Sridhar: Thanks for posting your comments, and delighted to note your views. The fact that we may not agree is what makes the interchange valuable. With respect to your queries:
#1.
- Memory producers have estimated that it requires $600.0 million to $800.0 million to set up new production capability at 64 Mbit or greater level.
- Note the size of MU's expenditures before they shuttered their new plant, and note that it was both empty and a long way from completion.
- Mu has stated that much of the recently raised money is for working capital purposes, and since the total sum raised was $500.0 million, it doesn't leave much for major new construction.
- Industry estimates currently suggest that 80.0 million 64 Mbit chips will be delivered this year. Prices for these chips have already fallen to $35. to $40. This rapid price erosion results from OVER-SUPPLY, and PC producers haven't really adopted them yet.
- Some memory producers are already delivering 64 Mbit chips from actual production runs, not "sampling" from their laboratories. Since MU is not so doing, their competitors enjoy a significant chronological advantage.
- I don't think you would argue that PC sales growth is not falling, and that this sector soaks up the majority of memory production. Unfortunately, a large number of new fabs are about to come onstream, promising further disruption of an already oversupplied market.
- I have some difficulty in accepting your comment that "MU's performance seems to indicate that they are ahead of the competition in terms of execution". The company has historically been a follower, not a leader in both product innovation and adaptation of new production processes (note that they have only this year made a full transition to 8 inch wafer). Spending hundreds of millions of dollars on an unfinished plant doesn't speak well of execution to me either. Additionally, given the remarkable profits the company reaped two years ago, decent execution would not see them cash deficient, and back at the junk bond trough, dependent on its successful placement for their very survival. From my perspective, MU was plain lucky this Spring in that the Koreans held back a ton of inventory as a part of their strategy to combat the "dumping" charge, providing MU with a temporary sales window. The Koreans can't afford to write this down or off (they've already experienced serious GDP problems as a result of poor semi export revenues and pricing), and they appear to be re-entering the lists. We'll see just how good MU's "execution" is as the pricing noose is tightened. 16 Mbit memory will shortly blow through $5. (Asian analysts expect $2. to $3.50 this year).
#2
- Find me a current PC or a current software program that really requires more than 32Mbit to function properly, and I might buy your argument. I find it ironic that you label my contention a "subjective" viewpoint, and then cite SIA's 30 to 50% bit growth PROJECTION (nothing but an extrapolation) as proof of same, and then add your own unsubstantiated opinion that you "think they are conservative" (not subjectivity on top of subjectivity?). You might also note that the average PC sale price point has dropped precipitously since September (from $2,000. to $1,000), and PC's are no longer going out the door with more memory (many low priced units sell with 16Mbit). No known laws of man or nature guarantee the continuity of extrapolated lines, and for the near term at least, memory requirements, will not continue to grow as they have in the past. Given pricing, this is a moot point in any event......who gives a darn about BIT growth when each chip contains order of magnitude more bits BUT SELLS FOR LESS THAN ITS PREDECESSORS, AND AT LOWER MARGINS, which is a fact in the current market.
- If you can show me current serious demand for these larger memory chips, or construct the heavy demand scenario from the current reality, I'm all ears. I sure don't see it in the field, and the plunging prices state an obvious opposing point of view.
#3
- Two years ago memory prices ROSE, (an anomaly, given the typical 30% decline in memory prices), then last year they FELL 80 to 85%. The rise created hoarding, and massive over-ordering by the box builders, which in turn gave impetus to the largest fab capital spending binge in history. I quit counting when the fiftieth new fab was announced (not that I think they will all be built or completed), and they are coming on stream like Dandelions in my lawn. Too bad the construction plans were drawn up so quickly. If my contention that the "supply side is endless" seems an overstatement, wait a few months as the new plants gear up. As an aside, if my contention is not accurate, why is pricing so weak, and getting weaker?
- MU "consistently ending up in the black" doesn't jibe with the company's financial records. (Dell has also had its past financial boondoggles). Why the heck are they tapping the junk bond market and selling chunks of their subsidiary if they are doing so well. Additionally, if I wasn't a gentleman, I'd suggest that their accounting practices of late are "picture painting", and designed to help bring in the needed dough.
- I can't recall ever saying that "there are no margins left anywhere". What I have endeavoured to do is to point out the simple facts that demand is slowly falling , that (much more importantly) supply has and continues to outstrip demand, and that this is eroding margins. This is a fact, whether investors wish to ignore it or not, and is patently obvious to anyone who digs into the business or who reviews the numbers. This quarter's pre-announcements have been consistently ugly, have underlined this erosion of margins and have been largely ignored.
#4
You crack me up with this one. MU successfully converted its lines from 4 Mbit to 16 Mbit. Ask them if they can convert those same lines to 64 or 256. Ask them what NEW production processes will be required to make the jump, and at what cost. Ask them how much time will be required to accomplish this. In fact, company officers have reluctantly spoken to these issues in telephone conference calls, and left listeners worried. If you think they have the expertise (Lowery left), technology (where, certainly not in house yet), or management (I'll refrain from comment on this), then post your proof. Certainly, they don't have the capital. The business requires big working capital bucks, and MU has a crummy balance sheet. The recent junk bond issue does little more than keep them alive, meanwhile there is that little matter of interest, (which of course can be deferred, I wonder why this is required?)
#5
"Memory, for all practical purposes, is a perpetual demand game".....(and I'M SUBJECTIVE?) If that were true, why are the prices so weak? It doesn't matter how efficient a company is, if that company can't price their products to reflect cost-of-production, plus margins, then it won't make a profit. MU's pricing is totally at the whim of the spot market, and that market is heading south because of oversupply.
- See earlier statements re company's "adroit use of resources", especially need for junk bond financing.
- Speaking of subjectivity, I can't wait to read SIA's reasons for stating that "the next boom should start in early 1998".....let's see that's 6-9 months from now.....utter nonsense, given current supply situation.
Boy, looking back over this note, I really got the old keyboard in gear.
Earlie |
