CNET también está disponible en español.

Ir a español

Don't show this again

Christmas Gift Guide
Tech Industry

2HRS2GO: IBM does techs a favor

Technology investors owe IBM Corp (NYSE: IBM) a big Thank You.

Big Blue in the first hour of today's trading dropped as much as 16.8 percent, or in line with the company's expectation of a 12 to 16 percent year-over-year drop in fourth quarter earnings. But concurrent declines in the prices of other mainframe and server-related companies -- Sun Microsystems (Nasdaq: SUNW), Oracle Corp. (Nasdaq: ORCL), Compuware (Nasdaq: CPWR), BMC Software (Nasdaq: BMCS) and Computer Associates (NYSE: CA), to name a few -- makes no sense once you look beneath the surface.



Have an opinion on this?



IBM's depressing outlook stems from a slowdown for hardware such as high-end servers and mainframes. The company blames Y2K. Maybe that's true, maybe it isn't; some analysts, such as Morgan Stanley Dean Witter's Charles Phillips, believe IBM's real problem stems from the fact that the G5 mainframe has been out for awhile. Y2K or no Y2K, G5 demand was bound to drop off as the product cycle matures, Phillips says.

Whether Y2K deserves blame or not, one thing stands out: IBM's problems are largely self-contained. Even IBM says so: "The slowdown in high servers, which is unique to IBM, remains an issue," CFO Douglas Maine said during yesterday's conference call.

No major U.S. firm makes anything comparable to a System 390 or the top end of RS/6000 systems. Not Sun, not Hewlett-Packard (NYSE: HWP), not Compaq (NYSE: CPQ), nobody. Those companies churn out hardware comparable to Big Blue's AS/400 and Netfinity lines, the latter of which wasn't affected by Y2K, IBM says. In fact, Netfinity did very well in the third quarter, executives said.

As for the software side occupied by CA, BMC and Compuware, sales of mainframe software have no correlation to hardware, as Credit Suisse First Boston analyst Wendell Laidley points out. While IBM, Hitachi and Amdahl try to shove processing power (measured in Million Instructions Per Second, or MIPs) down their clients' throats regardless of actual need, software vendors take a more practical view.

"Customers tend buy Mainframe software based on actual usage and therefore tend not to 'binge,' " Laidley points out. "Mainframe software's 'pay as you play' model encourages more consistent procurement patterns and is therefore less exposed to quarter-to-quarter spikes and dips in MIPs shipments."

Corporations can live with their current Big Iron set-ups for the next few months. That doesn't mean they won't keep up to date with software, which usually costs much less money than a new mainframe, so there's less risk involved. It's not much different from the average person's buying pattern: most of you reading this won't purchase a new PC in the next few months, but the majority of you will at least try out new software, if not actually buy some.

Morgan Stanley's Phillips found that BMC's business, which relies heavily on financial companies, had the strongest correlation to MIPs shipments. But even BMC is largely insulated by its long-term contracts, Phillips says.

"The duration on these contracts is roughly 2.5 years, so a one or two quarter blip wouldn't throw off a customer's planned consumption of MIPs over that time period," Phillips writes in a research note released today.

Big Blue may be the largest technology company in the world, yet its business model bears little relation to other tech companies. IBM plays in many segments, but it's only dominant in two: mainframes and services. You can't consider IBM a PC powerhouse anymore, just one of several major players. Its low- and mid-range server businessess continue to grow at slower rates than competitors such as Sun. IBM software units are powerful in databases, group suites and mainframe systems, but they're not the undisputed leader in any of them. Hard disk drives have done well, but that industry still offers fierce rivals.

No other company is comparable to IBM's melange of businesses. Conversely, that means IBM, even though it is big, actually isn't a very good bellwether. Except for mainframe hardware, it doesn't reflect any technology market particularly well, maybe because it is in so many of them -- there's something to be said for focus.

So the market shouldn't automatically assume technology industry weakness, just because IBM feels it. Today's tremors shouldn't be taken as anything serious. Instead, they're just a chance to buy some very good companies at lower prices. You can thank Big Blue for that.

Other issues:

  • Aether Systems
  • (Nasdaq: AETH) Maybe Wall Street has discovered a new sense of Nordic interest, given the Viking-looking name for this provider of wireless data systems. Today's surge certainly isn't powered by a great financial performance; Aether lost of $4.3 million on revenue of $787,000 for the first six months of this year.

  • Wall Data
  • (Nasdaq: WALL) It took the company less than two months to change its mind about a sale. In August, the provider of software that lets PCs and workstations connect to servers and mainframes decided not to seek a buyer for its businesses; today the company announced a deal to be acquired by rival NetManage for $94 million. A tender offer of $9 a share may not sound like much, but it's the best Wall shareholders can expect after the company's disappointments for the last few quarters. 22GO>