MB, Joseph's change of heart is interesting. He seems to be embracing some kind of new industry "gestalt" all of a sudden. IMHO, when you actually examine his observations and then his conclusions, there's little logical glue that holds his ideas together other than this appeal to "gesalt".
What makes it even more interesting is that his last prior report on MU was published only yesterday. So you could actually attempt to break down his ideas into an informal logical argument - yesterday's argument and today's and see specifically why the conclusions vary.
Start with pricing.
From yesterday's 7-13-99 report:
In the past week, prices for 64Mb DRAM have increased by about 30% in overseas markets and by about 15% in the U.S. to about $4.90-$5.00, which is still below contract prices, which is currently about $5.50. Brokers in Asia over the last couple of days tried to raise prices to $5.25-5.30, but so far have found few buyers.
From today's 7-14-99 report "MU: It's More Than Micron"
In the past 10 days, prices for 64Mb DRAMs have increased by about 60% overseas and 35% in the US to about $6.00 and $5.70, respectively.
An interesting difference in his suppositions re: DRAM prices.
7-13-99 Possible Explanations for Price Uptick:
1) a stronger than expected pickup in seasonal demands for personal computers
2) difficulty in transitioning to a new generation of shrinks
3) Industry cources believe that the rate of the price increase has been exacerbated by overbooking from OEMs and distributor inventory stocking, in an effort by OEMs to lock-in bargain basement prices
7-14-99 Possible Explanations for Price Uptick:
1) Medium-term demand...for personal computer components picked up strongly in June following nearly six months of slowdown. Much of this can be credited to a sharp increase in seasonal demand-it was just at this time last year that microprocessor and DRAM demand started to take off.
3) Worries of DRAM production problems. We believe Micron has had some yield problems in its shift to 0.18 micron process, but nothing out of the ordinary. Nevertheless, the company is devoting more wafers to Rambus DRAM, to DDR DRAM, and other next-generation product, while at the same time the overseas fabs come up a little slower than we would have forecast...net impact is Micron's bit out are expected to increase only about 15% per quarter over the next several quarters, rather than the 20-25% expected by the market (but not put into our model).
3) We also believe buyers became complacent and let inventories get too low, which is now resulting in some panic buying.
Not that much difference.
7-13-99 Expectations
1) Micron and other DRAM makers, still sitting on 6-7 weeks of inventory, often come into the market as prices rise
2) Also as manufacturers get back on track with their die shrinks, a new wave of product could be entering the markets this fall.
7-14-99 Expectations
1) Still some inventories to work off...Micron, Siemens (Infineon), and Hyundai, all of whom it is said have built inventory, are likely to begin selling as prices rise.
Not that much difference.
The difference is really in the conclusions.
7-13-99 Primary Conclusion:
This confirms our belief...that prices will stay in the low-to-mid $5.00 range over the next few weeks, and that prices have begun to flatten out again before possibly slowly trending downward...Although we are encouraged by the recent turn in prices, we are concerned the upward trend may not be sustainable.
7-14-99 Primary Conclusion:
We expect the price recovery to flatten off somewhat around the $6.00-$6.50 range, perhaps as soon as next week...Even so, the strength of the upturn suggests to us at least six months of sustainability as the PC segment heads into its traditionally strong season.
On their own, each would seem to make sense.
Yet, these were written by the same person on two consecutive days.
One can question the validity of either set of conlusions since the arguments when compared seem to negate each other. Where's the logical glue in the 7-13-99 (neutral) report if the writer can use essentially the same sets of premises and come to the opposite conclusion? And vice versa?
But I guess when you turn on a dime overnighht, it's tough to connect the logical dots. Which is why this appeal to the greater forces of an industry gestalt would make sense. You don't have to connect the dots. You just paint over everything.
Anyone who thinks I'm being unfair, can sign up for a trial at multex.com and make sense of it for themselves.
Good trading,
Tom |
