High Growth In Desktop Computers, Few Companies

Twenty years after the start of the PC revolution computers permeate every aspect of U.S. business but have penetrated only 30–40% of the households. In other developed countries that figure falls to 15%, and it plunges to 1–2% in the developing world. As the global economy continues rapid expansion over the next few decades, these percentages should rise correspondingly. Indeed, worldwide sales rose 16% during the 3^rd quarter of ’97; in the U.S. sales increased 20%. It means desktop computer manufacturing will remain an industry of high growth over the long term. Why, then, are there so few manufacturers when instead the high growth should be attracting new entrants? Given the potential, why is consolidation in the industry continuing to occur?

13 Companies
The U.S. has 13 publicly-held, computer manufacturers with over $1 billion in revenue (Table 1). Another handful of companies generates revenues totaling $100 million or less. Between the two groups exist few if any other companies.

Eleven of the 13 have shown an income in the last year with a median value of 4.5% net profit margin for the total group, which amounted to $141 million. That is to say, 7 of the 13 companies made $141 million or more. Despite the world’s enthusiasm over computers, not all these companies can claim a foundation of solid bedrock. Only 5 of the companies have stock prices over $40, and the reason, as usual, is earnings (Fig 1). Disregarding IBM at the high end and Apple at the low, average earnings amounted to $0.73.

Not Everything Is Rosy
Among this group of companies doing over a billion, Apple is in free-fall having posted an earnings deficit of $8.30, Silicon Graphics is wobbly and looking for a buyer despite showing strong growth last quarter, Intergraph Corp and Unisys have tepid performance, Gateway, although generally strong, took a $107 million loss for the quarter, and Digital Equipment Corp earned $141 million on $13 billion for ‘97 but acknowledges a 10% decline in sales over the previous year. IBM, which is second in sales to Compaq, is growing mainly on the strength of the world market, but losing market share at the same time. It’s also true of Packard Bell, a closely-held company currently fourth largest in the industry but with declining market share.

Clear Winners
The uncontestable winners are Compaq, Dell, and Hewlett-Packard whose sales growth for the 3^rd quarter of ’97 ranged between 56 and 68%. Compaq, far and away the leader, now owns 13.7% of the world market (Table 2).

Figure 2 shows the relationship among companies with respect to growth rate over the last year. This data, however, should not necessarily be interpreted as a measure of strength. Silicon Graphics managed 25% revenue growth, but is having its troubles.

Macro Trends Of The Industry
Aside from the high quality of management among the top companies there are a number of macro trends greatly influencing the industry at this stage of its development.

In 1981, when IBM entered the home computer market, it stood as the world’s preeminent computer manufacturer with unparalleled technical and marketing sophistication. It effortlessly dominated the industry. The resulting homogeneity of hardware architecture through the ‘80s, polarized between the IBM and Mac, may well have inhibited innovation, but it had a positive effect on the development of applications. It forced software companies to orient around no more than two clearly defined operating systems. This restriction allowed for the creation of large markets in software applications. Hardware would be constrained ever after in support of the software.

The Rise Of Microsoft And Intel
Inadvertently, IBM caused the rise of Microsoft and Intel, which together continue to exert a dominating influence on the industry. This grand accommodation between these two companies, for better or worse, has provided the stability and continuity needed to make software applications flourish. Together they captured such a sizable share of the world’s market that it flattened any opportunity for either hardware companies or software to take new, even what might have been promising, directions. Around 80% of the world’s computers are Wintel, that is, they use the Intel processor and run Microsoft Windows.

The Unix Moment
The emergence of the Unix operating system in the late ‘80s was the last serious opportunity to inject diversity into the industry. Unix had undeniable technical superiority over the obvious weaknesses of Microsoft’s DOS. Had it succeeded in breaking out of its engineering niche and capturing a share of the business community we would today see more market fragmentation in semiconductors, hardware manufacturing, and software. Three influences militated against that happening—the continued high quality of the Mac through ’93 as an alternative, a general inertia against the cost of change, and the introduction of Windows 95, which spelled the demise of Unix for the consumer.

Enter Windows 95 & Windows NT
Windows 95 and Windows NT have brought a level of sophistication to the operating system that previously was found only with the Mac. Windows 95 applications have achieved, and often exceeded, the quality and ease of use of Mac applications. This parity has put tremendous pressure on Apple, which, together with its own internal stumblings, leaves Apple in a state of crisis, its future very much in doubt.

Barriers To Entry
The rapid maturation of the computer industry over the last few years has created technical and financial barriers that will restrict the entry of new players who wish to make it a business. From the mid ‘80s till perhaps ’94, small companies could gather together all the components—motherboard, hard and floppy drives, power module, internal modem—integrate them into a case, and attach their logo. Such would be a computer. It still can be done, but this sort of product holds less and less appeal to the consumer.

Over the last few years the improved performance and changing uses of computers have vitalized the area hardware peripherals. Modems have reached speeds of 56,000 bytes per second, disk drives hold several gigabytes of data, CD roms are standard equipment, and processors are now optimized for multimedia. At precisely this time the Internet has been made available for commercial use putting an end, once and for all, to the days of the standalone computer. With the Internet we see a blurring together of many different technologies involving hardware, software applications, networking, communications, browsing, and multimedia, even television and telephone. The PC is now designed to remain on 24 hours a day throughout its lifetime.

The Importance Of Brand Name
In this environment, where the computer has become of central importance to one’s daily work, any problems at all are intolerable. With this in mind, a system purchased from the top manufacturers usually includes a 3-year limited warranty with one year of onsite service. This is the level to which the competitive bar has been raised. With the acceleration of the industry toward a high degree of maturity we see the consumer valuing brand name as a way of minimizing risk. Brand name assumes that not only is the system technically excellent but that its service is excellent as well. It also assumes that the companies operate at economies of scale, which allow them to price their products at competitive levels. Weaker players are now enduring greater pressures for lack of such a brand name. Any new entrant to the industry, without brand name identification, will have to compensate in other sophisticated ways, and it will be difficult to do. Thus, we can say with assurance that those who define the field of play today, with an occasional exception, will be the same ones doing so for at least the next 10 years. Their stocks will be supported by strong fundamentals.

The Rise Of Direct Marketing
Brand name is necessary, but it may not be sufficient. The last couple of years have seen the rise of direct marketing as a force to be reckoned with. Direct marketers sell directly to the consumer bypassing distributors and retail outlets; consequently, they’re able to offer products at lower prices. In the past they’ve marketed the consumer through catalogs and other unwieldy approaches, but the proliferation of the Internet has marked a fundamental change in the way they do business.

Dell Computer Corp and Gateway 2000, Inc. have posed a challenge to the very way computers are sold. Through the Internet the consumer selects in great detail all the specifications for the desired computer including the speed of the processor, the amount of memory, cache, and storage capacity, the kind of modem, the quality of speakers and CD rom, plus the operating system and application software. In a few days the system arrives in the mail with a long-term warranty and strong support. Retail outlets, which typically buy inventory to be sold as-is, follow the traditional model of selling goods. They lack the flexibility and resources that would be required if they were to let a customer request deep changes to the product.

This new approach toward the selling of computers, essentially computers designed by the customers themselves, is already being felt by Compaq, which depends almost entirely on distribution through retail stores. Although Compaq is a perennially strong company and will remain so, they’re worried about a softening of market share among that growing body of consumers who are serious users of the Internet. IBM, Packard Bell, and others trailing farther down in market share have much to be concerned about. Their elaborate systems of distribution through stores will not easily be changed. For the time being the market is able to accommodate both approaches, but it speaks to the uncharted terrain onto which the industry has been forced by the Internet when major players can feel themselves at such risk.

Dell takes the lead in direct marketing. The company is aggressive and well-managed with high-quality products at a competitive price and rapidly building brand-name awareness. We can expect Dell to remain in high growth for the foreseeable future. Any new companies emerging on the scene will most surely have to take this form.

The High-End Problem
A final major influence on the increasing strength of some companies and the weakening of others is the clash between the low end and the high. At the high end are very powerful but also very expensive computers. Silicon Graphics is the classic example. In the early ‘90s it stood alone in its ability as a desktop computer to handle process-intensive applications, such as special effects for the film industry or auto design. Although its pricepoint came in around $100,000, the solution it offered was still less expensive than what previously had been used. Specialized software was developed for the machine, and companies were formed around the use of these tools where power was essential.

In the meantime, the low end consisting mainly of PCs year by year improved their processing speeds until they became viable, cheaper alternatives for those Silicon Graphics machines at the high end. The industry has shown that it’s always easier for a low-cost system over time to increase in power than it is for a powerful system to come down in price. Silicon Graphics has been caught in this crunch. Even though its revenues increased to $3.7 billion in ’97 compared to $2.9 billion in ’96 making for 25% growth, its stock price was in decline throughout the entire year. In the first quarter of ’98, the company is showing a $37 million net loss, or $0.20 a share.

Either it moves to a higher level, again out of reach by the PCs but with a more restricted customer base, or it matches the PCs in price, which it may not be able to do because of its higher cost of manufacture. A third alternative, toward which it is moving grudgingly, is to de-emphasize its own hardware architecture and operating system and instead start selling low-end workstations based on the Intel processor and Windows NT. The problem would then turn on differentiating itself in the marketplace. How the company will resolve this dilemma remains to be seen. Layoffs and management changes are being proposed. Investors should be cautious during this period of repositioning.

To less of an extent Sun Microsystems suffers from the same problem. Its workstations, which were all the rage in the engineering community from the mid ‘80s until the recent run-up in processing power by the PCs, have transitioned well into the growing server segment of the market. But they are by no means immune from the PCs and the Microsoft Windows NT operating system. Sun’s latest success has been the creation of the Java programming language, which has swept the software development world. However, it’s not clear how they intend to make money from Java. Further, if Java is to flourish, it will have to come under the control of the International Standards Organization, and this, Sun is resisting.

Sun’s stock has had a good run over the last year, from a low of $25.50 to a high in October of $53.31 with earnings of $1.96. It’s received a great deal of press attention from its promotion of the so-called "network computer" as well as all its activities involving the Internet and especially its unrelenting verbal attacks on Microsoft. The visibility seems to have had a positive influence on stock price. Even so, there’s something unsettled about Sun’s direction and its ability to maintain growth.

IBM Shows How
IBM, whose existence teetered on the brink of disaster a few years back, is showing the others what is meant by quality management. Even though its desktop computer products are enduring a declining rate of growth, IBM has many lines of business, which are doing well. Its revenues have now reached an incomprehensible $76 billion. Its growth rate last year was 5.6%, $4 billion over the previous year. And its earnings were an extremely strong $5.84. Fundamentals such as these are the reasons its stock is, or should be, around $100 or more.

Figure 3 graphically shows one way IBM is keeping its earnings high. Compared to revenue, there’s very little stock outstanding, 982 million shares—less, in fact, than the amount Hewlett-Packard has in the market, which is a company half its size. Since January of ’95 IBM has spent $16 billion in a plan to buy back its own stock, that is, to pull it in and take it out of circulation in order to reduce dilution. When net income holds constant, then earnings per share necessarily go up. Investors should find this effort by the company of some interest.

One final comment about Hewlett-Packard. It grew 22% last year and showed $2.6 billion in earnings. Despite these impressive figures it managed a $64 stock as of this writing, which demonstrates the suppressive effects of dilution on stock price.