Editorial: Semiconductors: A mature industry?
08/01/1999
 Bob Haavind, SST Editor in Chief |
During the Industry Strategy Symposium early this year, a number of investment gurus referred to the semiconductor business as a "mature industry." When we think of a mature industry, we think of commodities like steel, paper, and even automobiles. In the high-tech arena, product categories such as calculators, VCRs, and transistor radios come to mind. These are industries with margins so paper-thin that only the very lowest-cost competitors can keep head above water. Mature industries are extremely unattractive for investment capital.
Is this really how the whole investment community views a vibrant, dynamic marketplace like semiconductors? If so, it is a bad omen. It suggests an industry that is hardly managing to stay afloat in spite of the noblest efforts, with little potential for any growth, let alone rapid growth.
Finally, on the last day of the symposium, I asked a question of a panel of financiers about this frequently expressed view. The semiconductor industry is entering a period of dramatic changes, in materials, processes, wafer size, and chip design. Progress is speeding even ahead of Moore's Law, the rock upon which the industry plans its Roadmap for the future. So far this approach has provided annual growth rates averaging in the 20% range, despite some dramatic ups and downs. Defining the industry as "mature" implies it has reached a plateau. Was that their view?
The answer was revealing. The new growth industries, it was explained, involve the Internet and the World Wide Web. Companies can be launched with little capital, yet they have fantastic upside potential. This is where investors want to put their money for maximum impact. Semiconductor plants are becoming prohibitively expensive, and even launching a new tool company requires huge investments that may or may not pay off. Therefore, the investment community decided to lump semiconductors right in with other "mature" industries. By using a convenient category like this, the need actually to analyze specific projects and business plans becomes moot. "It's in a mature industry. We don't invest in mature industries," provides a quick, easy out. It makes more sense, from this view, to give a neighbor funds to start an Internet enterprise, and then cash in when the new enterprise goes public (which may be only a matter of months, or even weeks, with a red-hot IPO market).
It is only fair to point out that some of the analysts participating in the ISS are themselves strong semiconductor supporters, but they were explaining the pervasive view that had overtaken Wall Street. Their firms' investment strategies were being shaped more by this consensus view than by their own analyses. Needless to say, there were murmurs of disapproval among the many top executives attending the symposium. Their feeling was that this attitude could hurt the industry just when there was an especially great need to fund the new technologies that will be required to provide the upgraded plants and infrastructure for the new millennium.
More recently, the ardor of Wall Street for those kick-start Web enterprises seems to be cooling. Some analysts are suggesting that investments in the infrastructure companies - those that supply the specialized chips and equipment to expand the bandwidth available for exchanging information on the Web or on intranets - might be a wiser play than the actual content providers. No matter who wins or loses among content providers and services, however, they will need the infrastructure that is now rapidly taking shape. At the same time, there is also tough competition among the communication carriers. One sure thing about all modes of communications: they depend on high-performance ICs, and the support chips and devices needed to turn them into effective systems. As chip markets give signs of a rebound from a couple of tough years, a few voices are beginning to point to the semiconductor/process tool industries as winners no matter what happens downstream.
If a few profit surprises begin to pop up in the semiconductor industry, we expect that by next year's ISS, those "mature industry" comments will have gone the way of bipolar TTL and bubble memories.