by John Rhea
WASHINGTON — Captains of the electronics industry — particularly those in the military and aerospace segment — believe fundamentally in their youthful vigor and in their future creative glories.
What they believe is essentially true, but it is also true that all industries move inexorably toward maturity and that none is immune to ironclad laws of economics.
Executives and entrepreneurs can conveniently ignore these laws when demand exceeds their companies' ability to satisfy that demand. As supply and demand approach equilibrium, however, it becomes increasingly unlikely there will be enough chairs for everybody when the music stops.
The music remains buoyant today, but maybe it's drowning out uncomfortable news that nobody wants to hear. Richard Guy, vice president for marketing at Amkor Test Systems in Wichita, Kan., cites the high levels of investment that board-level companies must make up front to stay abreast of rapidly advancing chip technology.
With 400-pin chips now becoming common in the commercial world, Guy says, the non-recurring engineering (NRE) necessary just to test those chips can run $40,000 to $75,000, even though the company may need only 20 or 30 pieces a year.
As a result, surviving companies in the so-called "VME corridor" will need sufficient sales volume to cover the required infrastructure, warns Edmond Hennessy, vice president for marketing at Sky Computers Inc. in Chelmsford, Mass. In a youthful market that could be $5 million a year. In a maturing segment it could be $20 million a year. Obviously it will be considerably more in a mature segment.
The near-term solution is increasing use of contracting out services to reduce NRE and infrastructure costs. Guy's company is in the business of supplying those services, and Hennessy's company is reconciled to the need for using them. Both can use equilibrium to their advantage.
"While 'outsourcing' has existed in the U.S. since George III hired Hessian mercenaries to suppress the American Revolution, it has assumed revolutionary commercial scale in 1990s America," says David Pearce Snyder, who describes himself as a "consulting futurist" at the Snyder Family Enterprise in Bethesda, Md. He estimates that outsourcing by U.S. corporations, all levels of government, and non-profit organizations amounted to less than $25 billion a year in 1990 and had climbed to $100 billion by 1996. He projects it to reach $318 billion this year.
The ironclad law of economics that has served the electronics industry so well to date is the well-known experience curve. The law is so familiar, in fact, that it is for all purposes self-enforcing because buyers and sellers alike understand it.
As first spelled out by the Boston Consulting Group (BCG) in 1966, the experience curve states that prices will decline at a constant rate of 25 percent each time accumulated experience (as represented by cumulative industrywide sales volume) doubles.
This is a universal law, and it applies to all industries, not just those in advanced technologies. Two thousand years ago the Roman engineer Vitrivius described how undershot waterwheels — those in which the lower part of the wheel was immersed in a stream so that the current turned it in the reverse direction — achieved efficiencies of 15 to 30 percent, which was adequate for milling grain.
The technological breakthrough that enabled millers to apply the experience curve was the overshot wheel. It raised efficiencies to 50 to 70 percent, thus driving down costs and making the price of the milled grain more competitive. Yet this approach required high initial investments in dams, millraces, sluice gates, and tailraces, as well as in gearing.
The decision to take that next step was analogous to the decision facing semiconductor companies today of whether or not to invest a billion dollars in state-of-the-art wafer fabrication facilities. In each case, the successful organizations have the financial resources to ante up and stay in the game while the ranks of the players thin.
This leads to another, less familiar law known as the rule of three and four. This law, which came from BCG's late founder Bruce Henderson, states that any stable competitive market never has more than three significant competitors. The largest of these, furthermore, has no more than four times the market share of the smallest.
Henderson cites two conditions that create this rule:
1) a ratio of 2:1 in market share between any two competitors seems to be where it is neither practical nor advantageous for either competitor to increase or decrease market share; and
2) any competitor with less than one-quarter the share of the largest competitor cannot be an effective competitor.
Of course, all these laws presume some semblance of free-market conditions and are inapplicable in totalitarian societies, communist or fascist, in which the economy is centrally controlled. That's why totalitarian societies, to paraphrase Karl Marx, contain the seeds of their own destruction.
What has postponed the day of reckoning for the electronics industry is they have not achieved the level of stability that Henderson envisioned. In one of his recent presentations, Snyder saluted the plunging costs of information technology; if the automotive industry had achieved the same degree of efficiency the computer industry has, he said, cars today would cost $10 a dozen and get 250,000 miles per gallon.
I gritted my teeth when I heard that old saw 30 years ago in Silicon Valley, and I'm still gritting my teeth. The reason that cars don't cost $10 a dozen is the automotive industry's inability to ship units as quickly as can the electronics industry. The automakers proceeded down the identical experience curve, just the same, albeit at a less visible rate.
Waiting for all industries, high tech as well as low tech, at the bottom of the curve is the inevitable shakeout. It would be foolhardy to think the electronics industry will escape this trauma. There's no such thing as eternal youth. "The faster the industry growth, the faster the shakeout occurs," Henderson concludes.