitability. The statistics in Table 2 are developed
from means taken over the entire period in which businesses are represented.
Thus, Declining High Performers and Rising Moderate Performers also show
relatively high average profitability. Declining Moderate Performers are secondlowest in rank.
Table 2 also shows the averages sizes in assets of businesses by category.
Categories with high average sizes are highlighted. Steady Moderate Performers
are over twice as large as both Sustained High Performers and Chronic Underperformers. The next column of Table 2 shows median size in assets. (The median
is defined so that half of observations fall below and half fall above the value.)
For all categories, the median size is significantly less than the average, although
the relationships between categories are similar. This means that the distribution
of businesses by size is strongly skewed. In all categories there are a lot of small
businesses and relatively few very large businesses. The very large businesses
draw up the averages. In the discussion of core examples, this issue is raised
again. Overall, the information on size by category suggests that high performers
become moderate performers by growing large, perhaps compromising their
original operating strategies. It also suggests that low performance occurs if a
high or moderate performer does not grow enough to achieve a minimum efficient scale.
80
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
FIGURE 5. Average Growth Rates by Category
Beginning
Performance
Ending Performance
Sustained High
Performers
15.3% (10.0%)
Declining High
Performers
13.7% (8.5%)
Fallen High
Performers
10.4% (4.5%)
Rising Moderate
Performers
14.7% (11.4%)
Steady Moderate
Performers
15.1% (8.9%)
Declining Moderate
Performers
8.5% (2.4%)
Turnarounds
15.5% (9.9%)
Rising
Underperformers
16.3% (19.6%)
Chronic
Underperformers
16.9% (22.4%)
Note: Simple averages with weighted averages in parentheses.The weights are the inverses of the variances of the estimates.
The last two columns of Table 2 show the average financial-market premium and financial-market value of businesses in each category.18 The financialmarket value represents the total amount of investors’ claims on a business in
absolute terms.19 The financial-market value is lower for Sustained High Performers than for Steady Moderate Performers. This occurs because Steady Moderate Performers are so large on average. Their size overwhelms the influence
of the higher financial-market premium on Sustained High Performers. The
regularity carries an important implication for strategy: Steady Moderate Performance may be a worthy objective in its own right if it creates more financialmarket value than Sustained High Performance.
Although a detailed analysis of the financial-market premium goes
beyond the scope of the study, several of the estimates are striking. The average
financial-market premium is greatest for Sustained High Performers, with Rising
Moderate Performers and Turnarounds distant followers. In general, improving
performers had a higher financial-market premium than declining performers,
which suggests that investor expectations about the future were related to the
trajectory of accounting profitability. Chronic Underperformers had a relatively
high financial-market premium despite their low accounting profitability.
(Recall that over 19% of the businesses in the dataset qualify for the category.)
This result means that, on average, investors expected better performance from
Chronic Underperformers than from Declining Moderate Performers. Perhaps
investors viewed Chronic Underperformers as prospective turnarounds.
Figure 5 shows average revenue growth rates by category. The figure
shows both simple and weighted averages. The weights are the inverses of
the variances of the estimates. The simple averages differ from the weighted
averages because the businesses with the highest growth rates also have the
highest variances in growth over time.20 Although the simple averages are
higher than the weighted averages, the relationships across categories are
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
81
Competition, Strategy, and Business Performance
FIGURE 6. Sectors by Category
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Health & Bus. Svcs.
Lodging/Ent. & Pers. Svcs.
Wholesale/Retail Trade
Transportation
Manufacturing
Agriculture & Mining
similar. The highest average growth rates arose in categories in which the businesses ended with high performance. The lowest average growth rates arose in
categories where businesses ended with low performance. The major exception
is Chronic Underperformers.
Figure 6 shows how businesses within each category were distributed
across broad sectors of the economy (e.g., agriculture and mining, manufacturing, transportation, wholesale/retail trade). There are some differences across
the categories: few Turnarounds arose in the wholesale/retail trade sector, and
transportation businesses were rarely Fallen High Performers. The agriculture,
mining, and manufacturing sectors were disproportionately represented among
Sustained and Declining High Performers, and were less common among Steady
Moderate Performers.
Table 3 shows additional information about the distribution of industries
across categories.21 This information shows whether Chronic Underperformers,
for example, were concentrated within just a few industries, or whether they
arose in many different industries. For example, the first row reports that 89.3%
of the 534 industries that contained Sustained High Performers held between
1 and 10 of them. Just 0.6% of the industries contained between 51 and100 of
the Sustained High Performers.
The results in Table 3 point to some important asymmetries that are not
apparent in Figure 6. Across categories, a high portion of represented industries
contain just 1-10 members. In particular, Turnarounds were not concentrated,
contrary to the hypothesis that they tended to be members of industries like
biotechnology/pharmaceuticals or software. Rather they were distributed across
a variety of agriculture, manufacturing, transportation, wholesale, retail, entertainment, and service industries. Overall, this pattern was common. Businesses
on all of the performance trajectories were widely distributed across industries.
82
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
TABLE 3. Distribution of Industries Across Categories
Count of Industries by No. of Members
No.
Bus.
No.
Ind.
1-10
11-25
26-50
51-100
100+
Sustained High
2,628
534
89.3%
9.0%
1.1%
0.6%
0.0%
Declining High
599
265
98.5%
1.1%
0.4%
0.0%
0.0%
Fallen High
160
98
100.0%
0.0%
0.0%
0.0%
0.0%
Rising Moderate
685
301
98.0%
1.7%
0.3%
0.0%
0.0%
Steady Moderate
5,517
609
78.7%
14.1%
5.1%
1.1%
1.0%
572
232
97.4%
2.2%
0.4%
0.0%
0.0%
Declining Moderate
Turnarounds
Improving Under’s
Chronic Under’s
74
57
100.0%
0.0%
0.0%
0.0%
0.0%
658
267
97.0%
2.6%
0.4%
0.0%
0.0%
2,654
465
87.7%
8.0%
3.7%
0.2%
0.4%
A few exceptions to this pattern are evident. Steady Moderate Performers were more concentrated by industry than businesses in any other category.
One percent of Steady Moderate Performer industries had more than one hundred members from the category. Of course, this regularity arises partly because
of the high number of businesses in the category and partly because of concentration of Steady Moderate Performers in just a few industries. Table 4 shows
that Steady Moderate Performers tended toward the transportation sector.
TABLE 4. Most Populated Industries Among Steady Moderate Performers
No.
Industry
Identifier
No. of Steady
Moderate
Performers
Total
Member
Businesses
Sector
Description
1311
Agriculture, Mining
Crude Petroleum
and Natural Gas
Extraction
180
488
4911
Transportation
Electric Services
155
168
4924
Transportation
Natural Gas
Production and
Distribution
131
145
5812
Wholesale/Retail Trade
Eating Places
106
242
4813
Transportation
Telephone
Communications
104
212
7372
Services
Prepackaged
Software
102
339
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
83
Competition, Strategy, and Business Performance
TABLE 5. Most Populated Industries Among Chronic Underperformers
No.
Industry
Identifier
No. of Chronic
Underperformers
Total
Member
Businesses
Sector
Description
1311
Agriculture, Mining
Crude Petroleum and
Natural Gas Extraction
121
488
7372
Services
Prepackaged Software
101
339
2834
Agriculture, Mining*
Pharmaceutical
Preparations
57
224
* Pharmaceutical Preparations are classified in “Agriculture and Mining” because they have an SIC in the 2000’s.The classification may reflect
historical links between pharmaceuticals and agricultural compounds.This study relies on the convention that agriculture and mining businesses
have SIC’s less than 3000, and manufacturing businesses have SIC’s in the 3000’s. Some other studies assign all businesses with SIC’s less than
4000 to “agriculture and manufacturing,” or SIC’s between 2000 and 4000 to “manufacturing.” Thus, care is needed in comparing the results with
those from other studies.
Chronic Underperformers were also relatively concentrated, with 0.4% of
industries hosting more than 100 members. One industry contained 57 members. Table 5 shows the three most populated industries.
The third most concentrated category was Sustained High Performers.
Although none had more than 100 members, 0.6% of industries had more than
50 members. Table 6 shows that they were the same industries as the most populated among Chronic Underperformers.
These results indicate important relationships between the categories. In
particular, they show that a subset of industries may be structured to support
both Sustained High Performers and Chronic Underperformers.22 The analyses of
growth rates, sizes, and financial-market premiums also suggests relationships
between categories as businesses evolved.
TABLE 6. Most Populated Industries Among Sustained High Performers
No.
Industry
Identifier
Sector
Description
2834
Agriculture, Mining
7372
1311
84
No. of
Sustained High
Performers
Total
Member
Businesses
Pharmaceutical
Preparations
85
224
Services
Prepackaged Software
70
339
Agriculture, Mining
Crude Petroleum and
Natural Gas Extraction
58
488
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
Core Examples of Competitive Strategy
This section offers additional information on the diversity of businesses
by category and examines “core examples” for each of the categories. Because
each category contains diverse businesses, no single example can represent the
full range of issues that arise in a category. The examples are selected because
their characteristics are near the averages for their categories and, in some cases,
because they are interesting in their own right. The discussion highlights only
some of the strategic challenges that can arise for businesses on each
performance trajectory.
Sustained High Performers
Most of the 2,628 Sustained High Performers were small business segments. Recall that the average Sustained High Performer posted profitability of
25.9% on assets of $459 million over the period. The average financial-market
premium for Sustained High Performers was higher than in any other category
at 1.64. A supplementary analysis indicates that on average Sustained High Performers belonged to attractive industries, were part of high-performing diversified firms, and had profitability well above the industry averages.23 Table 7 lists
several members of the category along with information on their performance,
size, growth, and financial-market premiums. Many are well known. Some are
household names.
Although Microsoft received great attention for its superior performance
in microcomputing software over the period, it was not typical of the category in
some important respects. Its average profitability, revenue growth, and financialmarket premium were all much greater than the averages for the category.
Microsoft also was not part of a diversified firm and was much larger in size than
average. The Gannett Corporation in the newspaper-publishing business is more
representative of the “core” of the category. Gannett’s profitability, financialmarket premium, and growth rate in newspaper publishing were near the averages for the category. Also, this business belonged to an attractive industry, was
part of a high-performing corporate parent, and posted profits above its industry
average. However, Gannett’s newspaper publishing business was somewhat
larger than the mean or median firm in the Sustained High Performer category.
The business is suggestive as a core example despite its large size because so
many of its other characteristics reflected the averages for the category.
What was Gannett’s business strategy over the period? During the 1980s
and early 1990s, Gannett participated in three industries: broadcasting, outdoor
advertising, and newspaper publishing. Over the entire period, Gannett
performed above the averages for its industries, perhaps because corporate management was particularly effective in managing its divisions.24 In newspaper
publishing, Gannett published USA Today and about 80 local papers with daily
and weekly circulation. Its principal national competitors pursued very different
strategies. For example, the Wall Street Journal concentrated on business news;
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
85
Competition, Strategy, and Business Performance
TABLE 7. Examples of Sustained High Performers
Average
Size in
Assets
($ mil.)
Average
Revenue
Growth
Average
FinancialMarket
Premium
31.7%
$9,260
16.2%
3.25
36.8
3,691
45.0
6.73
Average
Business
ProfitDescription
ability
Company
Industry
Identifier
Merck & Co.
2834
Human &
Animal Health
Microsoft Corp.
7372
Microcomputer
Software
PepsiCo Inc.
2096
Snack Foods
27.7
3,258
10.4
1.60
Kellogg Co.
2043
Food
Products
26.6
3,155
7.1
2.52
Home Depot
5211
Building Materials
15.9
2,942
49.4
3.77
Gannett Co.
2711
Newspaper
Publishing
26.4
2,207
8.4
1.91
Cisco
Systems Inc.
3576
Internetworking
Systems
40.1
1,452
100.7
7.94
Dell Computer
Corp.
3571
Personal
Computers
17.2
1,299
56.4
2.55
Tribune Co.
2711
Newspaper
Publishing
29.0
814
2.1
1.40
Armstrong
World Indust.
3296
Building
Products
18.8
469
5.7
1.02
Jostens Inc.
3911
MotivationRecognition
Products
21.8
392
9.7
2.06
Bandag Inc.
3011
Tread
Rubber &
Equipment
28.6
378
6.0
2.33
Houghton
Mifflin Co.
2731
Textbooks—
Educational
Material
27.1
277
10.5
1.41
Tootsie
Roll Inds.
2064
Candy
22.4
179
10.9
2.13
Boston
Beer Inc.
2082
Craft
Brewery—
Beers, Ale
16.5
78
18.1
3.36
Black Hills
Corp.
1221
Coal Mining
19.8
49
3.9
1.04
Drew
Industries
Inc.
3442
Manufactured
Housing,
Recreational
Vehicles
35.8
27
17.2
1.29
Mile High
Kennel Club
7948
Greyhound
Race Track
20.3
11
5.4
0.36
86
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
the Washington Post was known for political reporting; and the New York Times
covered national news, New York stories, and popular features. Because of broad
differentiation among the leading competitors, industry-average profits exceeded
the national average.
Gannett distinguished itself from other large newspaper companies
by focusing on the publication of local papers. It decentralized editorial control
but maintained centralized coverage of national and world news. Distributed
printing systems assured cost efficiency and timeliness of physical distribution.
Although USA Today was something of a financial disappointment, the company’s core newspaper operations in local markets continued to generate
impressive returns.25
This balancing act—sustaining high profitability within the industry
without damaging industry structure—required the careful execution of strategy.
Gannett’s continued high operating return on assets, the defining characteristic
of its category, attracted attention from customers, employees, and competitors
—all of which were interested in improving their own value capture. Apparently, however, Gannett’s tight operations network, local distribution, and journalistic expertise gave it an advantage that could not be replicated. The strategic
challenge associated with these kinds of specialized capabilities was growth:
How could Gannett build on its strength in newspaper publishing without
compromising its differentiation and without inciting retaliation from rivals?
Steady Moderate Performers
Steady Moderate Performers were much larger on average than Sustained
High Performers and Chronic Underperformers. With an average size of $967
million, Steady Moderate Performers were larger than businesses in any other
category. Steady Moderate Performance was also quite common (partly because
of the broad definition of “medium” profitability). Over 40% of the businesses in
the dataset qualified for the category. The large size of the average Steady Moderate Performer was associated with a high financial-market value despite the
lower financial-market premium than Sustained High Performers.
Several transportation businesses lead the list of industries most populated by Steady Moderate Performers. The largest business in the category was
Nippon Telegraph & Telephone in Telephone Communications (industry 4813)
with average profitability of 6.0% on average assets of $87 billion. The secondlargest business was AT&T in the same industry with average profitability of
12.7% on average assets of $54 million. This business was labeled “Information
Movement and Management” by AT&T from 1993 to 1995, and “Communications Services” from 1996 to 1997. Information was reported on the business for
only the five years from 1993 to 1997. Table 8 shows examples of other Steady
Moderate Performers with their characteristics.
Not surprisingly, the average Steady Moderate Builder posted profits very
near the norm for its industry. Overall, businesses in the category belonged to
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
87
Competition, Strategy, and Business Performance
TABLE 8. Examples of Steady Moderate Performers
Company
Industry
Identifier
Average
Business
ProfitDescription
ability
Average
Size in
Assets
($ mil.)
Average
Revenue
Growth
Average
FinancialMarket
Premium
Toyota Motors
Corp.
3711
Automobiles
8.0
47,017
15.9
0.95
BellSouth Corp.
4813
Telecommunications
11.9
30,105
6.2
0.98
British
Petroleum PLC
1311
On-Offshore
Exp.& Prodn
12.1
26,325
n/a
n/a
Federated
Department
Stores
5311
Department
Stores
7.4
12,419
24.6
1.03
HCA (Hospital
Co. of America)
8062
Hospital
Management
10.1
5,419
9.5
1.06
Federal Express
Corp.
4513
Express Delivery
11.5
3,760
21.6
1.07
Southwest Gas
Corp.
4924
Natural Gas
7.1
965
3.1
0.86
Starbucks Corp.
2095
Coffee Retailer
& Wholesaler
8.4
371
60.5
3.55
Stratus
Computer Inc.
3571
Computer
Systems
13.9
321
33.3
2.22
Carmike
Cinemas Inc.
7832
Motion Picture
Theaters
10.3
258
21.1
0.97
Summit Care
Corp.
8051
Nursing Care
Centers
10.9
127
24.0
1.26
Hamburger
Hamlet Rest.
5812
Restaurants
8.3
33
8.0
0.92
Spec’s Music Co.
5735
Retail Music
Stores
9.2
27
16.2
1.15
industries that showed profitability slightly above the average for the economy.
Consider Federal Express in express delivery and Starbucks in coffee retailing/
wholesaling as examples. These companies adopted well-tested business models
and continually renewed capabilities through internal development of critical
resources over time. Federal Express built its overnight delivery business by
centralizing all of its nightly sorting and redirecting functions at a Memphis facility. Starbucks became renowned by putting new stores in geographic clusters
around the country. As Starbucks, Federal Express, and other businesses built
interlocking systems of assets, their advantages became difficult to imitate.
The Steady Moderate Performers category contains a disproportionate
share of businesses in the wholesale trade, retail trade, and transportation
88
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
sectors. As noted earlier, the category contains more businesses concentrated in
specific industries than any other category. The most populated industry, Crude
Petroleum and Natural Gas Extraction (industry 1311), was also a leading industry among Sustained High Performers and Chronic Underperformers. Of the
thirteen industries with more than 50 Steady Moderate Performers, five were
in the transportation sector: electric services, natural gas production and distribution, telephone communications, natural gas transmission, and trucking
(except local). Two were in wholesale/retail trade: eating places and grocery
stores. Three of the industries were in agriculture and mining: operating
builders, paper mills, and petroleum refining. Two industries, prepackaged software and motor vehicle parts, were in services and manufacturing, respectively.
The sectoral composition may reflect underlying industry structures that
tend to support moderate performers. Federal Express and Starbucks did not
make large, preemptive commitments to establish their positions. Rather, they
invested incrementally to develop resources internally and to expand their geographic presence over time. In these situations, the central strategic challenge
lies in finding incremental growth opportunities that do not compromise the
underlying formula. Further research is needed to understand whether the challenges facing Federal Express and Starbucks apply to a wide range of businesses
in the category.
Chronic Underperformers
The “Chronic Underperformers” category contains businesses with long
and short histories: American Motors in general automotive, Bethlehem Steel in
steel-related operations, Lionel in toys and leisure products, McCaw Cellular in
cellular telephone service, Nova Pharmaceuticals in drug research and development, and THQ in video game software. On average, businesses in this category
posted profits significantly below their industry averages for a long period. This
suggests that the average Chronic Underperformer had strategic problems that
differed from those of direct rivals.
A significant minority engaged in a process of exit. Many exited only after
several years of investment in new assets, however. These businesses may have
started with an expectation of better performance, but eventually accepted that
the business would never recover the costs of ongoing investment. Exit often
took the form of consolidating mergers between members of the same industry,
or between a firm and one of its customers or suppliers.
What kinds of strategic problems confronted Chronic Underperformers?
Consider McCaw Cellular and Nova Pharmaceuticals as examples of newer
businesses in the category. Both of these businesses constructed major valuegenerating assets over a long period. The assets had to be fully developed
before products could be brought to market: McCaw had to build a cellular
network and consumer awareness, and Nova had to find a viable new pharmaceutical product. Their strategies involved years of low profit while resources
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
89
Competition, Strategy, and Business Performance
accumulated for future return. The firms bore the risk that rewards would not
materialize, or materialize later than anticipated.
Some older firms in the category, including Bethlehem Steel in steelrelated operations, faced more complicated strategic challenges. Bethlehem
Steel and others like it weathered great technical changes in their industries
but invested for the future by rationalizing assets. In these situations, established
customers and suppliers often had little incentive to encourage the adoption of
new ways of doing business. Older Chronic Underperformers had to invest in
the hope of either attracting new trading partners or compelling established
partners to change.
In an important sense, both the newer and the older Chronic Underperformers were prospective Turnarounds at the end of the period. The persistent
refusal to exit indicated an ongoing commitment and the implicit intent for a
recovery. Some of the firms made compelling cases that their turnarounds were
imminent. McCaw Cellular sold to AT&T in 1993 for an unprecedented $12.6
billion despite low accounting profitability.
The analysis raises a hypothesis that some industries were structured
to encourage long-term investments with differed reward. The most populated
industries among Chronic Underperformers were the same as among Sustained
High Performers: crude petroleum and natural gas extraction, prepacked software, and pharmaceutical preparations. Pharmaceuticals, for example, hosted
both established leaders and a range of low-performing public biotechnology
companies.
Declining High Performers
Unlike Microsoft in microcomputer software and Gannett in newspaper
publishing, some businesses that began as high performers subsequently experienced a decline in profit. About a quarter of the businesses that started with
high performance ended with a lower ranking. Most became Declining High
Performers with medium profitability. The “Declining High Performers” category
includes Whirlpool in major home appliances, Benihana in restaurants, Carnival
Corp. in cruise lines, Coca-Cola Bottling in soft drinks, Rite Aid in retail drug
stores, and Snap-On in tool manufacturing and distribution.
What went wrong for Declining High Performers? On average, they lost
distinctive performance within their industries, but their industry effects did not
deteriorate.26 The analysis shows that Declining High Performers started with
somewhat lower performance than Sustained High Performers in the first place.
Although industry effects eroded over time, the greatest decline by far occurred
in the business-specific effects.
Consider, for example, Benihana in restaurants. Benihana operated a
chain of Japanese restaurants, and appeared as an independent entity in the
dataset from 1983 to 1992. Restaurant industry profits exceeded the average
throughout the entire period. Initially, Benihana performed well above the
90
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
industry average. The restaurants were famous for an entertaining style of food
preparation in which chefs cooked food to order on grills that formed part of
patrons’ tables. By some accounts, Benihana is credited with originating the
“eatertainment” concept, which augmented the dining experience with entertainment.27 By 1992, Benihana’s business-specific effect had declined substantially, and profits dropped below the industry average. Customers apparently lost
interest in the novelty associated with Benihana’s approach and simply switched
to other dining and entertainment experiences. The main issue was not direct
imitation of Benihana’s teppanyaki cooking approach. Rather, customers wanted
novelty, and entrepreneurs responded by creating new entertainment restaurants using other approaches.28 For Benihana, the innovations that spawned
high performance had two lingering effects. The first was that the business continued to generate value for a core group of customers and suppliers and
retained medium profitability.29 The second was that the high profitability
of the business attracted competitive attention. Competent rivals discovered
ways to overcome customer and supplier switching costs, although not by imitating directly. Instead, they learned from the leader’s approach and extended
it sufficiently to lure away the leader’s customers and suppliers.30 The strategic
challenge was not in fending off direct imitators, but in competing against rivals
that extended and adapted Benihana’s approach.
Fallen High Performers
Businesses in the Fallen High Performers category include Boeing in
military transportation equipment, Fair Grounds Co. in horse racing, L.A. Gear
in shoes, and Sierra On-Line in entertainment software. For the average Fallen
High Performer, the industry, corporate, and business-specific effects all fell off
quickly and dramatically in about the same proportion. Profit typically declined
rapidly. On average, the businesses in this category were low performers just a
few years after posting high performance.
Consider L.A. Gear, a maker of high-end fashion sneakers and shoes.
L.A. Gear’s revenue in the shoe business increased from $36 million in 1986
to a phenomenal $902 million in 1990. The company’s products appealed to
teenagers and young adults interested in fashion shoewear. L.A. Gear’s customers had very low switching costs, however. In fact, the fashion quality of the
product may have generated negative switching costs: customers may have preferred a different brand just for the difference in fashion once they tried L.A.
Gear shoes.31 At a minimum, L.A. Gear shoes were unlike Benihana’s dining
experience in that there was no compelling reason for repeat purchases.
By the 1990s, competition from Converse, Reebok, and Nike greatly
eroded L.A. Gear’s sales and profitability. In 1994, L.A. Gear sales in the shoe
business were $416 million, less than half of the 1990 peak of $902 million.
The company did not unwind its investments at nearly the same rate, however.
In every year between 1991 and 1994, operating profitability was negative.
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
91
Competition, Strategy, and Business Performance
Several problems plagued L.A. Gear. The first was increasing price competition. Given low switching costs for customers and suppliers, competitors
attracted L.A. Gear’s customers with lower prices. The second problem was specialized assets that locked the businesses into position. L.A. Gear held significant
inventory and sold products at locations that were tightly packed geographically,
for example.32 As competition escalated, the company could not exit the industry without walking away from valuable assets.33
It is difficult to know whether these kinds of strategic challenges were
widespread across the category, but they probably did arise for a significant
group of Fallen High Performers. Fallen High Performers were small on average
with sales of $223 million. For a company like L.A. Gear, larger competitors had
greater bargaining power and access to scale economies. Although Fallen High
Performers were often small, they did not have a greater tendency to exit than
businesses in the other categories. In fact, the disinclination to exit may have
contributed to long-run fragmentation in the industries populated by these
businesses.
Declining Moderate Performers
The Declining Moderate Performers category includes businesses that
began with medium performance and ended with low performance. Digital
Equipment in computers, Pizza Inn in restaurants, Crown Books Corp. in retail
book stores, Everest & Jennings in medical equipment, Bowl America Inc. in
bowling centers, and Todd Shipyards in shipbuilding belong to this category. The
average industry effect on Declining Moderate Performers was initially slightly
positive but quickly became slightly negative. As industry structure shifted, the
average business-specific effect became significantly and persistently negative.
Case studies of some of the companies in the category reveal an interesting dynamic. Like the core examples of Steady Moderate Performers, some
Declining Moderate Performers had a history of investing gradually to achieve
medium profitability. Unlike the Steady Moderate Performers, however, the
Declining Moderate Performers became highly committed to positions that could
not be quickly adapted as customer tastes and supplier technologies shifted.34
For example, Everest & Jennings continued to manufacture parts for vinyl-covered, standard wheelchairs through the early 1990s because many hospitals and
nursing homes demanded them. Everest & Jennings’ customers—the hospitals
and nursing homes—had co-invested by building fleets of the older wheelchairs
that sometimes needed new parts. As the Invacare Corporation developed and
commercialized newer models of standard wheelchairs using different raw materials and designs, it gradually attracted customers from Everest & Jennings.35 For
Everest & Jennings, conflicting incentives arose. To serve established customers
and suppliers, the business continued to offer traditional standard wheelchairs.
Investment in new products may have appeared too risky or too difficult. As a
result of these conflicting incentives, the strategic problems facing Everest &
Jennings were particularly challenging.
92
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
There is some evidence that these challenges may have confronted other
Declining Moderate Performers as well. Many of the businesses in the category
engaged in significant restructuring and downsizing programs during the late
1980s and early 1990s. Perhaps as a result, the average Declining Moderate
Performer was small, with assets of $367 million, and grew slowly over the
1981 to 1997 period.
Turnarounds36
Some of the most celebrated businesses of the 1980s and 1990s began
with low performance and ended with high performance. The Turnaround
category includes Turner Broadcasting in news and the Sprint Fon Group in
long-distance communications service, each of which contributed to a revolution in its industry during the period. The category also contains companies
that received less attention, including the Ackerley Group in broadcasting,
Minnetonka in personal care products, and Pittston in home security systems.
Turnarounds typically began in slightly unattractive industries with businessspecific effects substantially below average. The poor business-specific effects
quickly dissipated, however.
Consider Turner Broadcasting in news. Early poor performance occurred
as the business invested for future return. After an initial period of low profitability, profits turned marginally positive, with a positive corporate-parent effect
associated with the improvement. After some years, the business showed profitability above the industry average. Once Turner Broadcasting’s news business
posted above-average profits, the industry average began to improve. The turnaround stimulated widespread customer interest in a new class of television
broadcasting services.37 Other broadcasters offered differentiated products and
services to meet the new demand.
Turner Broadcasting’s performance in the news business suggests a pattern that may arise for other Turnarounds as well. The new business model
required heavy initial investment in physical equipment, brand capital, and
other capabilities. Only after construction did the assets begin to generate value.
As a result, revenue growth accompanied low profitability for a significant
period. Because the business developed specialized capabilities before attracting
attention from rivals, it secured a position that would have been too costly to
imitate. Turner had secured advantages that became impossible to imitate during
the late 1980s and 1990s.38
The Turner Broadcasting example suggests that revenue growth tied more
closely to the achievement of high performance than to the sustainability of high
performance.39 The factors that made Turner in broadcasting a Turnaround—
early investment, followed by secure differentiation and high profitability—tied
directly to the pattern of revenue growth in the business. Turner Broadcasting’s
profitability turned sharply positive during the mid-1980s and remained high
through the 1990s.
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
93
Competition, Strategy, and Business Performance
What characterized Turner Broadcasting as a core example of a Turnaround? It invested for a long period despite outsiders’ skepticism about how
the investments would pay off in the market. By the time rewards materialized,
rivals could not imitate effectively. Customers and suppliers found the approach
compelling and quickly flocked to the leader. Followers faced the prospect of
long asset-development programs before they could even match the leader’s
approach, and they often chose differentiation instead. In many cases, the industry benefited as differentiated rivals learned from the innovation while the
leader preserved its position. These kinds of bold investments required strong
leadership.40 A visionary leader saw the payoff to asset development years before
it occurred.41
Rising Underperformers
The Rising Underperformers category contains many businesses that pioneered their industries, including International Paper in packaging, Houghton
Mifflin in general publishing, and Ford Motor in manufacturing autos and parts.
The accomplishments of these firms are especially impressive in light of both
their average sizes and average growth rates. Rising Underperformers were
larger on average than businesses in any other category except Steady Moderate
Performers and grew even more rapidly on average than Steady Moderate
Performers.
Early in the period, the old advantages of many of these leaders had
become disadvantages. These companies dealt with a large cadre of loyal customers and suppliers that were committed to the old business model, which
made change particularly difficult. By the mid-1990s, however, profits recovered
to moderate levels—the same category achieved by Toyota Motor, Federal
Express, and Starbucks. It would be a mistake to understate the accomplishments of the firms first in envisioning new opportunities and second in finding
a way to adapt old resources to new business models. For these Rising Underperformers, better profitability depended on working with customers and suppliers to help them make major transitions.
The Rising Underperformers category also contains some newer businesses with short histories of investment. For these businesses, prior investment
started to generate return by the mid-1990s. In many ways, the challenges faced
by the older and newer businesses were similar: Investment in durable resources
initially constrained profitability and then started to pay off. In other ways, the
challenges differed: many older firms operated on a much larger scale and faced
conflicting incentives to adapt, whereas newer firms operated from a clean slate.
Was there anything that clearly distinguished a Rising Underperformer
from a Turnaround? On average, Rising Underperformers began the period with
performance above that of Turnarounds. For Turnarounds, however, corporate
effects were associated with an early positive influence on profitability. In contrast, no corporate effect was associated with early investment for the average
Rising Underperformer. Industry-average profitability stayed low. By the end
94
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
of the period, however, industry effects turned slightly positive. The size and
history of the core examples of Rising Underperformers made steady, incremental improvement the priority. Value creation could not be deferred given the vast
asset base of these businesses.
Rising Moderate Performers
The Rising Moderate Performer category includes businesses that worked
up to high performance over time, including Gillette in appliances, Kroger in
food stores, Intel in integrated circuit components, and Sherwin-Williams in
paint stores. On average, Rising Moderate Performers started in attractive industries and belonged to corporate parents of above-average profitability.
Consider Kroger in food stores as an example. During the 1980s and
early 1990s, the food-store industry reaped the returns of early rationalization
and showed returns slightly above the economic average. Through impressive
asset management, Kroger consistently increased its profitability until it became
a high performer. Although sales increased steadily, assets in the business
remained at about the same level over time. As a result, operating income
moved evenly higher over time. Kroger generated high performance through
a strategy of careful investment in a system of tightly linked assets. For Kroger,
profits improved through enhanced efficiency in the delivery of value to customers. As a result, profits improved incrementally and reflected less basic risk
of failure.
In one important sense, Kroger was not typical of Rising Moderate
Performers: It did not belong to a diversified company. As with Turnarounds,
corporate effects were often associated with profit improvement among Rising
Moderate Performers.
The Enduring Logic of Competition
The analysis of both statistical features and of core examples suggests that
an enduring logic of competition influenced the achievement and sustainability
of performance between 1981 and 1997. In many cases, the dynamics of competition played out over a period that spanned years and even decades for some
businesses.
Businesses with high profit in the end, including Gannett in newspaper
publishing, Turner Broadcasting in news, and Kroger in food stores, engaged in
long processes of asset development for future reward. In some cases, the development of necessary resources and capabilities occurred before returns materialized. Turner Broadcasting built formidable, inimitable advantages before rivals
perceived the full value created by its business model. Kroger crafted a strategy
based on incremental improvements in asset utilization and cost structure to
increase profits gradually over time. Despite increasing pressure for better performance, these businesses achieved enduring advantages from persevering
commitment to carefully developed business models, often under considerable
risk of failure.
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
95
Competition, Strategy, and Business Performance
Many low performers showed signs of engagement in programs to
develop assets and capabilities. McCaw in cellular telephone and Bethlehem
Steel in steel-related operations invested for future return. For some low performers, prospects were dimmer. The assets that once created great value and
deterred imitation locked the businesses into poor position. Companies such
as Everest & Jennings in medical devices faced powerful incentives to adhere to
old approaches because of sunk investments and demands from established customers and suppliers. For L.A. Gear, low customer switching costs meant that
revenues declined quickly as fashion-conscious buyers turned to competing
products. The cases show that some low performers faced challenges arising
from difficulties in adapting to new business models. In many cases, the adaptation was constrained not by a lack of willingness to change, but rather by the
conflicting incentives associated with serving an established base of customers
and suppliers that adhered to old ways of doing business.
Some businesses with medium performance developed resources incrementally rather than through overwhelming investment under uncertainty
about return. Federal Express in express delivery earned profits while investing
moderately over time. The risks associated with this strategy were not as great in
the sense that the business generated returns as growth occurred. The analysis
showed that Steady Moderate Performers more often came from the wholesale
trade, retail trade, and transportation sectors than businesses in other categories.
With assets of nearly a billion dollars, the average Steady Moderate Performer
was larger than the typical business in any other category and over twice as
large as the average Sustained High Performer.
Why Did Accounting Profit Decline During the 1980s and Early 1990s?
So where did competition intensify in the economy? What explains the
decline in accounting profitability over the 1980s and 1990s? Why did financialmarket premiums increase over the same period, putting enormous pressure on
managers? The evidence suggests that business conditions changed during the
1980s and early 1990s in several ways.
First, standards of performance increased in many industries. A statistical
analysis shows that conventional industry structures remained an important
influence on accounting profitability.42 The analysis of financial-market premiums indicated that, overall, investors expected that conventional industry
boundaries would remain important to profitability in the future.43 However,
a finer look at each of the categories of performance suggests increases in the
requirements for operational effectiveness within many industries.44 Deregulation of the airline industry, for example, had a profound impact on the kinds of
large commitments necessary for superior profitability. Some airlines continued
to redeploy assets twenty years after deregulation took hold. In the postal-service industry, the standards for speed of delivery increased dramatically after
Federal Express implemented state-of-the-art technology. In supermarket retailing, Kroger relentlessly achieved great improvements in asset utilization over
96
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
time by integrating detailed knowledge of customer preferences with its operational expertise.
Second, the evidence suggests a complicated set of tradeoffs between
size and growth. The statistical analysis shows that growth was associated with
higher profitability and higher financial-market premiums. Of course, growth
generates increased size. Although larger businesses had higher accounting
profitability than smaller businesses on average, smaller companies had higher
financial-market premiums. The detailed analysis of performance by category
points to a two-part tradeoff between size and profitability. At the high end of
profitability, Sustained High Performers were less than half the size of the Steady
Moderate Performers. At the low end of profitability, the tradeoff went the other
way, with Steady Moderate Performers over twice as large as Chronic Underperformers. Businesses that improved from low performance—Turnarounds and
Rising Underperformers—had high growth rates. Growth was associated with
the achievement of better performance, but sustained high performers were relatively small. These patterns suggest that, during the 1980s and early 1990s, successful businesses had to navigate a varied set of strategic challenges over time.
Organizations had to adapt through phases of major investment and growth;
leadership and profitability; and competitive challenge and incremental investment. In some cases, businesses were penalized for too much growth with a loss
of strategic focus that subsequently led to lower profitability.
Third, the evidence suggests that important competitive challenges came
from businesses that adapted and extended the successful approaches of their
rivals. Although direct replications clearly hurt L.A. Gear, a Fallen High Performer, it was adaptation by imitators that hurt Benihana, a Declining Moderate Performer. The analysis showed that Declining Moderate Performance was more
common than Fallen High Performance, which suggests the need for further
study of adaptation among rivals. The case examples also showed that Chronic
Underperformers engaged not in price wars but rather in major investment projects that reflected new business approaches. In particular, the core examples of
Chronic Underperformers engaged in long-term investment programs with the
intention of becoming Turnarounds. For older firms that fell into low performance, the case examples indicated conflicting incentives for updating their
approaches. When a company held assets tailored to an old way of doing business, the cost of updating its business approach was effectively greater than for
newcomers to the industry. For this reason, the accomplishments of many Rising
Underperformers are truly impressive. For example, several of the businesses in
this category engaged in comprehensive programs for updating their own positions as well as those of critical suppliers. In sum, the nature of the challenges
facing low performers did not always have to do with the short-run competitive
interaction that characterized direct imitation.
Finally, corporate-parent effects were less important than industry and
business-specific effects on profitability, although corporate parents may have
been somewhat more important for businesses that improved profitability over
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
97
Competition, Strategy, and Business Performance
time.45 The statistical and case research shows that corporate effects may have
been quite important in specific types of situations, however. For Turner Broadcasting in news, a positive corporate effect brought profitability above the economic average several years before business-specific profits became positive.
Corporate effects were especially influential in the achievement of high
performance.
These observations show an enduring logic of competition that shaped
interaction between firms during the 1980s and early 1990s. Industry structure
and corporate diversification remained important sources of ongoing advantage.
Many businesses earned high returns within their industries by commercializing
new products after years of investment. Low performers often innovated aggressively. The standards for moderate performance rose substantially. For managers
charged with improving their operating strategies, the critical challenge is understanding the enduring logic of competition in an increasingly competitive
environment.
Notes
1. The ratio, also called Tobin’s q, is greater than one when financial investors expect
the firm to generate more value from its assets than if the assets were sold off at
their replacement values. This measure incorporates the values of stocks, bonds,
and other invested capital, and it controls for accounting anomalies by calibrating
the premium on the replacement value of assets rather than on earnings.
2. See Anita M. McGahan and Michael Porter, “How Much Does Industry Matter,
Really?” Strategic Management Journal (July 1997), pp. 15-30.
3. The studies are McGahan and Porter, op. cit.; Anita M. McGahan and Michael
Porter, “The Persistence of Shocks to Profitability,” Review of Economics and Statistics
(Spring 1999); Anita M. McGahan and Michael Porter, “The Emergence and Sustainability of Abnormal Profits,” Harvard Business School Working Paper, May
1997, rev. April 1998; Anita M. McGahan and Michael Porter, “What Do We
Know About Variance in Accounting Profitability?” Harvard Business School
Working Paper, December 1997; Anita M. McGahan, “Profitability Data on U.S.
Industries and Companies,” Harvard Business School Note 9-792-066, 1992, rev.
1988; Anita M. McGahan, “The Performance of U.S. Corporations: 1981-1994,”
Harvard Business School Working Paper, July 1998, rev. December 1998; Anita
M. McGahan, “Do Competitors Perform Better When they Pursue Different
Strategies?” Harvard Business School manuscript, March 1998. The analysis also
provides the foundation for a new advanced course in strategy at the Harvard
Business School called “Strategy and Business Performance.”
4. For additional information on this measure, see Anita M. McGahan (July 1998),
op. cit.
5. Richard Schmalensee, “Do Markets Differ Much?” American Economic Review, 75/3
(June 1985): 341-351; Larry H.P. Lang, and Rene M. Stulz, “Tobin’s q, Corporate
Diversification, and Firm Performance,” Journal of Political Economy, 102/6 (1994):
1248-1280.
6. Because this figure is the average ratio of operating income to assets, it is higher
than the average return on assets in Figure 2.
7. The classification scheme used here was developed in an effort to achieve simplicity without obscuring important regularities identified through statistical analysis
98
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
in the previous research. For example, McGahan and Porter [(May 1997), op. cit.]
use a Markov analysis to show that the likelihood that a business would post
high, moderate, or low profitability depended statistically on the entire available
profit history of the business.
For example, the ranking by category of the average sizes of businesses is the
same when only long-lived businesses are included in the analysis (although all
average sizes are larger).
Accounting profitability is measured by the ratio of operating income to assets.
For the “beginning” categorization, each business was classified as a high,
medium, or low performer based on its average profitability ranking among all
businesses during their first four years in the dataset. For the “ending” categorization, each business was classified into a category based on its relative average
profitability ranking among all businesses during their last four years. The categorization roughly conforms to natural breaks in the economic characteristics of the
businesses. See McGahan and Porter (May 1997), op. cit. Businesses in the second
and third quartiles are grouped together in the “medium” category because their
economic characteristics are similar and because of high degrees of movement
between the second and third quartiles.
When a business appeared for less than four years, then its position was based on
the average profitability for the years in which it appeared. Recall that single-year
appearances are screened from the dataset.
The thresholds map loosely to the range of cost-of-capital figures obtained in
case research on companies in the U.S. economy. It also suggests that a significant number of businesses may not be returning their costs of capital over long
periods.
If there were no systematic relationship between beginning and ending categories,
then 6.25% of businesses would be classified as Turnarounds (i.e., 25% of the
25% that started with low performance would randomly end with high performance).
If falls from high to low performance occurred randomly, then 6.25% of the businesses would qualify for the category.
If the retention of moderate performance occurred randomly, then 25% of businesses (i.e., 50% of 50%) would qualify for the category. The 40.7% figure is
much higher than the number that would occur randomly.
These results are verified statistically in McGahan and Porter (Spring 1999), op.
cit.; McGahan and Porter (May 1997), op. cit.
One type of analysis does shed light on issues of exit, however. A separate study,
McGahan and Porter [(May 1997), op. cit.], using a similar Compustat dataset,
established that the rate of business “exit” from 1981 to 1994 was about 15% per
year for businesses that ranked in the bottom 10% by profitability. Among businesses that ranked in the top 90% by profitability, the annual rate of exit was
nearly uniform at about 10%. Even high performers exited at this rate, which
suggested that their owners might have been inclined to sell at a performance
peak. Thus, the rates of business exit for high, medium, and low performers do
not follow a simple pattern, partly because “exit” from the dataset occurs for a
variety of reasons.
The average is obtained from the mean rate of profitability for each business
over its reporting history. Theoretically, this average is slightly biased because it
gives equal weight to the profitability of businesses with long and short reporting
histories. In practice, the average is not substantially different from the unbiased
estimator, which weights each observation by the inverse of the variance of the
estimate. The simple averages are reported here to make them clear to the general
reader.
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
99
Competition, Strategy, and Business Performance
18. The financial-market premiums in Table 2 are the means across businesses. The
value for each business is the mean for its corporate parent over the entire period
in which it appears in the dataset.
19. The financial-market value for a business is equal to the financial-market value
for a corporation multiplied by the portion of the corporation’s assets in the
business.
20. The growth rates here should not be used for forecasting and should be interpreted to apply only to the 1981-1997 period. Even in the weighted averages, a
bias arises from the end points of the series on each business. See McGahan and
Porter (Spring 1999), op. cit., for discussion of the problem and of unbiased
estimators.
21. Industries are defined by 4-digit Standard Industry Classification codes.
22. McGahan and Porter (July 1997), op. cit., report that the influence of industry,
corporate-parent, and business-specific effects varied by economic sectors. Over
all sectors, the fixed effects of industry, corporate parents, and specific businesses
accounted for 19%, 4%, and 32%, respectively, of variance in accounting
profitability.
23. See McGahan, “Facts on the Logic of Competition,” Harvard Business School
Working Paper, November 1998; McGahan and Porter (December 1997), op. cit.
The term “attractive industry” indicates that average profitability among industry
members was above the overall average.
24. The tendency of a corporation to perform differently than the averages for its
industries defines a “corporate effect” for the firm.
25. See Pankaj Ghemawat and Scott B. Garrell, “USA Today: Making Headlines Across
the Nation,” Harvard Business School Publishing Case Study 792-030; John
Deighton and Anthony St. George, “USA Today Online,” Harvard Business School
Publishing Case Study 598-133.
26. The figures were constructed by examining the average of coefficients in a nested
ordinary-least-squares regression of industry, corporate-parent, and businessspecific influences on profitability. See McGahan and Porter (December 1997),
op. cit.
27. “Benihana Inc. Opens 62nd Restaurant, Introducing New Sushi Theme Concept,”
Business Wire, July 20, 1998; Richard Gibson, “Restaurants,” Wall Street Journal,
June 30, 1998.
28. See Rona Gindin, “Market Segment Report: Casual Theme,” Restaurant Business,
91/17, November 20, 1992, p. 169; Jeffrey A. Trachtenberg, “Planet Hollywood
Founder Sees a World of Theme Cafes,” Wall Street Journal, December 13, 1995,
p. 4; The Associated Press, “Theme Restaurants Abound,” The Columbian, December 12, 1995; and Deborah Adamson, “’Entertainment’ a Hot Dish Theme Dining
Serves Up Some Popular Recipes,” Los Angeles Daily News, February 9, 1998, p. B1.
29. Following Michael E. Porter [Competitive Strategy (New York, NY: Free Press, 1980)]
and Adam Brandenburger and Barry Nalebuff [Co-opetition (New York, NY: Doubleday, 1996)], this approach emphasizes the importance of both the suppliers to
the firm and the customers of the firm. Suppliers account for the firm’s cost structure, while customers account for the firm’s revenue structure.
30. The business press contains many reports that themed restaurants like Benihana
compete to attract customers interested in the “eatertainment” experience,
for example. See “Benihana Profits up 35.6% as Same-Store Sales Grow,”
Nation’s Restaurant News, 31/45, November 10, 1997, p. 12; Tracy Kolody, “That’s
Entertainment Restaurants Are Catering to Diners Who Have a Taste for Fun and
Excitement for the Whole Family,” Sun Sentinel, September 22, 1996, p. 1G; David
Farkas, “Theme Dreams,” Restaurant Hospitality, 82/1 (January 1998): 36-44.
100
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
Competition, Strategy, and Business Performance
31. L.A. Gear targeted products at teenagers and young adults, demographic groups
famous for shifts in their fashion interests. See “Teen Spending Keeps Climbing,”
American Demographics (January 1998).
32. In 1993, Forbes reported that “By 1991, L.A. Gear counted 12 million pairs of
shoes in warehouses and only 1.5 million in cash. [New investors brought $100
million to the company in late 1991, and installed a turnaround specialist as president.] . . . Goldston’s first move upon becoming president was to stop inventory
from piling up by shutting down production at contract factories overseas for 70
days. . . . Goldston then began a fire sale of excess inventory, a move that further
hurt the already soiled brand image, but it gave the company close to $100 million of much needed cash.” Damon Darlin, “Getting Beyond a Market Niche,”
Forbes, November 22, 1993, p. 106.
33. L.A. Gear’s Annual Report for 1991 indicates a new strategic direction emphasizing
“price and value,” a sign that the company had to lower prices. It introduced a
new marketing program intended to “differentiate L.A. Gear from its
competition.”
34. For information on lock-in and sunk costs in the U.S. Airline industry, see “The
U.S. Airline Industry in 1995” Harvard Business School case 795-113 and “The
U.S. Airline Industry in 1995, Teaching Note” Harvard Business School case
799-023.
35. The source for this paragraph is Anita M. McGahan, “Sunrise Medical, Inc.’s
Wheelchair Products,” Harvard Business School 9-794-069.
36. Here, “Turnaround” indicates only that the accounting profitability in a business
improved significantly over the 1981-1997 period. This definition of
“Turnaround” does not imply that good management replaced bad management;
indeed, the definition encompasses the possibility that a business is confronted
with a dramatic new investment opportunity.
37. CNN’s success in cable news preceded broad investments in other kinds of specialized cable formats, including nature-oriented, weather, sports, and children’s’
channels.
38. Hank Whittemore’s book CNN: The Inside Story (Boston, MA: Little Brown and
Company, 1990) describes the long period of investment by Turner in broadcasting, and the kinds of tangible and intangible resources that emerged as a result.
The book also describes initial, widespread skepticism about the viability of CNN’s
format and the subsequent attempts to imitate the approach after it proved
successful.
39. McGahan and Porter (May 1997), op. cit.
40. See Gene N. Landrum, Profiles of Genius (Buffalo, NY: Prometheus Books, 1993),
chapter 18, for an account of the skepticism that preceded Turner’s success in
broadcasting, as well as Turner’s personal leadership and the accolades that followed the turnaround of CNN.
41. For a detailed framework for analyzing these kinds of asset-development opportunities, see Pankaj Ghemawat, Commitment (NY: Free Press, 1991).
42. McGahan and Porter (July 1997), op. cit.
43. McGahan (July 1998), op. cit.
44. In “What is Strategy?” [Harvard Business Review, 74/6 (November/December
1996): 61-78], Michael Porter made a distinction between operational effectiveness and competitive strategy. His definition of operational effectiveness is used
here.
45. Nitin Nohria documents the trend toward networked organizations in “From the
M-form to the N-form: Taking Stock of Changes in the Large Industrial Corporation,” Harvard Business School Working Paper 96-054 (June 1996).
CALIFORNIA MANAGEMENT REVIEW
VOL. 41, NO. 3
SPRING 1999
101
Purchase answer to see full
attachment