This semi business had always been boom or bust cyclical business. Thanks to massive over investment in capacity by the Koreans, we have a bust that last 3 years or two process generations instead of one.
If you buy that assumption, which is proven historically in my 20 yrs in the industry, now is a great time to be buying stock. Want to be conservative, buy XLNX, ALTR, STM, TI and you'll be rewarded with 3x by 2001. Feel like gambling a little but don't want too much risk - go for LSI, probably will return 5x to 6x from where it is today, downside 2x. Want to really gamble, buy ALSC, SIII, IDTI, NSM.
IMHO, it's simply whether you believe in this cyclical thing. I bet you when it comes, and I see that coming in 9 months or less, we would see 25 to 30% growth for 3 years - 6 months to have money people start believing, 6 months to get money together, 1 year to build fab, 6 months to ramp fab, 1 year for the me-too slow guys to also get on board and get their fab built and ramped, and it will be 3 years of prosperity before, saliving here, we go bust again. Yes, ride it on the way up, ride it on the way down. After cashing on 2001, my plans is to ride it down on DRAM and foundry stocks in year 2002.
If you are upset and want relief, I agree you should be out of LSI. Not worth the grief, much easier way to play this e.g. XLNX, ALTR or Broadcom.
I can see that you are upset over your loses, and I think a lot of people are with you here. But rather than say you are being emotional about it (nothing wrong with that, that's just human), you seem to put your focus in this # or that #.
Just for the sake of argument, let's take your R&D #. First of all, by cutting old fab and incurring big write off and starting new fab and incurring new expanses, I read management as being able to swallow bad #s for a couple of Qs to prepare for future growth. So I would trust them to cut back on R&D because I do not think they would do that if that will hurt future business. Secondly, whether they are still cutting back on R&D is still questionable, owing to the combined effectiveness of the two companies. That's why there are good take-overs and there are bad take-overs. E.g. the best merger is if VLSI would merge in. Can cut off almost all of VLSI without hurting much. End R&D #s, like any overhead #s, will shrink like you wouldn't believe. In my view, besides for some design folks, the two companies are just about completely overlapped. Finally, if it's EDA tools that they had been spending money on, I have heard from one design manager who's been working on a 0.25um design that the tools out there are just not enough to do the jobs because at 0.25um, a lot of surface factors as vs bulk factors, start taking over. And the tools out there are having a hard time predicting and quantifying the surface factors. Perhaps LSI did spend a lot of money in getting over this hurdle and not that they got over it, this can carry over to 0.18 or even 0.12 generations. Thus they can well back off in spending money on it. I do not know if this is true, but I do know that as outsiders and not hands on managers, sometimes we do have to take their word for granted. If not, should sell the stock and move onto Intel or CSCO.
patrick |
