Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Applied Materials No-Politics Thread (AMAT) -- Ignore unavailable to you. Want to Upgrade?

To: Cary Salsberg who wrote (6373)	7/3/2003 1:04:55 PM
From: Pink Minion	Read Replies (1) \| Respond to of 25522

>> Previous upcycles were mostly a function of the nature of fab construction Are you referring to the out-sourcing model? Don't you think it had to do with the hypergrowth of the PC sector? It use to be 75% of the chips made. When that's growing at 20-30% per year, it's kind of hard to match demand. If we see a product that cost $2600 and uses $1200 bucks in chips with that kind of growth, I'll bet the out-sourcing model would have problems keeping up. When they're all booked up, isn't that when companies build their own fabs?

To: Cary Salsberg who wrote (6373)	7/4/2003 12:44:46 PM
From: runes	Read Replies (5) \| Respond to of 25522

<<There is a phrase about "generals fighting the last war." >>. ,,,And, as you have noted there is a phrase about failing to learn the lessons of history.... SO much for phraseology. But you are correct that predicting the future (extrapolation) is, at best, an uncertain proposition. There are no sure things but there are more reliable and less reliable techniques. ...Blind assumption that nothing changes is the first and simplest technique (generals fighting the last war) and is not necessarily a bad thing to do. But it can go horribly bad when the underlying mechanism changes. Which leads to a better method of extrapolation - to extrapolate based on an understanding of the underlying mechanism. And if you will be so kind to go back and re-read my post, you will see that I pointed out a very specific mechanism - the high cost of capacity expansion coupled with fluctuating demand and significant time lags between the need for capacity and when that capacity comes on line. ...And I went even further by pointing out that the same mechanism is not something unique and specific to the semis. The same thing pops up in natural gas, in petroleum refining (refining capacity is also becoming constrained) etc. etc. But now - let's get down to the nitty-gritty. You kind of raise two points (if I am reading you correctly). First - the playing field has changed with the emergence of the foundry/outsourcing model. ...Anectdotally - the foundries were in dominance in the '98 - '00 bubble. XLNX, ALTR, CSCO, JNPR etc. etc. were all outsourcing at that time. TSMC was literally auctioning off their capacity in blocks with a 2 or 3 year contract lengths. UMC, Chartered, Tech, MVI etc. were all going into foundry in the early ot mid '90's. And the Cost of ownership model actually came into practice in the mid '90's so that's not new. ...Fundamentaly, foundry vs captive fabs do not change the investiment/leadtime dynamic. If you will notice, a number of them put greenfields and buildouts on hold in the last downturn and they are just now starting to murmur about adding capacity (as the leading edge capacity is starting to hit the magic 90% utilization number). ...I DON'T SEE THIS AS INVALIDATING THE EXISING MECHANISM. ...(FYI - in Taiwan new fabs are exempt from taxes for the first 5 years. So there is actually a strong incentive to build new fabs rather than retool older fabs). SECOND - you seem to feel that because the semi market is no longer PC dominated and is driven by a wide variety of products, that the magnitude of the demand fluctuations (the forcing function to my mechanism) will be abated. ...Anectdotally - the PC component of the demand has been steadily declining since it's peak in '83 (advent of the IBM PC). And yet the '98 - '00 boom was the biggest one yet. ...Fundamentally - the problem is that almost all of the demand components are correlated to the economy. When the economy is bad, tech investment stinks. When the economy is good they can't invest fast enough in the latest fad. (Or dot com as the case may be). ...SO THE FORCING FUNCTION TO MY UNDERLYING MECHNISM STILL HOLDS. A piece of trivia - Jim Bagley gave an interview back in '99 where he predicted that the semi boom-bust cycles were over because the investment community was no longer willing to support overbuilding capacity. WHOOPS!