*** Pricing and Cost info on MU 64MB DRAM's ***
Ok, with a title like this, Skeeter, I suppose I have to deliver some actual numbers:
In Q2/99 (ending in February), Micron's estimate of "ASP for 64Mb equivalent (i.e., 64Mb DRAM) is $8.85, which is down from $8.96 in Q1. Estimate for Q3 (ending in May) is $8.80, and for Q4 it's $8.60.
Cost of goods sold for 64Mb equivalent over those same quarters were/are: $7.91 (Q1), $6.33 (Q2), $5.06 (Q3), and $4.05 (Q4).
The fully loaded costs of those same 64Mb equivalents over the same periods were/are: $10.68 (Q1), $9.07 (Q2), $7.42 (Q3), and $6.05 (Q4).
I'm kind of a SlowThinker when it comes to this stuff, but it does seem to me that the profitability picture starts looking _much_ better beginning next quarter. Let's look out into FY2000 now:
Micron's estimate of "ASP for 64Mb equivalent (i.e., 64Mb DRAM) for FY2000 are estimated to be: $7.91 (Q1), $7.20 (Q2), $6.62 (Q3), and $6.09 (Q4).
Cost of goods sold for 64Mb equivalent over those same FY2000 quarters are estimated to be: $3.64 (Q1), $3.21 (Q2), $2.83 (Q3), and $2.49 (Q4).
The fully loaded costs of those same 64Mb equivalents over the same periods were/are: $5.55 (Q1), $4.99 (Q2), $4.48 (Q3), and $4.02 (Q4).
So what you were saying below, Skeeter, that within 12 to 18 months the cost of one 256Mb DRAM chip will be less than 4X the cost of a single 64Mb chip is hard for me to place credence in. Four times $7.20 is $28.80. While I do believe that 256Mb DRAM's will come down rapidly in price, I have a hard time believing that they will be down in the sub-$28 price range by then. And even if they are, it's even harder to imagine Micron not fielding a 256Mb product once it's economically beneficial to do so.
More important to note from all the numbers above, though, is that the profitability picture is going to change drastically in the coming quarters. Specifically and based on the numbers above, EPS will go from about $.70 this fiscal year (1999) to at least $2.70 next year. Semiconductor gross margins will go from 29% this quarter (Q2/FY99) to 43% next quarter to 53% in Q4/FY99. Next year, the numbers will begin in the mid-50%'s and trend up toward the high 50's.
Might the stock still trend downward somewhat this next week or two as people focus on spot market prices that may reflect some seasonal weakness? Sure. But by the way, is this seasonal weakness appearing to be less than it usually is? Yes. And are all the right medium- and longer-term trends pointed in the right direction for a _strong_ move up in MU's profitability, EPS, and stock price. Yes, for a third time.
To me there are two ways of looking at this.
The first is to just look at the strategic picture and not worry too much about numbers. That's what a lot of folks are doing with Internet valuations these days. From this standpoint, Micron looks like a shoe-in to be one of the top 3 DRAM producers worldwide (with Samsung being one of the other two; NEC for the third slot?). If we assume that management is smart enough to figure out how to make a profit at it, then they are in very good shape.
The other way is to just do a valuation based on the net present value of the stream of future earnings. If you believe that we are at the start of a strong DRAM cycle and that the $.70 in EPS this year will be followed by $2.70 next year and even more earnings growth in 2001, then this method also yields a much higher stock price than its current $50 price.
This is what is driving the $200 three-year target by Roberson-Stephens, the $70 target by CS First Boston (do I have this number correct?), and the $80 target by Unterbein Towbin (number correct here?).
This is a big picture opportunity here. If you focus out beyond this week's (or tomorrow's) spot DRAM prices, it's a compelling investment. If, however, you want to focus on spot pricing and issues like "how will Micron's current 64Mb chips ever compete a year or year and a half from now with the 256Mb chips that Samsung will be producing then?", then I think you're missing the big picture or focusing on non-issues, or both.
[The numbers are courtesy of Robertson Stephens' recent analyst report. If you have substantive disagreement with the numerical analysis, it would be constructive to offer specific points of disagreement, as opposed to personal diatribes on any particular analysts. Whether you believe it or not, these folks generally sweat over getting the numbers as right as possible.]
Slow-Thinkingly yours,
>actually, it was more like 6 months. btw, stop looking to the
>past ;-) within 12-18 months, the cost per bit for a 256 mb chip
>will be less than 4 64 mb chips. people won't pay more, david,
>no matter how much you think they will. |
