RealMoney Silver - AMD's Going Nowhere

AMD's Going Nowhere
By Bob Faulkner
RealMoney Contributor
12/11/2007 4:03 PM EST

On Thursday, Dec. 13, AMD (AMD) will hold an analyst meeting to review its outlook. Based on sell-side comments, this has to be one of the more underanticipated events of the year. Little if anything is expected. But who knows? Maybe AMD will surprise investors just as it has every 90 days or so for the last three quarters.

None of that success has stopped the weekly rumors that AMD will be taken private by Firm X or Firm Y. Never say never -- but if that were the case, I'm certain that any purchasing firms would not be especially pleased to have to pay up for an extra 49 million shares that were sold to the Mubadala Development Co. just last month.

Do I think AMD is going the way of the buggy whip? No. Do I think it's going to be acquired? No again. Here's why.

The financials are only part of the problem, but let's start there. It has been a tough year for AMD and a full four quarters since it last made an operating profit. What makes this worse is the fact that it's been a pretty good year for overall PC demand.

In addition to having lost $2.2 billion over the last 12 months, the company's debt position has ballooned from $690 million to $5.3 billion. Part of this funded an ill-advised acquisition of graphics chip supplier ATI Technologies and part is for fabrication expansion. What makes this debt situation so difficult for the company is that over the same four quarters it has burned nearly $950 million in cash from operations.

None these circumstances present a particularly attractive foundation for a private equity firm.

But the true believers (in the rumors) make the leap of faith that business will improve, and that's based, in part, on management's expectations. Investors need to ask just how realistic that potential outcome is.

Most of the new microprocessors from AMD that I have seen reviewed on various technology web-sites have received generally favorably comments. However, that alone will not overcome two hurdles for the company.

1. Market share. From late 2005 through the end of 2006, AMD experienced its greatest market share gain. There are several market research firms that track this data and all have slightly different metrics, but for the sake of argument let's say AMD went from the mid-teens to the mid-20% range. It was during this time that AMD saw its gross margins explode into the high 50% range as well as double-digit operating margins.

Much of AMD's market share gains came from its dual-core Opteron processors, which were, at the time of their introduction, in a class by themselves. And that's the point. Intel (INTC) was completely asleep at the wheel and was hit between the eyes by a 2x4! It had no competitive alternative so it was easy to gain market share.

Now the competitive landscape is completely different. Intel's products are just as good and in some cases better than its competitors'. The market share data for 2007 show limited changes and where they do, it's in basis points.

2. Lateness. Not only did Intel bounce back with new designs, it leveraged one of its most powerful weapons against AMD -- its manufacturing prowess. As I have noted on this site and in The Telecom Connection, silicon area is the fundamental basis for the cost of a die. By pushing its process technology hard and fast to 65nm and now to 45nm, Intel was able to boost the performance of its parts and reduce their cost far faster than AMD.

In the fourth quarter of 2006, Intel had essentially 50% of its microprocessor production on 65nm before AMD announced its first shipments at that node. In early November, Intel commenced shipping the first 45nm parts, while AMD is expecting to deliver at that node in the first half of 2008 via one of its foundries.

This lag creates two problems for AMD and dreams of a buyout. First and foremost, it gives Intel pricing leverage that won't go away, which will also make it far more difficult to regain any market share. Furthermore, you can forget any fantasies of gross margins retiring to the 50% range and that, in turn, will continue to pressure AMD's cash generation ability.

Second, having units available at a new process node is one thing; having them in volume is something else. At AMD's foundries, they are going to share that "bleeding edge" process with a number of other fabless semiconductor companies. The question becomes, will they be able to generate sufficient volumes to satisfy the Dells (DELL) and Hewlett-Packards (HPQ) of the world.

The financial, competitive and technology issues do not seem to make for an attractive circumstance for a private equity investor. But you never know.

Despite my negative outlook, I really don't see AMD disappearing from the landscape. Quite the contrary. AMD's customers will keep it around in the same way they keep disk-drive companies in business. Business will continue to go to AMD as long as it produces viable products, and AMD helps the OEMs of the world (such as Dell and H-P) keep Intel on its toes.