If you've decided that companies like

Nortel Networks

(NT)

,

Cisco Systems

(CSCO) - Get Report

,

Network Appliance

(NTAP) - Get Report

and

Broadcom

(BRCM)

and their stocks are never coming back, the technology sector is easy to figure out these days. Just stay away. Wipe the stocks off your radar screen. Forget they even exist.

For the rest of us, however, the sector remains a difficult puzzle. Which stocks will come back strong and which are down for the count? Which stocks will rebound quickly and which will only slowly stagger off the floor? Which stocks are really cheap and which stocks are merely cheaper than they were?

I've tried to answer those questions more than once in columns this year, giving you my take on individual stocks and on how different types of technology stocks would respond to the

Federal Reserve's interest rate cuts. I'm sure you've read or listened to lots of other opinions, too. In fact, by this point in our journey toward the bottom of the bear market that has ruled the technology sector since last March, you've probably heard more opinions than you need -- many contradicting each other. But how do you put them all together? How can you put all the stocks you've heard mentioned on a level playing field and compare them with each other to decide which is best for you, right now?

I don't claim to have the definitive system for doing that, nor the most sophisticated. But in my own efforts to sort through this sector, I've come up with a rough-and-ready method for comparing technology stocks. This has helped me understand the differences between the many potentially attractive stocks in this sector and to rule out some equities entirely.

My system uses just three yardsticks for measuring stocks -- fundamentals, timeliness and valuation -- but it has helped me divide the technology sector into the good, the neutral and the ugly. The idea is to give each stock and company you study a positive, negative or neutral score in each of these three areas and then combine the scores into one ranking. No stock is likely to come out with a perfect average. But you're not looking for perfection, just for the best in the sector. (And to stay away from the worst.)

Fundamentally Speaking

My fundamental yardsticks here are market share, product mix, new products and cash.

Good times tend to cover up problems. Lose an order or two because products don't arrive on time? With customers pounding at the door, it doesn't seem to matter much. A competitor's built a new, more efficient factory? With revenue growing almost faster than a company can handle, why worry about a little slippage in operating margins?

Tough stock markets and tough economies test companies and their stocks. Suddenly, those inefficiencies matter. Minor holes in a company's product line turn into chasms that swallow all revenue growth. A slip in market share turns into a landslide as once less-demanding customers desert to suppliers they preferred all along and who can now meet their orders.

The stocks you want to buy when the economy turns around are those of companies that did well, or at least held their own, when the economy was lousy. Look for companies that are taking market share from their competitors, such as

Brocade Communications

(BRCD)

,

Ciena

(CIEN) - Get Report

,

Advanced Micro Devices

(AMD) - Get Report

and

Nokia

(NOK) - Get Report

-- and mark down companies that are losing share, such as

QLogic

(QLGC)

,

Lucent Technologies

(LU)

,

Intel

(INTC) - Get Report

and

Motorola

(MOT)

.

Search out companies where hot products -- those that are selling well even in an economic slump -- make up a good part of total sales. Examples would be

Xilinx

(XLNX) - Get Report

,

Applied Micro Circuits

(AMCC)

and

Novellus

(NVLS)

. Give failing grades to companies with a cold product mix such as

Hewlett-Packard

(HWP)

or Lucent.

Look for companies that have introduced or are about to introduce what could become the hot products of tomorrow. In the chip area, some examples are

LSI Logic

(LSI) - Get Report

with its CDMA chip for wireless phones and

Atmel

(ATML)

with its high-end flash memory for Bluetooth applications. And, of course, take points away from companies that seem to be having trouble getting important new products out the door. An example would be the trouble that Cisco Systems is having with its optical switch.

Last but not least, make sure that the company has enough cash, or the ability to raise it at a reasonable cost, to survive until the good times and to execute any critical parts of its business strategy while the slump lasts.

Measuring Timeliness

My timeliness yardstick: early, midrange and late turners.

Take the technology investor's rule of thumb: Buy six months before the expected turn in a company's business. Combine it with what we know about which sectors recover first after the Federal Reserve starts cutting interest rates. That tells me which technology subsectors are likely to bounce back first and which are likely to stay dead money for a while.

Early turners are likely to be concentrated in the consumer technology and traffic-supported technology subsectors.

Dell Computer

(DELL) - Get Report

, Intel and

Nvidia

(NVDA) - Get Report

and all PC stocks are examples of the consumer technology subsector.

eBay

(EBAY) - Get Report

,

Yahoo!

(YHOO)

and

Mercury Interactive

(MERQ)

all belong to the traffic-supported subsector. Revenue at those companies rises as Internet use increases and as ad revenue rises. This should happen when consumer confidence returns as the economy starts to recover.

Technology companies that support infrastructure for consumers or that sell peripherals to corporations should be midrange turners.

Applied Materials

(AMAT) - Get Report

and Novellus are members of the first group -- they sell the equipment to make the chips that go into wireless phones and PCs. Companies like

Extreme Networks

(EXTR) - Get Report

,

EMC Corp

(EMC)

and

Sun Microsystems

(SUNW) - Get Report

sell gear to corporations that extend their infrastructure, but that don't require a corporatewide commitment to re-engineer or replace complete existing systems.

Deep infrastructure companies, on the other hand, dominate the late-turner group. Examples include Cisco Systems, Nortel Networks,

Inktomi

(INKT)

and

Critical Path

(CPTH)

.

Putting a Value on Techs

My valuation yardsticks: historically cheap, historically reasonable but not cheap, cheaper but not cheap.

Despite the punishment inflicted on the Nasdaq during the past 11 months, very little in the technology sector is historically cheap even now. You can get a feel for the bottoms and tops of multiples on any stock on

MSN MoneyCentral

by using the 10-year

summary of key ratios found under the "Financial Results" tab in the "Stocks" section of the site. For example, in 1997, Intel traded at a

price-to-earnings ratio of 21, about where it is today. But historically, the stock tended to trade between a high multiple of 14 and a low of 11, a far cry from where we are now.

Even if we limit ourselves to recent history -- say, a look back to 1997 and the period before the great run-up in technology valuations -- few stocks make the grade.

Oracle

(ORCL) - Get Report

, definitely. The stock has a current price-to-earnings ratio of 21 and showed a multiple of 33 in 1997. Intel, yes. Atmel, pretty close; at a recent P/E of 26, the stock is in the historical ballpark.

A larger group of stocks are historically reasonable but not cheap, which is to say, still above the historical low but well below recent highs. In this group, I'd put stocks such as Applied Materials, LSI Logic, Sun Microsystems and

Vitesse Semiconductor

(VTSS)

.

And then, finally, there are those stocks that despite huge tumbles are still expensive in historical terms: EMC, Cisco Systems,

JDS Uniphase

(JDSU)

and Broadcom.

10 Tech Picks

Put these yardsticks all together and which technology stocks match up best? I'll give you my 10 best:

1. Intel: Improving fundamentals, early turn timeliness, historically cheap.

2. Applied Materials: Improving fundamentals, midrange timeliness, reasonable valuation

3. Novellus: Improving fundamentals, midrange timeliness, reasonable valuation.

4. Oracle: Improving fundamentals, midrange timeliness, historically cheap.

5. Nvidia: Improving fundamentals, early turn timeliness, cheaper but not cheap.

6.

Wind River Systems

(WIND)

: Improving fundamentals, early turn timeliness, cheaper but not cheap.

7. Atmel: Improving fundamentals, early turn timeliness, historically cheap.

8. LSI Logic: Improving fundamentals, early turn timeless, reasonable valuation.

9. eBay: Improving fundamentals, early turn timeliness, cheaper but not cheap

10.

SanDisk

(SNDK)

: Improving fundamentals, early turn timeliness, historically cheap.

Plug your own favorites into this system and see how they come out. I think you'll find it useful -- not for calling the bottom of the market, perhaps, but for figuring out what to own when you decide it's safe to put new money to work in technology.

At the time of publication, Jim Jubak owned or controlled shares in the following equities mentioned in this column: Applied Materials, Applied Micro Circuits, Atmel, Brocade, Ciena, Cisco Systems, EMC, Extreme Networks, Intel, Inktomi, LSI Logic, Mercury Interactive, Nokia, Nortel Networks, Nvidia, Oracle and Wind River Systems.