To: Earlie who wrote (54076 ) 3/31/1999 5:28:00 PM From: Earlie Read Replies (7) | Respond to of 132070
Earlie from Earlie: As most folks on this thread know, I have a low opinion of the "research" that emanates from cheerleaders IDC and Dataquest. Having had a chance to peruse Dataquest's recent commentary entitled "DRAM Market Back In Gear", that opinion has just been reinforced. Herewith a few comments. Dataquest expects the DOLLAR VALUE of Dram to rise 50% in 1999 after falling 27% in 1998 and following a THREE YEAR continuous slide. At first I thought this was a typo, but no, this is their actual, in print expectation. The provided figures are even more awe-inspiring. They expect Dram sales to rise from $15.1 billion in 1998 to $22.6 billion in 1999. I find this amazing, especially when this report is dated March 29 (almost a full quarter of the year has passed, and for the last several weeks, Dram prices have been in free-fall). They also think we are starting a new THREE YEAR growth cycle in Dram. By the year 2001, they see Dram revenues at $59.9 billion. Let me get this straight,.....the Dram industry, which has experienced falling revenues for three years straight and which last year had sales that totalled $15.1 billion, will grow to $59.9 billion by 2001, providing growth of 300% over the intervening 3 year period. Reasons provided for this astounding, positive reversal include rising demand, and stable prices. What words should one employ to denounce this type of shear mathematical lunacy? In the interest of maintaining the "family entertainment" rating of the thread, I'll refrain from using the profanity this poppycock deserves. Data quest should be outright embarrassed by it, but of course it upholds the company's long-standing tradition of foisting dross on the unsuspecting investment public in lieu of responsible research, hence I suspect there won't be any red faces. Here are some of the wondrous "facts" upon which this drivel is based: - "We have not seen prices move much, but there have been indicators since last July that the market has begun to turn". He doesn't cite these wondrous "indicators". Any of you folk see them? - "The fact that prices did not drop late last year, even after suppliers were able to lower their costs through production and design enhancements, indicated to me that demand was already ahead of supply." What baloney! What planet is he from? How does he reach such a stretched conclusion from this observation? Sure prices held up late in the year,....a big over-build was underway, and it is the traditional busy season. What about the rest of the year (which incidentally was a nasty comparison with the preceding year) during which prices fell like bricks? And what about the fact that prices have been CRATERING for weeks of late (64 Mbit under $8.00)? That's down 15% to 20% and we are just getting underway. He (Jim Handy, principal analyst) also notes capital spending has slowed "to a crawl" in recent years (from $20.0 billion in 1998 to $10.0 billion in 1999,....SOME CRAWL). From this he leaps to the conclusion that overcapacity is coming to an end. Too bad he didn't bother to check out the DOUBLING OF YIELD over the last 12 months in many fabs. On the demand side, Dataquest expects "PC sales" (uh, is that UNIT sales or Revenues,.... it is of course the latter, but note the choice of words) to rise 14% this year. Hello out there,....please check this quarter's figures. To be charitable, the recently posted DECLINE doesn't suggest these figures are to be carved in stone. And in spite of spreading global recessions, PC sales will hit 157 million by 2002. Of course. No mention is made of the fact that the channel is still suffering from Q4 constipation and we are almost through March. Also be aware that PCs will continue to absorb more and more memory, (even though there are no new applications and even though most PC sales are at the low end of the price spectrum). He sees a bit growth rate per PC of 70% per annum, compounded. Someone had better tell them out in the field. The part I loved the most was his comments on Rambus. It will represent 70% of that massive new total by 2002. That's right, 70%. This in spite of the memory manufacturers openly suggesting that they don't see the buyers and therefore they don't want to make the required investments in expensive equipment. How the heck will Rambus grow so quickly if only Samsung comes to the party? Ever hear of DDR? As others have already commented, I'll refrain from pointing out the complete insanity of their $1.30 per mbit "stable pricing" FOR THE NEXT THREE YEARS. Looks like they forgot how to extrapolate for this item, while extrapolating everything else to the moon. The 64 mbit chips are already well below this and the year is young. At least they got one thing right,....the Dram boys will be "forced to sell parts at below cost". He thinks they will lose $5.0 billion (on projected sales of $22.6 billion,....YUK). They also see this glorious cycle being short-lived, turning down again in 2002. What a great business. This report does raise an interesting question; How is it possible for some analysts to predict a $200 share price? ("Hello Dan,.....seen these predictions of LOSSES,.... and from a cheerleader to boot? How's the view from Mars?") This piece also goes into my bulging file of IDC/Dataquest excrement. Best, Earlie