Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Micron Only Forum -- Ignore unavailable to you. Want to Upgrade?

To: DavidG who wrote (43941)	3/17/1999 1:08:00 PM
From: TREND1	Read Replies (2) \| Respond to of 53903

David you wrote As a result MU company is in a much better position than most of their competitors... David A better position on a sinking ship (g) Larry Dudash

To: DavidG who wrote (43941)	3/17/1999 2:37:00 PM
From: A. A. LaFountain III	Read Replies (2) \| Respond to of 53903

Re MU chip cost "IMO MU is still in the best position in this speculative DRAM commodity sector with a cost per chip now around the $6 level." This gets to the crux of the matter. You qualify your statement by acknowledging that it is your opinion. To the best of my knowledge, MU has never claimed that it is the lowest cost DRAM producer, only that others say it is. I've never seen any hard data that details the costs by producer. What we're dealing with is glorified conjecture. This not to say that the claim isn't correct. It's just that without underlying data points, it's not right for us to give it complete credence. Forget the horse - my kingdom for hard data! - Tad LaFountain P.S. Is the $6 cost figure a blended average for 16Mb and 64Mb parts? Here's why I ask - the 10Q states s/c revenues of $409.5 million and the MD&A states that 64Mb parts generated 65% of s/c revenues (=$266 million). In the conference call, 64Mb ASPs were said to be about $9.50, so units can be derived of 28.0 million. Taking out a small amount of SRAM and flash from the s/c total leaves $127 million in revenue from 16Mb and less. Now s/c gross margin was 9%, so s/c CoGS would have been $373 million. If the 64Mb part had a cost of $6, then 64Mb CoGS would have been 28.0*$6=$168 million. Non-64Mb DRAM s/c CoGS would have been $373-168=$205 million, which gives a non-64Mb DRAM s/c gross profit of 143.5-205= -$61.5 million or a gross margin of -$61.5/143.5 (which is an unappealing [and unlikely] -43%). If I take the $373 million in s/c CoGS and divide it by 85 million total units (28 million 64Mb DRAMs plus 55-60 million units of 16Mb and other, I can get to $4.40 cost per part on a blended basis. Now, that factors in a $2.25 ASP for the 16Mb and below parts, which may be too low; increasing that ASP by 50% to $3.375 takes the total units down to about 65 million for a blended ASP of $5.74. If this is along the lines of how you derived your $6 estimate, what costs do you ascribe to the 64Mb part and what to the 16Mb (because a simple reading of your post might lead to the conclusion that MU has a $6 cost for 64Mb parts that sold for $9.50 in 1QF99, and we wouldn't want people to get the wrong impression). - Tad LaFountain

To: DavidG who wrote (43941)	3/17/1999 3:23:00 PM
From: Earlie	Read Replies (2) \| Respond to of 53903

David: <S.E.Asian Dram producers in trouble>. Care to compare the large profit recently posted by Samsung with the $570.0 million operating loss posted by MU over the last year? (g) <MU then became the company with the ability to quickly make cap expenditures for upgrades> How's that? By its own admission, MU expects that it will have to spend "up to $1.0 billion" (10-Q) to upgrade its plants this year, with much of that relating to the TXN plants. MU people have quietly commented that those TXN plants are a mess and that they are well behind the technology curve. (i.e. expensive products). The company has about $1.0 billion in cash, but it also has much more than that in debt. It's also been losing a bunch, and has taken write-offs. Its debt paper has been downgraded, and dram prices are falling again. I doubt that they would run the cash down to zero or even close to this under the circumstances as it would be corporate suicide. I also doubt that they will spend the required money on upgrading, given that INTC might not be happy (that $500.0 million was primarily for Rambus) and given that doing so would merely exacerbate an already over-supplied market. Then too, some of that dough needs to be kept on hand to pay those pesky interest bills (which will require more than $100.0 million this year). So where does the dough come from for the upgrades? Remember that the corporate junk bond market is in lousy shape and with the current bloated share price, heavy insider selling, heavy debt, and non-existent earnings, nobody is going to buy their equity. <MU in a much better position than its competitors> I cannot follow this argument in any way. The dumping charges may or may not work this time for MU. Korea, which is the only Asian country to have shown even a tiny bit of progress at climbing back out of the black hole, depends almost exclusively on chips for its export dollars. If this gets cut off, they slide back down. The U.S. has so far seen fit to act as the buyer of last resort for Asian (and other) countries trying to regain their footing. I suspect that the U.S. would be loathe to slam the door on the only country that has demonstrated a bit of progress, although the pressure to so do is growing (witness steel charges against the Asians, Brazil, and Russia). We'll see. Re your 16/64 cross-over comments. The fact is that 16 Mbit chips enjoyed the shortest period of viability to date. It appears each new rev. has a shorter life span. Now how does this help MU again? As you point out, the big bucks are usually earned by the early deployers, not the slow-to-the-party players. Does this not have something to do with margins? The fact that MU lost a bundle last year (as some of us accurately predicted) sort of supports that idea. You are dead wrong about the driving force behind the acceptance of 64 Mbit chips. The Asians moved quickly into them when the margins started to free fall in 16 Mbit chips. They wanted to regain margins, so they moved well before there was any "pull" from the PC sector. While I agree that many boxes are being populated with more memory, the actual need for this is considered a joke by the industry. Aside for games, there simply isn't a need for a major increase. In the past, more memory has been required by new apps. Unfortunately, new apps haven't come along. I'd sure as heck like to know where you get that $6.00 cost per 64 mbit chip figure from. It makes no sense. If it had been anywhere near this, the company would not have lost so much money. Did it just come about in the last few days? I recall you and I (and Skeeter) having a similar debate last year. I said they would lose money (I was wrong in suggesting that they could lose $100.0 million per quarter,....on average it was higher), based on a much higher all-up cost per chip figure than you thought possible. Remember your comments of the time? You might wish to fill me in on whatever lesson you thought the Asians might or might not have learned. I'm at a loss on this one. Re: Micron's contract prices be higher than the spot market prices? The fact is that at times they have sold product below spot (to move product before a quarter end). Also,with spot prices declining, it's tough to get a premium price. And whether it is slightly higher or lower, does it matter? In the end, a company is supposed to make money. <By the end of the year, a 124 Mbit system will be the norm> Shall we bet on that? This is your hope or dream, not a quantifiable statement. PC prices have been cut dramatically during the last 18 months, yet household penetration has been minimal (up less than 10%). And you say that all houses will have two or three near term? I'll refrain from asking for at least something to support this contention. Should not some of your statements be labelled as opinions? One last thought. The stock is at $52. There are over a quarter of a billion shares outstanding. To support this price, one needs some sort of future earnings. Provide us with some earnings projections and some arguments to back it, given a world-wide PC sale situation that is unlikely to exceed last year's 89.0 million (and still falling) estimate. Yes, I know IDC thinks there will be a 14% unit increase this year. Use that if you want. Have fun. No matter how you stretch it, a plausible earnings projection just cannot be created. Maybe we'll have some new massive memory uses pop up, like DVD, etc. (that's a joke Dave). Best, Earlie

To: DavidG who wrote (43941)	3/17/1999 8:23:00 PM
From: PAinvestor	Respond to of 53903

Korean and Japanese producers have followed a business model pioneered by Toshiba whereby they move to quickly ramp up next generation chips when the penetration is still only in high end servers etc. where significant price per bit premiums can be secured for a time (6-12 months or so). Current 128meg prices are still over $20 last I checked (anyone have any updates?) but costs are around $15-16 at the top Japanese makers and dropping fast. The Japanese have quite a lead here over the Koreans and MU. In fact they should produce over 50m units of 128meg this year ahead of the Koreans which should produce about 35m or so. MU unfortunately will not even be close to reaching these figures or even be in volume production by year end. As Fabeyes has stated they haven't even shipped samples yet. Hopefully their 64meg shrink will be successful enough to make up for what they are missing out on the 128meg. That is their strength, but again, I have yet to hear how successful their use of phase shift masks is proceeding given that their exisiting ASML steppers have sub 0.6 NA's. Has anyone heard anything further of yields?