Dave, The arguments that I present are basically supply related. That current capex in leading edge technology is insufficient for total global bit demand in the the coming 12-18 months. The boom/bust characteristics of the silicon cycle is unchanged. The main driving factor is changes in capex, not fluctuations in demand (though that plays a more minor role).
Capacity that was build 12-18 months ago is now obsolete. It cannot produce leading edge products that are demanded. You cannot mass produce a PC-100 64meg SDRAM chip unless you have sub-0.25micron lines, high yields, a chip size of less than 100mm2 and expect to make money. You can count chip makers globally that can do that on one hand. Alot of those lines have been converted to logic and cannot be easily switched into memory without significant retooling and significant cost.
That is where a shortage has developed. It takes time and money to retool lines. More importantly if you do not have the technology, you lose money and you drop out. The Koreans now understand this and realise that they too have to make money. There will be no "flooding" of the the market this time. Making chips below 0.25micron is a different game.
By the way, fact that chip prices have remained steady (slightly rose on the case of high end 64meg) for the last couple of months is an effective price rise. Chip shrinks mean that those that are successful in shrinking and raising yields are seeing their costs drop whilst prices remain firm. Even if prices began to tick down again (which logically, they should considering learning curve pricing) those that can shrink will make money, and lots of it. I include Micron in that group.
Basically, I think MU has more to go, they haven't really announced anything in the way of earnings surprises yet, and I think there will have to be some "survivor premium" on the company. Besides all this optimism seems to be spilling over into lagging tech stocks from the internut plays. I'm still holding long. |
