Despite the sensationalism surrounding Martha Stewart’s sentencing on Friday, the disappointing earnings report out of Intel late Tuesday was the true big market event of the week. Shares of the chip maker lost more than 10% in Wednesday trading as investors questioned the legs of the technology recovery.
While I agree that the reaction to the quarterly report doesn’t bode well for Intel, especially considering that it met the profit forecast, the momentum of the tech sector as a whole hasn’t been compromised. As I see it, Intel’s problems are Intel’s problems.
The big reason for the sell-off, and the one that has gotten all the coverage in the media, is Intel’s margins slipped a bit during the quarter. Margins are important because Intel invests vast amounts to develop new technologies to build state-of-the-art chips, and manufactures them in huge volumes. High and stable margins assure investors that Intel is in control of its high-stakes game. Indeed, Intel offered as one explanation for the drop in margins expenses associated with a little boo-boo in the factory that cost it $38 million (in the delicate parlance of the industry, it called the snafu a “manufacturing excursion”).
As a result, Wall Street analysts have constructed elaborate and arcane models for understanding and predicting Intel’s margins. So when there’s a margin surprise, Wall Street has to take a hard look at everything it thought it knew about Intel. And when there’s a negative margin surprise, it’s not a happy day.
Wall Street’s confusion about Intel’s margins got pretty intense in the company’s earnings conference call after the close on Tuesday. Several analysts threw questions at Andy Bryant, the chief financial officer, that were so complex and so theoretical that Bryant had to say more than once that he didn’t understand what he was being asked. Sure, it’s embarrassing for the company not to be able to handle tough questions. But at the same time, investors should take a couple of steps back for a big-picture look at what Intel’s margins really mean.
When you strip away all the models and all the quarter-to-quarter noise, the plain fact is that over the long term, Intel’s margins probably have nowhere to go but down. Why? Because high margins arise when a producer has the power to keep prices high in a product that is growing rapidly. Intel found itself in just that situation when it and Microsoft won the personal-computer lottery in the early 1980s. They found themselves with virtual monopolies on the two must-have components of one of the fastest-growing technologies in history — Microsoft had the operating-system software, and Intel had the microprocessor.
But once you’ve won the lottery once, it’s not likely you’ll win it a second time. With the hypergrowth of the PC market now in the past, both companies have had to diversify into more competitive and less profitable lines of business. Intel, which two decades ago abandoned the low-margin, commodity computer-memory business in order to focus on high-margin microprocessors, now finds itself going back into that business by fighting for market share in cheap “flash memory” chips. Intel has been doing very well in flash memory. That’s good news for revenues, but bad news for margins. The numbers tell the story. In 2000, Intel posted record revenues of $33.7 billion, with earnings per share of $1.57. This year the company will probably set a new revenue record. But the consensus earnings forecast is only $1.22 a share — a 22% drop from 2000’s high-water mark.
In the conference call, one analyst threw a very straightforward question at CFO Bryant: Why are you eroding your margins by going into these low-margin lines of business like flash? And Bryant gave a straightforward answer: “We’re in this business to increase the overall profit dollars, not the percent.”
True enough. But that kind of straightforwardness doesn’t get you anywhere on Wall Street. Talking about “overall profit dollars” is what stodgy, mature go-nowhere companies do. Growth companies talk about soaring revenues and high margins. So what kind of company do you now think Intel is? And what kind of valuation does Intel deserve? You get only one chance to answer correctly.
OK, I’ll tell you: Intel is a stodgy, mature, go-nowhere company. But at the same time, Intel faces technology challenges that make it totally unlike the kinds of companies that you normally think of as stodgy and mature. And that brings me to the other, less publicized problem that came out in Tuesday’s earnings report. Intel reported that inventory levels grew by $427 million during the quarter. That means that, of the $8.05 billion in revenues that the company reported, 5% was what amounted to selling stuff to itself.
Some of that is to be expected. Every manufacturing company has to produce goods and store them for a while before customers show up and buy them. And as CFO Bryant said several times in the call, for a chip maker it’s a fatal mistake not to have inventory available when customers demand it. In the technology business, customers are always in a hurry.
Bryant also said that some of the inventory was actually the result of a positive surprise — some of Intel’s manufacturing processes had been more efficient than expected, so more chips had been manufactured than the company actually intended. It’s the dirty little secret of the chip industry that manufacturing semiconductors is actually a lot like farming: You never really know how big the crop is going to be until you harvest it. This quarter Intel got a bumper crop.
Good news in some sense, but don’t tell that to a farmer who has barns full of soybeans that no one wants to buy. In the conference call Tuesday Bryant assured investors that there are plenty of people who want to buy Intel’s soybeans. The company sees global demand continuing strong. But it remains a reality that in the current quarter, Intel is going to have to burn through that inventory before they can start doing business that will make it to the bottom line.
And speaking of bottom lines, here’s mine. Intel’s falling margins are Intel’s problem. There’s no reason to think that this indicates a problem for the whole sector. As an investor, face the fact that Intel is a mature company like Coca-Cola or Pfizer, and if you want something growthier you’ll just have to look elsewhere.
And Intel’s inventory problems don’t mean that technology buyers are on strike. It means that Intel got lucky and harvested too many chips. Again: It’s not a problem for the whole sector.
I continue to think we’re going to have a difficult summer, with all stocks (including technology stocks) trading back down toward the low end of this year’s trading range. There’s just too much uncertainty about the presidential elections, as I wrote about here last week. But while we wait for that to play out, corporate earnings continue to grow. And stocks are starting to get interestingly cheap again.
The above is an “Ahead of the Curve” column published July 16, 2004 on SmartMoney.com, where Luskin is a Contributing Editor.