An interesting thought occurred to me. First of all, clearly you are right that MU has huge control over pricing, and the huge ramp they are planning will probably adversely affect prices. Niles made the statement that falling prices are good for MU. This is too bizarre a thought for Niles to have come up with on his own. Interestingly I wonder if it reflects the thinking at MU, as voiced through Niles. Upon reflection I believe that MU could well intend to drive the price down both at 64mb and at 16mb.
Why would they want to do that, you ask? First of all, they are much further along in converting to small geometries than most if not all competitors, though Samsung is in the same range and Hyundai is probably not far behind. Almost everyone else is 6 months to a year behind. Six months ago there was despair in the the DRAM industry as all players were losing money. Suddenly you see more cheer, and players from Nan Ya to Japanese are announcing plans for re-investing in the DRAM business. Certainly if they do, MU will be hurt in the long run.
Therefore the best plan for MU may well be to scorch the earth. Unlike their competition, they have the dollars in the bank to keep up this pace of investment for at least two more years, and perhaps three even without making money. There are plenty in the industry who are not sure they want to keep plowing money into this industry, but who will if they think they see profit around the corner. If DRAM prices fall to the $5 range for 64mb, how many players will blink and say, "ummm maybe there is something else we can make in our fab?" Or will some players simply run out of money for investment?
MU is in a position of strength right now, with both strong technology and plenty of cash. If MU can push aggressively to .18µ by the end of the year at both 16mb and 64mb they can once again take the profit out of the industry. If everyone runs at once to 128mb (which is what it appears they are planning to do), they will be worse off than if they stay. A 128mb chip will cost more than twice as much as a 64mb chip to make, and if ample supply is available at 128mb, the 128mb chip will sell at an equivalent price per bit, and will thus be less profitable. Of course if only a few people move to 128mb, it could be more profitable, but it doesn't look like that is going to be the case.
Of course if RMBS takes off, all bets are off and a DRAM shortage will be here sooner, rather than later. This will keep all the players in. MU will be profitable too, but it will make for a shorter up cycle. If MU can bury the competition now, the next up cycle could last quite a bit longer as people will think twice before taking on MU, Samsung, and the other survivors. And make no mistake, there will be another up cycle, and both MU and Samsung will be there to see it.
Carl |
