To Intel Analysts: Margins Aren’t Everything

Today’s Nasdaq sell-off was relatively steep (although you have to remember that with the Nasdaq near 3,000, 71 points is just not panic-inducing). And while picking proximate causes for market movements is a fool’s game, I’ll be a fool and point to two important reasons: bond prices and Intel.

With the yield on the 30-year bond now at a two-year high, at 6.3 percent, bonds are looking like a tastier investment, and there’s still a lot of concern that the Federal Reserve will raise interest rates at its Nov. 16 meeting. Neither of those things are good for stocks. As for Intel, it came in with quarterly numbers that were a couple of cents shy of analysts’ expectations and got crushed as a result, with much of the rest of the tech sector falling in sympathy (or empathy, or whatever you want to call it).

Intel is a true tech bellwether, since its chips go into most of the PCs in the world, meaning that if it’s having trouble, a lot of other companies probably are, too. And given that in the past few months expectations for the semiconductor industry have risen sharply, Intel’s failure to blow away estimates–something it hasn’t done, by the way, for a couple of years–came as a shock. But there are good reasons to think that today’s sell-off was overdone and that whatever trouble Intel is having has less to do with the “tech sector” (whatever that mythical beast is) in general and more to do with Intel in particular.

The company, for instance, took great pains to point out that demand for its chips was not the problem. In fact, in a curiously un-Intel-like performance, it was manufacturing the chips, especially its newest chips, that tripped the company up. Intel delayed the release of its highest-powered Pentium processor, Coppermine, and a new chipset for memory chips because it couldn’t hit the production schedules its customers needed. Had they shipped as planned, Intel would have had no trouble meeting estimates. Traditionally, it’s Intel’s competition, most notably Advanced Micro Devices, that finds deadlines bothersome. But apparently even masters of manufacturing can trip up.

There’s a bigger point than just manufacturing mishaps at stake in the Intel sell-off, though, and that’s what you might call the “fetishization of margins” problem, which I touched on in last week’s discussion of’s zShops initiative. Nearly all the analysts who marked down Intel today emphasized that the company’s gross margins–which equals revenue from products minus the cost of sales for those products–fell in the most recent quarter. Now, they fell 0.2 percent, from 58.9 percent to 58.7 percent, which is statistically insignificant. But the important point, for the analysts, was that they didn’t rise.

What this demonstrates, supposedly, is the impact on Intel of having to sell so many more of its cheap chip–the Celeron–rather than its higher-priced Pentiums. As consumers have got used to the sub-$1,000 PC, Intel has had to make chips that cater to that market, and naturally they earn less profit on the $80 Celeron than on the $250 Pentium III. Analysts see this as a slippery slope, so that pretty soon Intel will watch its margins driven down to a sliver.

Well, maybe. But somehow I doubt it. In the first place, Intel’s margins are still remarkably high for a manufacturing company (and would have been much higher had they shipped Coppermine). More to the point, falling margins are not, in and of themselves, evidence that things are falling apart. If you do things right, turning inventory more rapidly, you can make it up on volume. The value of a business depends not on how hefty its profit margins are but on how efficiently it uses capital, and you can use capital efficiently by setting up a revolving door of low-margin products (Wal-Mart), a steady stream of high-margin products (most banks), or a mix of the two (which is where Intel is going). So let’s wait a little while before deciding that Intel is really on the road to destruction.