Milo, an interesting take on Intel's current situation.
ted
____________________________________________________________
Intel's New Pricing Scheme Will Chip Away at the Bottom Line
By Jim Seymour
Special to TheStreet.com
7/30/01 2:42 PM ET
Everyone's struggling with analogies for the semiconductor industry these days. Comparisons to optical are commonplace. I want to suggest an analogy no one seems (yet) to have thought of. It should strike some unease -- if not fear -- into the hearts of investors in semiconductor manufacturers -- and by extension, investors in semiconductor-equipment companies.
The analogy to dot-coms doesn't work across the board, and I'm not suggesting that semis are headed for the junkyard, complete with R.I.P tombstones next to those of so many dot-comas. But in at least one area, this analogy is very troubling.
Five key structural problems in dot-com land, circa 1999, led inevitably to the collapse of so many apparently promising Web businesses. My list:
Too much capital, available too cheaply;
Arrogance about the supposed replacement of economic reality by a New Economy;
Arrogance about the dot-coms' ability to create demand from thin air ... or at least to dramatically grow existing demand;
Inept and inexperienced management;
And finally, the real killer: A focus on building market share, not making money.
It's that fifth item that worries me so much in semiconductors today: Across the board, semis are cutting prices like crazy, clawing and scrabbling for market share -- almost always in ways contrary to what the history of the semiconductor industry has taught us about how to rake in the chips in ... chips.
Intel (INTC:Nasdaq - news - commentary) is the poster child here, though Advanced Micro Devices (AMD:NYSE - news - commentary) ranks high. (On the other hand, AMD's been playing this game for a long, long time.) And even the specialty chipmakers are succumbing to a too-short-term perspective.
Ironically, so many good things seem to be happening for Intel these days. CPU speeds are pushing up to and beyond 2 gigahertz, at least presumptively creating demand to replace older, slower machines. Intel's new Itanium 64-bit chip is close to making what I think will ultimately be a big market impact. Intel has co-opted or simply intimidated important competitors (including Hewlett-Packard (HWP:NYSE - news - commentary), whose PA-series CPUs would have been a serious threat to Intel's 64-bit efforts, and Compaq (CPQ:NYSE - news - commentary), which in a huge turnaround two weeks ago dumped further development on its high-end Alpha chip and pledged its troth to the Itanium). And Intel has launched what ought to be an important second front, in communications chips.
And yet at the same time, market softness -- what a gentle term that is in semis! -- and resulting competitive forces have forced Intel to simply abandon the pricing model that made the company such a powerhouse.
As it began establishing its "x86" series chips, with the 8086 and 8088 in 1980, Intel established itself with a business model based on rolling out new, superior CPUs at very high prices, intended to grab early profits and repay the cost of development -- then six to 10 months later, cutting those prices. Those cuts opened room at the top of the line for successor chips, and also kept competitors, such as AMD and National Semiconductor (NSM:NYSE - news - commentary), at bay.
The model worked superbly: Intel's huge return on investment at those high, early prices helped fund development work on future chips -- and dropped fat profits to the bottom line.
Intel also measured carefully the pace of its improvements: Yes, it subscribed to the idea of "eating your own children" -- introducing chips that outperform your current top-of-the-line parts, before someone else does -- but since Intel was usually well ahead of its competitors, it could also afford to lag the pace of innovation and still stay on top.
That also contributed to extending the market life of specific chips -- and thus to corporate profitability.
Intel lives in a different world now. AMD is constantly nipping at its heels -- and often wins apples-to-apples performance comparisons. So Intel has given up on the "high early prices, then cut prices only as needed to find off the competition" model that made its executives and shareholders wealthy.
Now Intel introduces new, top-of-the-line, high-performance chips at prices, and with margins, that are perhaps half of what it would have set a year or two ago. Intel may have no choice in this, given AMD's push and lagging PC sales ... but by aggressively pursuing market share at almost any cost, it dumped the model that produced such stellar returns.
Anyone who watched dot-coms -- specific ones, or the whole pack -- claim quarter after quarter that the name of the game was gaining market share and that profitability would have to wait now shivers when a huge, expensive design-and-manufacturing machine such as Intel moves in that direction.
The change means that Intel has to be revalued, away from historical models, in terms of likely future earnings and a "right" price-to-earnings ratio. I think Intel, around $30 at midday Monday, is still richly priced, with a P/E in the high 30s.
One more point on this incomplete analogy. There is more than a whiff here of my third point about why dot-coms failed. They not only valued market share over earnings; they also believed they could create massive demand, or huge shifts in demand, by their actions.
Intel claims that its new pricing model will kickstart PC sales by cutting typical street prices. Yes, the CPU is by far the most expensive component in most PCs ... but will a change of $100, $200 or even $300 in selling prices -- the range likely to result from Intel's price cuts -- really send demand soaring?
I doubt it. Both consumers and corporate IT buyers these days understand that we're past the era of "speeds 'n feeds," that a faster CPU doesn't necessarily mean they'll enjoy using their PCs more.
Moreover, while both those classes of buyers display some price sensitivity, in this current market -- almost entirely now a replacement market (and the more so for the higher-end PCs Intel is trying to get moving) -- $200-$300 difference in the price tag is not going to move most buyers to make the purchase.
Again: I don't mean to push this incomplete and inexact analogy too far. It fails in many ways. But consider carefully if you're thinking about jumping back into semi stocks what these new pricing dynamics mean to chipmakers' bottom lines -- and to their share prices.
The glory days are past. The gory days are here.
--------------------------------------------------------------------------------
Jim Seymour is president of Seymour Group, an information-strategies consulting firm working with corporate clients in the U.S., Europe and Asia, and a longtime columnist for PC Magazine. |
