It's valuation and macro-analysis time for Micron.
Valuation is the net present value of the profits in each of the years going forward. The net present value is obtained by applying a discount rate to the estimated profits in each of the years going forward. For example, if we apply a 20% discount rate to $2.00 in profits one year in the future, the net present value of those profits is $1.67 ($2.00 / 1.2). If profits two years in the future were $3.00, and if we assume a 20% discount rate in each of the next two years, then the net present value of that $3.00 today would be $2.08 ($3.00 / (1.2 x 1.2)). This is old hat for some people, not for others. This is the way companies are valued.
Why is it relevant to remind people of this? Well, with a company like Micron that sells what appears to be a commodity, it's very tempting to focus discussion of its stock price on what the current pricing for DRAM is. That does, after all, have an effect on their near-term profitability. And that near-term profitability should indeed have an effect on the stock price.
However, don't forget that a true valuation of the stock includes summing up the net present value of the ENTIRE future stream of earnings--not just the current quarter's earnings.
Let's do some macro-analysis of Micron's potential for future earnings beyond this current quarter.
First, Micron's share of total DRAM megabyte shipments has gone from 5.7% in Q3/95 to 11.7% in Q2/98. During this time, overall industry DRAM megabyte shipments went up over 4 times. So, Micron's trend lines are going in the right direction here.
A follow-on question would be whether the world is going to be able to soak up the DRAM that will be produced over the next few years. One of the big users of DRAM is the PC industry. In Q4/98, the average base memory in shipping PC's was 73MB. (Side note: this is really amazing when you remember back to early 1994 when the base memory in a PC was 6MB.) Base memory in 1999 is projected to be 102MB. In the year 2000, it's projected to be 145MB. CAGR from 1996 to 2000 is (projected to be) 72%. Yes, prices are dropping, but at the same time, demand is soaring. So as long as Micron can keep grabbing more market share, their units shipped and their total MB shipped should burgeon over the next few years.
(And lest you think that PC's are all that will be driving memory growth, just remember these other users of semiconductor memory: telephones and pagers, automobiles, set-top boxes, HD TV's, and other consumer electronics. Also, within the PC market, drivers of memory growth will include: an accelerating upgrade cycle, memory-hungry applications, Windows NT, Internet/Intranet servers (!), DVD, increasing graphics and 3D graphics, and more sophisticated games.)
Second, yes, it is better when DRAM pricing doesn't go down that rapidly (or even goes up temporarily.) The fact is, though, that the long-term trend is DOWN in DRAM pricing. It's the nature of the semiconductor beast. What's important in Micron's and every other semiconductor manufacturer's case is whether costs are going down more quickly or more slowly than prices are. After all, what we care about are earnings, and earnings equal Price minus Cost. It's ok if Price goes down, as long as Cost goes down faster. Right?
Third, when it comes to cost, a key cost driver is how fast a company can successfully shrink die size so that they can harvest more saleable chips off their wafers. One way a company does that is by shrinking the die size while at a given process technology (say, .21 micron). Micron Technology does extremely well at this, and they are famous for being aggressive with their die shrinks. Another way a company does this is by moving to smaller process geometries, as Micron is currently doing in moving from .21 to .18 microns. This will also yield more saleable chips.
Fourth, it takes capital--and lots of it--to build and debug new semiconductor fabrication lines. Micron has that capital, particularly with the $500 million infusion from Intel a few months ago (which essentially is money that Intel paid to Micron to do things that Micron would likely have done anyway.)
Micron's competition overseas in Japan and Korea either do not have the capital, or they don't have the will to stay in this ever more expensive game, or they are missing both. (See excerpt from TechWeb article, included at the bottom of this post.) Once a company is out of the DRAM game, it is very difficult and very expensive to get back in. The fact that DRAM-producing companies ranging from Fujitsu to NEC to Hitachi to Oki to Mitsubishi to Matsushita to LG Semicon to Hyundai have ALL cancelled or postponed indefinitely their plans for one or more fab lines is a STRONG indication that they are not going to be able to stay profitable long-term in this game--or even stay in the game long-term. Toshiba and Samsung have both postponed fab expansions but only to later 1999 and 2000. By the way, once you turn off plans to expand, it's not as quick as turning on a light switch to get the plans back on-line again.
Longer-term, it's fairly clear at this point that there will be a consolidation in the marketplace, and the surviving players will likely consist of Micron, Samsung, and at least one Japanese player, probably NEC.
So, bottom line:
1. Longer-term, Micron will be one of the three or so big players in DRAM.
2. Micron is likely to be a profitable player in the market, based on their historical profitability performance.
3. DRAM demand is going way up over the next few years.
4. Micron will likely make a _lot_ of money as demand ramps up (i.e., industry prices don't fall as rapidly as industry costs) and as they are able to drive their own costs down (through aggressive die shrinks and capital investment in smaller geometry fab lines.)
5. The stock price will likely head up, as predicted by Robertson Stephens, CS First Boston, and CE Unterberg Towbin. Will it be a smooth ride up? Probably not, based on the stock's historical performance, but I do expect that within a few short months and also within a year, we will look back on this time as a historical buying opportunity in Micron.
6. The big picture on this is that the company is in the process of FUNDAMENTALLY CHANGING THE INDUSTRY STRUCTURE and its place within it. This more than justifies a sharp appreciation in valuation.
Thoughtful and bigger picture (i.e., not "oh, the price of DRAM just dropped a nickel! the sky is falling!!") replies would be appreciated.
slow-thinkingly yours,
---------------------------------------------------------
From: techweb.com
"Samsung said it has no near-term plans to add new fab capacity, and discounted a report in the South Korean press that the company would renew construction on Fab 9 in Kiheung, which was postponed when DRAM prices crashed. However, Samsung will double output at its Austin, Texas, fab later next year to 25,000 wafer starts per month when it completes the second stage of its construction plan there.
The relentless DRAM-production contest is fast outstripping the ability of other chip makers to keep pace, said observers. Japan's DRAM companies, for example, are also raising 64-Mb output. But even that country's largest producer, NEC, is capable of only about 10 million units per month.
Some analysts believe such a production disparity could change the DRAM industry's market dynamics, as a few high-volume producers return to profitability by spreading costs over a far broader base of chips than their smaller competitors."
------------------------------------------------------ |
