Print-friendly version
Send this to a friend
Posted 10/18/2004
Contrarian Chronicles
About Contrarian Chronicles
Learn the Contrarian Chronicles lingo
Subscribe to Market Rap on Fleckenstein Capital
Contrarian Chronicles
Recent articles:
• The frantic search for a silver lining, 10/11/2004
• The government's inflation con job, 10/4/2004
• Watch out for the hurricane named Fannie, 9/27/2004
More...
| | Contrarian Chronicles
Intel: All risk, no reward
If you look closely at Intels earnings release, you have to conclude the chip giant is really stuck. It has too much capacity. All cutting production or prices will do is wipe out profits.
By Bill Fleckenstein
A triumph of obfuscation over fact. That's my description of Intel's third-quarter earnings release from last Tuesday. Here's a fact: Intel (INTC, news, msgs) now battles a three-headed monster -- overcapacity, stagnant demand and a better 64-bit "mousetrap" made by Advanced Micro Devices (AMD, news, msgs).
Intel 101
It's worth spending a lot of time on Intel, as I think there are many lessons to glean from it. (For review, see the July 19 Contrarian: "Intel's twin woes: excess capacity, slack demand.") I've long maintained that Intel and IBM (IBM, news, msgs) are classic examples of mismanaging your business in the long run to maximize your stock options in the short run. (And there are many others.) In my opinion, many tech companies are not in whatever business they claim to be in. They're really in the business of producing a rising stock price -- with the underlying business merely a sidelight.
Stock options: Keep back 200 feet
As circumstantial evidence to try to prove my case that it's all about stock options, please note that Intel crowed that it used $2.5 billion to buy another 106 million shares. And of course, it has led the disingenuous fight about the impossibility of calculating what options are really worth. I find it more than slightly ironic that this maker of the engine of the machine on which you are reading this article claims to be unable to value stock options. And, as you will see in a minute, Intel claims to be unable to add up a handful of different items included in its inventory charges.
Intel's fundamental problem is that it has tried to spend its way through a downturn that's really a saturation problem. Folks don't need more PCs, and they don't need faster processors. Yet, Intel needs, above all else, to charge high prices for its products. That, of course, is not exactly a good way to stimulate demand in a market that's saturated. Meanwhile, as Intel has tried to spend its way out of capacity -- and spent gobs of time cheerleading/defending its stock price -- AMD has produced a superior mousetrap and is now busily taking market share.
To summarize: Intel's fundamental problem is that it has put up too much capacity and is creating more product every day than it can sell. In fact, you will see that if Intel were to bring capacity in line with demand, it would likely make no money.
What's in the write-down?
Now let's turn to what happened on Intel's call last Tuesday night. Most people expected Intel's inventories to go up, as did I. There was a cry of relief amongst the bullish contingent that inventories actually dropped $43 million on a base of $3.2 billion. However, Intel took an inventory write-down and a reserve for inventory. Inexplicably, the company would not divulge the total amount of those adjustments. Though pestered about this a couple times on the call, they absolutely refused to give a number and used a whole bunch of weasel words about why:Question: "Could you please give some quantification of what the reserve was in dollars in Q3?"
Answer: "I (Chief Financial Officer Andy Bryant) won't answer your question specifically. I'll try to give you a little better feel." Then Bryant goes on to spew a whole bunch of nonsense, such that the questioner comes back with: "You lost me with the reserve."
Answer: "There are three buckets of reserves. Some are OK reserves, some are bad reserves." (He never mentions the third reserve.)
Question: "So what was the magnitude of the reserve?"
Answer: "I didn't give you the specific." Then, near the end of the conference call, the question came up again: Question: "Can't we just get an overall inventory-hit number out of you guys?"
Answer: "Not this time. If it were one big item, I would. But . . . I've got a series of five to seven, eight to 22 million-dollar things. So it's very difficult to cull them out, because there's not a single thing that's important enough to be culled out." Of course, what the questioner wanted to know was the total magnitude of the inventory adjustments, but Andy Bryant went into his comments about the specifics so as not to answer the question. Bottom line: He refused to tell everyone how big of an adjustment Intel took this quarter.
(Editors note: Listen to Intels conference call.)
Inventory . . . the envelope, please
However, it is possible to deduce approximately what the size of that charge was, and here, I am indebted to my friend Fred Hickey, editor of the High-Tech Strategist newsletter. He pointed out that, for at least the last four quarters, Intel's cost-of-goods-sold has ranged from about $3.185 billion to $3.275 billion. In other words, $3.2 billion has been a good running average for their cost-of-goods-sold.
Related Intel news and commentary
I need to make an oversimplified point here: Intel has a high-fixed-cost business. It put sand in the front end and out came high-margin chips on the back end. The cost of doing business stays pretty much the same, pretty much no matter what its volumes are. Its margins move up and down with its revenues. It's an important point to understand.
In any case, last Tuesday night, we saw that its cost-of-goods-sold jumped to $3.752 billion, which was up $483 million from the prior quarter and $477 million from a year ago. Some of that increase may be a legitimate uptick in cost-of-goods-sold. However, the majority of that number is likely to be the approximate size of the inventory adjustment, both write-off and reserve, that Intel took. This is roughly greater than 10 times the size of the decline in inventories. Even though without a breakdown between the size of the write-off and the reserve, we can't know for certain what happened, it would be almost impossible for the units in inventory not to have risen.
And, if you go back and look at the last time Intel had to finally start biting the bullet on an inventory problem (in the third quarter of 2001), you will see that its cost-of-goods-sold popped up to about $3.6 billion. But back then, its inventory declined from $2.8 billion to $2.35 billion, i.e., a drop of about $450 million, as that inventory got flushed through the cost-of-goods-sold. Currently, inventory is still $3.2 billion.
Cat got a chip's tongue
Though CFO Andy Bryant was totally evasive about the size of this write-off, he did crow that the company had whittled inventory down by $43 million. At that rate, it would take about 80 quarters or 20 years to get rid of it all. They obviously can't wait that long, but the point survives that the "progress" Intel made is pathetic, especially given the size of the adjustments it didn't want to disclose.
So, where does that leave Intel, and why does all this matter so much? As I started out saying, its fundamental problem is that it has too much capacity. How much, we can't say for sure, but we can make a stab at it. In the last four quarters, it has built up nearly one-quarter's worth of inventory. Again, even though the inventory on the books is only $3.2 billion, since its gross margins are 50%-plus, it works out to be about $7 billion worth of inventory, or roughly a quarter's worth of revenues. (Once again, I am oversimplifying this, because not all of Intel's inventory is processors. But for the purpose of illuminating the example so that everyone can follow it, I have to make some broad assumptions.)
No pent-up demand for Pentium 4
Intel now has a quarter's worth of revenues in inventory, up $1 billion, or 50%, over the last five quarters. Well, what would its profits look like if it cut its output by, say, 25% to bring supply and demand in line? Virtually nothing would happen to expenses. But 25%, or about $2 billion, would be cut from revenues, which would simultaneously shrink profit by $2 billion, as well. Given that Intel's pretax operating profit is only $2.3 billion, you can quickly see how fast its earnings would evaporate. Thats why it is so reluctant to cut capacity.
High-fixed-cost businesses are wonderful once you can produce above the break-even level, because the revenues from the incremental products all fall to the bottom line. But leverage cuts both ways, and profitability can reverse rather quickly.
Intel would face a similar problem if the company decided to try to cut price rather than capacity. Obviously, its cost-of-goods-sold wouldn't change much, but its revenues would shrink, and it would have the same problem I just outlined. For Intel to move inventory, it is going to have the problem of trampling prices in the market. If it was forced to try and sell all of the extra quarter's worth of inventory on the books in one quarter (which it wouldn't), your guess is as good as mine as to what price it might receive in an attempt to sell everything.
If Intel was able to sell it at, say, 60 cents on the dollar, then its income statement would look something like this: $3.2 billion for the cost-of-goods-sold for what was made this quarter and $3.2 billion for the cost-of-goods-sold for what was in inventory. That's about $6.4 billion for the cost-of-goods-sold. The $7-$8 billion of revenues that the inventory grossed up to, plus a similar amount that the quarter it was in grossed up to, equals $14-$15 billion. The latter multiplied times 0.6 is $8.4-$9 billion. Subtracting the combined cost-of-goods-sold of $6.4 billion, plus their operating expenses of $2.3 billion, once again leads to almost no profit margin. (Again, I am being overly simplistic to illustrate the point.)
Hiding in plain sight
What I find staggering is that Intel thinks it can get away with this stonewalling, even though the problem is hiding in plain sight. Intel has engineered itself into a giant predicament, but Intel is not alone in this. Many companies in chip land (and their predicament obviously affects the chip-equipment companies), as well as other industries, have built up too much capacity. However, due to their reluctance to admit to that, they resort to all kinds of games. In the long term, this inability to face up to their mistakes will only put them into a worse position.
Hopefully, this discussion makes it clear why you would not want to pay 4.2 times sales for Intel now that the companys revenues have been stagnant for more than four years, compared to the approximately 2 times sales it traded for from 1988 to 1994. That was when Intel was still growing rapidly and had the bulk of the PC boom still in front of it. Intel today is a textbook example of all risk and no reward.
Bill Fleckenstein is president of Fleckenstein Capital, which manages a hedge fund based in Seattle. He also writes a daily Market Rap column on his Fleckensteincapital.com site. His investment positions can change at any time. At the time of publication, Bill Fleckenstein was long IBM puts, short Intel and long Intel puts. Under no circumstances does the information in this column represent a recommendation to buy, sell or hold any security. The views and opinions expressed in Bill Fleckenstein's columns are his own and not necessarily those of MSN Money.
|
