Robgera, I believe that that writer (as well as the provider of the data on current and next year capacity increases) are missing a major point. The 1995 peak was purely a "semi" affair, over expansion, but during a general economic environment where growth did not slow materially. Right now, however, I fear we are faced with a "double whamo", not only it seems that we are facing over capacity (that article mentioned that most of the capacity increase in the last quarter and year, was from process improvements, not new increases in monthly wafer starts) next year, when the bulge of the cap-ex acquired this year will come on line, we may be faced with a small market contraction due to a possible emergence of an almost simultaneous world-wide recession. (World-wide, in this case, means US, barely a recession, Europe some decline and SE Asia, some real troubles).
The only "silver lining" I see, is that there will be very little 12" raw wafers capacity to fill all these new 12" fabs coming on line in the next 18 months (something like 250,000 12" wafer per month may be required, and the wafer boys have no money to install 12" capacity), which will reign in flooding the chip market from that new source of chips. The down side of that, will be major push out of plans for additional 12" new fabs. In Mighty Taiwan, Winbond has already cancelled (or pushed out "indefinitely") one of two planned 12" fabs.
By the way, about 2.5 months ago (around Sep 15), when AMAT was still above $75, we had a heated discussion on this thread whether we did or did not see the peak for this cycle (my definition then and now, was peak in stock prices). With AMAT breaching $40 and only $15 or so from what I suggested earlier might be considered the low for this cycle, are we going to admit we are in a down cycle? If we are, the discussion should really start and move to what would be typical parameters to watch for a bottom in the current down cycle. Other questions that we may want to address would be "is it too late to sell?"
Zeev |
