Semiconductor slowdown is imminent
03/01/2000
Mark Diorio
Although you are reading this column in the March issue of Advanced Packaging, it was written in January, which is, coincidentally, the time of year when set my annual forecast. Actually, I started my forecasting research for 2000 in the fourth quarter of 1999 (4Q99), but I didn`t really solidify it until the first week or two of January. I only hope that it reaches you in time for you to have a chance to revise your forecast for the 2000 calendar year.
My suggestion to you: Be conservative with your 2000 calendar year forecast because a semiconductor slowdown is imminent. In fact, it is probably overdue. The IC industry has been overheating (accepting double bookings and building over capacity) and history has proven this to be a dangerous situation.
It may seem incredible to you that I am making such a pessimistic suggestion when the current mood in the industry is quite positive. After all, the economy has never been better. Semiconductor applications are increasing. Companies are surging. Stock markets are soaring. But, on the other hand, DRAM pricing has been deteriorating, non-memory IC average selling price has been in decline, and comments among some IC vendors may indicate cancellations and push-outs. The factors that are present today are very similar to those in 1995 and 1997, which led to the subsequent downturns in 1996 and 1998.
You might recall last year I forecasted 1999 to be a strong year - that while the early portion of 1999 would be relatively slow, the second half would be dynamic. And you may also recall that I anticipated that items that would bring about the resurgence to the business levels would also provide the mechanisms for decline over time. Those mechanisms are now beginning to present themselves, so it`s prudent to prepare accordingly.
When plotting a course, it is always beneficial (and more accurate) to have at least two points of reference. Therefore, don`t take my word alone - use another source. One such source that I would recommend is Advanced Forecasting Inc. (Cupertino, Calif.), a group that has consistently forecasted the industry`s turning points during the 10 years I have used them. Advanced Forecasting has stated that "the probability of an immediate slowdown of the IC market is estimated at 80 percent."
Advanced Forecasting bases its forecasts on "true demand models" that start out of economic factors that influence semiconductor (and other components) consumption by various end-equipment industries. It is worth noting that Advanced Forecasting`s methods to determine the probability factor are based on similarities reflected in the last four recession periods during the past 15 years. Now this in itself may be the only glimmer of hope in that the anticipated downturn is erroneous: Both Advanced Forecasting and I take into consideration past experiences, our models consider history instead more human-type factors. Past experiences may have very little bearing on the future because what is happening in the market is beyond all logical rationale. By this, I mean the high stock valuations, the record breaking stock trading levels and the very stable interest rates. The economy, in general, is on fire and analysts are saying that investments and the economy should be strong at least through November 2000 and quite possibly into 2001. So there may be a chance that the logic I and other forecasters have applied in the past has no bearing on 2000; our models have not taken into consideration today`s market conditions and are, therefore, outdated and inapplicable.
There is a new phenomenon fueling today`s investments, and our models do not take this into consideration: speculation. Speculation is a human reflex that closely resembles gambling. It is unpredictable, a bet against time and dangerous. Speculation cannot be quantified. I try to avoid speculation because my past 20 years in this industry has taught me this. (I have been wrong more times than naught when I have purely speculated on the outcome.) The only bets I ever win are the ones in which the odds are heavily in my favor.
So do you really want to follow the wave of public emotion? Still don`t think it possible that the semiconductor industry can hiccup and slow down? Well, if you consider two factors - double booking and building over capacity - the strong possibility exists that we will have a slowdown. And if you still don`t believe it...how about we make a bet?
MARK DiORIO, chief operating officer, can be contacted at MTBSolutions Inc., 2685 Marine Way, Suite 1220, Mountain View, CA 94043; 650-960-3203; E-mail: [email protected].