Carl,
I'm not sure what the break-even point is for the semiconductor manufacturers. I do believe that the 4MB DRAMs are passe, though.
I recall reading an analysis of Integrated Device Technology (IDTI) late last year by Michael Murphy of the California Technology Stock Letter (CTSL) that theorized IDTI could continue to make a profit on the new 16MB SRAMs manufactured in the company's new Oregon fab even if the price dropped to $5. This new fab is just ramping up. I'd guess that the break-even for most memory chip makers is, on average, higher than that in the older fabs
I've seen a lot of ink on this subject lately. Most of it theorizing that the memory makers were, in many cases, selling below cost. They seem willing to canabalize each other because many are stuck with new, multi-billion dollar fabs that have to be paid for, and they have to generate some cash flow even if it meant taking some short term losses.
I don't believe that the cost of the silicon wafers is the real cost driver for semiconductors. In addition, the large number of chips flooding the market right now means that a lot of wafers are going to get sold, even in a weak (price-wise) memory market.
No question, though, if a lot more semiconductor capacity gets taken off-line, then MEMC will be affected. The issue is, will this happen?
The current bull thesis for the semiconductors is that we're in an inventory correction from which we will start to recover during the fourth quarter. The bears feel that the problems go deeper than that, that the glut of semiconductors is mostly the result of overcapacity which will continue to squeeze margins for another year or so.
In either scenario, I still think that wafer demand will be good and MEMC will thrive. Some older fabs may get shut down and some that are under construction will be delayed. But companies can't afford to idle capacity in many cases and seem willing to live on reduced margins as we work our way through the glut.
Bob M. |
