Predicting the size of the global chip market is never easy. Just ask the Semiconductor Industry Association -- it never gets it right. Making it even tougher is that the chip industry almost never hits its average annual growth rate.
Revenues don't even come close to matching the average annual growth rate of the global industry because of the wild boom-and-bust cycles that it has taken throughout the '80s and '90s.
But surprise -- 1999 was different. For the first time in more than two decades, chip revenues actually grew at the industry's historical average of 17% a year.
"I can't believe it -- 1999 came in exactly at the industry average of 17%," acknowledges Bill McClean of IC Insights Inc. Only two years ago, the veteran chip analyst caused a bit of a stir as well as some chuckles from chip executives when he observed that the chip industry hadn't come within seven percentage points of its "historical growth average" since 1978.
"Just when you say something never happens, it does in this crazy business," quips McClean.
No one really expects the chip industry to record another average growth year anytime soon. Analysts and industry managers are now predicting one of the chip industry's "non-average" growth years of 25% or more in 2000. And for 2001, most industry forecasts are looking for another growth year on the high-side side.
Forecasters say 2002 is a difficult one to call, but it now looks like another strong growth year, followed by a slump in 2003. And who knows what kind of roller coaster ride the industry will have in 2004, 2004 and 2006?
Not long ago, semiconductor companies learned the hard way never to say "never" when it comes to ruling out the one consistency of semiconductor sales trends: the industry's inconsistency and its wild boom-and-bust cycles. In 1995 -- when chip revenues were growing 42% over the previous year -- some managers started concluding that downturns would be a thing of the past and that stable and more predictable growth would be the norm in the future for global chip markets.
And several even thought chip revenues would keep climbing at the 1995 rate and reach $300 million by 2000. We all know happened next -- the unprecedented 1996-1998 slump.
It took 1999's 17% growth rate to finally surpass the industry's high-water market of $144 billion in 1995. Even with a 25% spurt over the next 12 months, chip sales will end up $120 billion short of the ambitious $300 billion target set five years ago.
"We made a lot of mistakes, and one of them was not forecasting the 1998 downturn," observes Jean-Philippe Dauvin, vice president and chief economist at STMicroelectronics in Paris. "But we also remain optimistic. . . . The $300 billion mark will likely be reached in 2003, and it might even be achievable by 2002," he says.
The one thing Dauvin won't predict is an end to the chip industry's boom-and-bust cycles. And for good reason.