Tom:
Please ask Andrew Hay (Reuters reporter who wrote the goose-down piece) to check in from time to time. We need him to listen up.
Loved the way you knocked off that 18% market share nonsense. You should stop being so nasty with those publicly available figures. (g)
I've been working the same problem, but from a slightly different angle, that being the shameful baloney that Kip the Lip put out at the conference.
Start with the "Micron sees 80% increase in industry demand in 1999".
Now lets see. Mu sells 72% of its memory products to the PC industry. Unfortunately, the PC sector's memory needs are not going to be up at all in 1999, but probably down. We both know why.
- "Stuffed channel",.... it will take a while to clear this at the retail end, and more than a while to do the same at the reseller end.
- Corporate sector. IDC, the "cheerleader" says that it is "saturated". It also says that the I.T. guys are not enthusiastic about buying any PCs this year. The crummy numbers and comments from the reseller sector also reinforce this. The corporations buy the big fraction of PCs.
- Of course the "direct channel" could be "growing like a weed", but Dell actually dropped a bit of market share globally, so using Dell as a proxy, it doesn't suggest that a massive growth spurt will materialize here.
- The box builders don't like to see the percentage cost of memory rise as a percentage of the total box, and it hasn't done this anyway of late. With reduced PC ASPs (and reduced overall global PC revenues) likely in 1999, the pricing pressure from this source alone will be significant.
- Yields are rising fast. Several market participants anticipate a doubling of output this year. New 128 mbit chips are already being sampled. DDR chips will be available within the next few months. It's hard to see pricing doing anything but falling, (which it is doing as we speak,...64s @$9.50 to $9.75 in spot, Mu reported to be selling at around $9.65) . My expectation is that the 64s will sell at $5.75 to $6.50 some time during the Summer.
So where's the 80% rise in industry demand come from? Even stuffing every PC with twice what it really needs (and there is no viable reason why this should occur), one obtains at best, a desultory percentage point or two rise in PC memory mbit demand. Given the pricing pressure, this translates to falling revenues This means that the other sectors (28% of MU's production) must experience a rise in demand enumerated in orders of magnitude, and in less than a year. Do you see anything on the horizon that needs big memory increases? Me neither. (g) Does Kip think silly assertions such as this one will actually be believed? My neighbours budgie can see through this sleight-of-hand slime.
The firm is going to "take its share of the computer memory market to as high as 25% by August." Sure it will, and my grandmother will swim across Lake Ontario next week using only her left arm. As you point out, MU is 10% at best of the market at the moment, so we are going to see the company raise its market share by 150% in the next six months. Wow,...more than one percent per business DAY! I'll watch for that. Perhaps the company has acquired some laser-guided ordinance, and will knock out most/all of the plants operated by Toshiba, LG Semi, Hyundai, Samsung,etc.
Aside from the 80% rise in demand, the old Kipper sees only "70 percent rise in supply". Uh, Kip, you might want to revise that figure downward a bit before the public gets its hands on it, and plugs that figure into the PC/1999 equation. It, uh sort of suggests a,.... ummm,....... glut. Also, Kip, get those guys operational with the weapons soon, as those #@$^%# Orientals just keep on improving the yields.
According to Kip, memory prices were "stable,... actually improved since November". He didn't lie, but it sure didn't last long did it?
And Kippy, how great a yield improvement does the company expect over the next few months? Enough to offset another round of price cuts? Enough to offset the impact of sales siphoned off by those new 128s? And what if they have to go into inventory because they don't sell at all?
Not to muddy the waters, but I find it interesting, that in light of the previous week's revelations,
- Mr. Bedard said not a word about "errors" in the company's reporting.
- Those TXN plant problems? What problems?
A closing note:
The company is hinting that those test problems are solved. Perhaps, but does it matter? Sometimes it is what is not stated that tells the story. To my knowledge, no info has been provided that deals with the TXN plant difficulties. The former is a problem in the backyard, the latter is a (multiple) disaster in faraway lands.
Best, Earlie |
