It is no small measure of the goldilocks nature of the technology market that Intel
Intel reported earnings last night of $1.9 billion, or $0.30 per share, which is 5% higher than a year ago. Back out the fact that Intel's effective tax rate for the quarter only hit 21%, substantially lower than its expected 31% and well below last year's level, and earnings are essentially flat from last year's result. The company's revenue came in at $8.47 billion for the third quarter, a slight gain of 5% over the previous quarter. This is substantially lower than Intel's normal third-quarter sequential gains, even over the last tepid few years, of 7% plus. Most importantly, it seems, is that Intel's inventory levels declined over the quarter, from $3.2 billion to $3.18 billion, a reversal of a recent trend that some Intel watchers have labeled "surprising."
I'm shocked, shocked!!
Surprising?!? Only perhaps because the company had guided that the levels would be $100 million or so higher. But as we'll see in a minute, even this drop isn't what it appears to be. Intel's inventory has been at record levels for the better part of the year, and we've been watching its customers' inventory levels rise and rise and rise. Inventory rising is not in and of itself a bad thing: Companies build up inventory when they believe that the demand for their products is going to increase. What has happened over the last nine months at Intel has been anything but. Five percent sequential revenue gains over what traditionally is the weakest quarter of the year are abysmal. They built it; not enough came.
So Intel decided to bite the bullet and bleed off inventories. Intel's gross margins tend to sit near 59%, and Intel watchers expected them (due to the company's commitment to burn off inventories) to drop to 58%. Instead, gross margins sank to 55.7%, and the company announced that it anticipated that they would remain at about that level for the next quarter. Apparently, this wasn't enough, so the company wrote down even more inventory value.
It must be said that the format and information that Intel provides in its quarterly press releases is wonderful. I'm struck by the fact that, for the first time, Intel's sales in both Europe and Asia (excluding Japan) exceeded that of its sales in the Americas. Its Asian sales are now, at 48%, approaching half of Intel's total revenues, while its sales in the Western Hemisphere have sunk, not only as a percentage of the whole, but in absolute terms, from $2.17 billion a year ago to $1.8 billion this quarter, a drop of 17%.
In this past quarter, Intel repurchased more than $2.5 billion of its shares on the open market. One would hope that this would result in existing shareholders having a larger claim on earnings, but this is not the case: Intel's diluted share count of 6.54 billion today is higher than its basic share count of 6.53 billion a year ago, this in a time when the company's stock has sunk by more than 22%, so the diluted share count wouldn't simply inflate as a result of more options being in the money. Come on. That use of shareholder capital to offset option dilution is just sort of par for the course -- to me it means that the earnings and cash flows Intel does generate aren't nearly worth as much as the market grants them, and that's too bad. Owner earnings aren't worth jack if they aren't deployed to the benefit of the owners.
Where's my pencil?
But here's what's really frightening to me, why I think Intel pushed as hard as it could and still couldn't beat its earnings: that old inventory bugaboo. Think about it -- yes, inventories are down. But the company took a higher reserve against them (automatically devaluing a segment of its inventory), it accepted lower gross margins than it usually does, and it still only managed to decrease its total inventory by $43 million, or 1%. Want to know how big the writedown of inventory was? Tough luck -- it wasn't disclosed, and in the company's conference call, Chief Financial Officer Andy Bryant noted he couldn't give an exact number.
Well, I have a guess, but it is necessarily following a little bit of arm waving. If the company had revenues last quarter of $8.1 billion and cost of goods sold of about $3.2 billion, and as I mentioned, its historical cost of goods sold is generally remarkably flat, one could infer that its cost of goods sold should be, for $8.47 billion in revenues, in the range of $3.32 billion. Instead it was $3.75 billion. Take the delta, or $427 million, and that is going to be in the neighborhood -- the very general neighborhood -- of where the combination of inventory writedown and product discounts was for the quarter. Where'd that $100 million projected gain in inventory go? It's just a green eye shade's memory.
So Intel is sitting on record high inventories -- more than 78 days' worth of sales (obviously, not all of Intel's sales are microprocessors, so this is more or less the actual condition), and its big customers and competitors are bloated with their own inventory indigestion. So, how on earth is Intel going to keep from writing off a ton of inventory in a fairly pouty sales environment, where companies like Celestica
You guessed it -- Intel's going to have to cut its prices and/or cut production. Most likely both. These are not good things. The big contortion act is coming to what may be a shocking conclusion. Remember when tech companies were all seeing dramatic improvement for the second half of 2004? It's not turning out that way.
See also:
Bill Mann's The Tech Inventory Hair Ball
Tim Beyers' Intel Swings, Misses
Tim Beyers' What's Wrong Inside Intel?
Philip Durell each month digs deep and finds some of the most intriguing value opportunities around. What does he have up his sleeve this month, you know, besides, ideally, an arm? Take a free trial to Motley Fool's Inside Value newsletter and check it out!
Bill Mann owns none of the companies mentioned in this article. Please view his profile for a complete list of holdings. The Motley Fool is investors writing for investors.
