Just a better posting, so I can rad it !
Hal
................
I am the analyst at Needham that covers the semis and was interviewed for the Barron's article on Micron. Since my views were counter to the theme of the article, I'm not surprised that they were not used. Readers of this thread have some interest in a somewhat different analytical approach from the apparent mainstream, so I'm attaching sections of a recent Research Note that we put out on Micron (it may be a little difficult to understand at first, but if you hang in there, it should become clear):
We understand that recent management guidance at investment conferences has been for 2QF99 bit growth to be substantially higher, with some reports stating that 40% or more may be an attainable goal.
We believe that the underlying arithmetic may be eluding investors, who are likely to be mesmerized by the size of the numbers being bandied about. A simple model, which we offer below, may be of help in understanding the dynamics of the situation.
We are assuming that Micron had a 12% share of the market in its fiscal fourth quarter (the base quarter). The market has been growing at a relatively steady rate of about 95% on an annual basis, which is equivalent to 18% on a quarter-to-quarter basis, so we grow the non-Micron segment of the DRAM market at 18% each quarter in both production and shipments. For Micron, we take 1Q production up 10% and shipments down 10%. This leads to an increase in inventory equal to 20% of the base amount of production. For the remainder of the year, we grow Micron production at 21% to enable its annual production to grow at the 95% rate enjoyed by the rest of the market. We also assume that all the inventory is shipped in the second quarter, so that for the rest of the year production and shipments are equal.
Table 1: Micron/Industry DRAM Model
(no scale, arbitrary units with industry base level set to 1,000 and Micron base set to 12% of industry)
Base 1Q Q/Q 2Q Q/Q 3Q Q/Q 4Q Q/Q
MU
Production 120 132 10% 160 21% 193 21% 234 21%
Shipments 120 108 –10 184 70 193 5 234 21
Inventory 0 24 0 0 0
Rest of Market
Production 880 1,038 18% 1,225 18% 1,446 18% 1,706 18%
Shipments 880 1,038 18 1,225 18 1,446 18 1,706 18
Inventory 0 0 0 0 0
Total Market
Production 1,000 1,170 17% 1,385 18% 1,639 18% 1,940 18%
Shipments 1,000 1,146 15 1,409 23 1,639 16 1,940 18
Inventory 0 24 0 0 0
MU/Market 12% 9% 13% 12% 12%
This leads to the following observations:
1) Merely getting to the industry rate of production growth and recapturing the unshipped inventory would take Micron's 2Q shipment growth rate to 70%. A report of shipment growth below that level creates a question of share loss.
2) Gross margins in the second quarter would reflect first quarter costs for 13% of the DRAMs shipped (24/184), so if they are improving, not all the benefit will be seen in the quarter.
3) Such a shipment pattern, if attained, would set the stage for a sharply lower growth rate in the third quarter, unless production rates are able to growth faster than the market (which is entirely possible).
4) The company's goal (as reported by Reuters) of 25% market share by the end of the year is specious. To get from 13% share in 2Q to 25% share in 4Q would mean that Micron's shipments would have to increase by X where (184+X)=.25*(1,706+184+X). Solving for X shows an increase of 385 in shipments from the second quarter. But with 2Q shipments including the 24 units of inventory from 1Q, the production increase would have to be from 160 to 569 – a 255% gain in six months. We don't think it's in either the cards or the wafers. Additionally, doing so would take the entire market's shipments to 2,275, which is 128% growth in a market that is perceived to be in balance when bit growth runs at 70-80%. The paradox is that unless the rest of the producers stop growing, Micron's incremental production to reach its goal will be sufficient to destroy supply/demand balance and lead to a resumption of severe price declines. This would make its "victory" entirely Pyrrhic.
-----
I hope that this is helpful to those who wish to approach the situation in an unemotional a manner as possible (unlike the jihad that this stock often acts like). I have been a big believer in the semi cyclical rebound since last summer, and had 11 Strong Buys on 15 stocks under coverage as of last Labor Day. But I have felt (incorrectly, so far) that the ongoing supply additions in the DRAM market would cause investors to delay giving MU additional value until concrete evidence of sustained profitability was in hand. As shown above, my numbers still don't suggest a decent rate of return for several more quarters and I am amazed at how willing investors are to pay up so far in advance for alleged profit increases.
Tad LaFountain (aal3@idt.net)
(Sorry about what appears to be the formatting, but I wasn't able to get the table formatted without the text failing to wrap)
|
