Today's Fribble - 01/17/1996 - Is the DRAM Slide Good for Semis?, by MF Chips

Wednesday, January 17, 1996
Is the DRAM Slide Good for Semis?
by MF Chips

We all know by now that Rick Whittington recently downgraded a host of semiconductor stocks based on his theory that these companies would enjoy "profitless prosperity" in 1996 as revenues increased, while profits remained flat. According to the January 11 MF Evening News, Whittington based his conclusions in part on the dramatic shift in the DRAM market which caused shares of Micron Technology <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:MU)") else Response.Write("(NYSE:MU)") end if %> and Texas Instruments <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:TXN)") else Response.Write("(NYSE:TXN)") end if %> to tumble in recent weeks.

Let's explore what the DRAM pricing pressure means to some of the companies in this sector. Specifically, let's choose a fictional non-commodity fabless semiconductor company, we'll call "Sparky Semiconductor." Sparky is a designer of extremely proprietary, whiz bang communication chips used by the likes of Cisco, Bay Networks, Synoptics, 3Com and others. In fact, Sparky owns a significant market share of this specific (fast-growing) whiz banger niche.

Sparky has been around for 5 years, going public about 2 years ago. Sparky first turned a tidy profit in 1994, and it appears that it will double that profit in 1995. Sparky is growing by all accounts at 100% per year. It's current trailing P/E is 30, yielding a PEG of .3 (funny how hard these numbers are to work with).

Now the REAL question is, will Sparky continue to grow at the expected rate? It has become clear that certain segments of the Semiconductor industry are seeing price erosion, and a general equilibrium of supply and demand seems to have been achieved?

Let's look at a few possible indicators. . . .

Sales growth? Well, it seems clear that the communications infrastructure will continue to blossom in the foreseeable future. Enough said. If you don't believe this, you're not on America Online, you're living in an ice cave on Antarctica.

Market share or, said another way, looming competition? Let's see, all of those DRAM manufacturers need to build something. What if they decide to build whiz bang communication chips? The answer to this question (and I believe that Wall St. does NOT comprehend this) is simply. . . they don't know how, or they wouldn't be building DRAM's in the first place. These are manufacturers. Building DRAM's is like building tires. Every car needs 5 of them (including spare), compared to just one transmission per car. Sparky is a "design technology" company, and a good one. Plenty of semiconductor companies would love to have the intellectual property that Sparky has developed in the development of specialty chips for the inter-networking marketplace. Sparky contracts with others for manufacturing, currently using 3 foundries for supply. Don't expect Sparky to be threatened by one of the DRAM makers in it's marketplace. The design technology barrier is significant.

Product costs, the amount of money it costs to build and sell your product? Well, we already know that those DRAM manufactures really would like to build some whiz bangers. If they can't design a whiz banger, then maybe they'd like to build some for Sparky? With DRAM prices going down, maybe the costs of "foundry," the manufacturing services that are key to Sparky's costs, will return to historical levels. Reduced fixed costs---now that would be a positive development in Sparky's financial outlook.

Product availability, i.e. more sales growth? Sparky has been limited to shipping only about 70% of their orders throughout most of 1995, due to foundry capacity constraints. Foundry capacity loosening should allow Sparky to ship more of their orders.

Even more sales growth? Looks like DRAM's will no longer constrain sales growth of Cisco's systems. Cisco will ship as many systems as they can get whiz bangers for. Oh yeah, by the way, a major cost of Cisco's system, those darned DRAM's are now cheaper, thus allowing Cisco to stimulate further demand by lowering it's end product price.

So there we have it. The DRAM dive of '95 has set Sparky free to reach it's full potential in '96. We see:

- Increased demand from Sparky's customers driven by growing end-markets.
- Memory allocation constraints have been removed for end products.
- Lower memory prices are reducing the costs of end system products.
- Relaxed fab capacity constraints are reducing Sparky's contracted fab costs.
- High entry barriers in Sparky's niche results in stable pricing for their products.
- And we've not suggested that Sparky raise prices. Its relationship as a "proprietary" supplier to Cisco would prevent this. We have pricing stability. Long term relationship. World class customers.

So now it all makes sense. Commodity makers are coming under pressure, so we've decided to bail on our chip holdings entirely. And although this may be the short term mentality of the street, Fools believe that a few good earnings reports may change this mentality. Sparky has dropped in value over the last 60 days by 50%.

Disclaimer: Sparky is a fictional company and is purely a figment of this Fool's imagination. Any similarities to an actual company is purely coincidental. If you find such a coincidence has occurred, please email me at MFCHIPS@AOL. I will take care of this "unfortunate" situation immediately. Hello, broker?

Seriously, I do not suppose that ALL of these scenarios will pan out. I can say that there are certainly Sparkies out there, with great earnings situations, great market niches, and most-excellent futures. And trading at huge discounts to this growth. I believe there is a relationship between Micron and Sparky, however, this relationship seems more like an inverse one.

[Editor's note: this Fribble is taken from the Semiconductor Industry folder in the Fool's new Industry Research section on America Online. Thanks MF Chips!]