SAN FRANCISCO -- One painful lesson investors have learned in the past 15 months is that most tech stocks are cyclical. An axiom among value stock managers is that the time to buy cyclicals is when their fundamentals look weakest. If you hold both these as truths, is now the time to consider buying tech?

Analysts who decry tech's cyclicality and simultaneously say the group is still too expensive "want it both ways," said Gary Farber, partner at

Nightingale & Farber Capital

, a Seattle-based hedge fund. The best time to buy automakers or any other cyclical is "on trough earnings when they're very expensive" on a price-to-earnings basis, he noted. "Why not treat tech like other cyclicals?"

Buying tech when it's expensive on a P/E basis

because

it's cyclical is a provocative concept. Adoption of the mindset may partially explain the

Nasdaq Composite's

33% rise from its April 4 low. (However, encouraging earnings news from

WorldCom

(WCOM)

and

PeopleSoft

(PSFT)

, plus optimistic comments from

Intel

(INTC) - Get Intel Corporation (INTC) Report

today failed to aid the Comp, which fell 1.2%.)

Notably, Farber and partner Walter Nightingale are decidedly

negative

on the fundamentals for most tech firms right now and are wading cautiously into the sector. But while admitting "we're not that smart to call the exact date" of a bottom, the hedge fund recently started getting "more aggressively long" -- as much as 60% -- than anytime since its launch in April 2000, Farber said. Currently, the $3 million hedge fund, up about 4% year to date, is running 45% long and 5.5% short. (Nightingale & Farber also run about $21 million in managed accounts.)

The hedge fund's long positions include the common stock of

Qwest Communication

(Q)

,

Sonus Networks

(SONS)

and

Sprint

undefined

, as well as call options on

Lucent

(LU)

and

Nortel

(NT)

. Obviously Nightingale & Farber isn't the (ahem) biggest fish in the pond -- and tech stocks only represent about 20% of its long positions -- but it's certainly not alone in searching for a rationale to buy (or hold) tech.

Good Idea, Wrong Time to Execute

The big caveat to the notion investors should buy tech now based on the laws governing cyclicals is that the sector's earnings have yet to hit their nadir. That's the prime reason many observers expect the nascent tech rally will soon crumble (and smell) like so much blue cheese.

"Obviously

stocks will turn before the bottom but you have to have some sense you're approaching that bottom," said Chuck Hill, director of research at

Thomson Financial/First Call

. "I don't think you're seeing that: The rally would appear to be premature."

Tech-sector earnings expectations remain "in free fall," Hill said, losing on average two-thirds of a percentage point a day. Heading into today, negative tech preannouncements in the second quarter totaled 203, up 77% vs. the same time in the first quarter, the current record-holder for warnings.

Looking ahead, the researcher observed tech earnings are now expected to fall 49% in the second quarter, 36% in the third and 11% in the fourth quarter vs. respective year-ago levels. But analysts are currently expecting tech earnings to grow 48% in the first quarter of 2002.

"We're talking one hell of a recovery. Is that realistic?" Hill wondered. "It doesn't sound to us like a backdrop to be loading up on tech stocks."

J.P. Morgan

strategist Thomas Van Leuven expressed a similar view (echoing the

dour forecasts of market strategist Douglas Cliggott).

TheStreet Recommends

"Our sense is that we're not at the trough yet," Van Leuven said. "Tech-sector earnings could continue to come down for a while yet. We don't think that expectation is reflected in

today's prices."

In mid-March, the consensus expectation was for S&P tech-sector earnings to fall 4% this year, he said, citing figures from

I/B/E/S International

(acquired in September by First Call's parent

Thomson Financial

). In mid-April, the consensus had fallen to a decline of 22%. Yet for 2002, the consensus is for tech-sector growth above 30%.

"Earnings expectations have come down a lot for the immediate quarters but it still seems to us like a fair amount of optimism a little farther out," Van Leuven said. "That makes us think the earnings-revision process ... has more to go. Next year's numbers still look increasingly optimistic to us."

At the risk of piling on, Thomas McManus, equity portfolio strategist of

Banc of America Securities

, offered yet another reason to dispel the hopeful thesis proposed at the top of this column.

In addition to believing tech earnings have yet to bottom, McManus noted that because technology changes so rapidly, tech franchises are fleeting. Save for a handful of firms such as

IBM

(IBM) - Get International Business Machines (IBM) Report

,

Motorola

(MOT)

, and

Hewlett-Packard

(HWP)

, few firms have survived multiple tech cycles.

Tech holdings can "go to zero" if not chosen wisely, the strategist warned. "To compensate you for that risk, the growth potential of the entire sector ought to be available at a discount" to the overall market.

In mid-April, the P/E-to-growth (or PEG) ratio of the S&P 500 tech sector was 1.42, he reported, while the nontech portion of the S&P had a PEG of 1.25.

Again, more empirical evidence for why it still pays to be wary of tech, even as many fund managers find themselves with fingers itching to trigger tech purchases. If and when that occurs, the quandary for investors still overweight tech will be whether to "double down" or take defensive measures. Those who've learned the

most

painful lesson of the past 15 months shouldn't have to quibble much.

The Never-Ending Story

A long time ago (but in the same galaxy) I

wrote critically about

PurchasePro.com

(PPRO)

, which today fell 26% to $3.

Shares of the B2B software developer have fallen precipitously from their split-adjusted high of $75.87 in conjunction with the broader dot-com sector. But this week, PurchasePro.com confirmed long-standing fears it was playing

fast and loose with its public statements and

accounting. The firm preannounced an earnings miss the day of its scheduled reporting date, which was

subsequently postponed.

Today, the company reported a first-quarter loss of 26 cents a share vs. a mean estimate of an 8-cent loss. My San Francisco office mate

Joe Bousquin

has done

yeoman's work following the PurchasePro story in recent months.

Way back when, I wrote that PurchasePro.com "in many ways typifies the New Economy phenomena -- for good and bad." Looks like I was half-right.

Aaron L. Task writes daily for TheStreet.com. In keeping with TSC's editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet.com. He also doesn't invest in hedge funds or other private investment partnerships. He invites you to send your feedback to

Aaron L. Task.

Check out

TheStreet.com's

new portfolio tracker. The new tracker, powered by

Money.net

, provides streaming, real-time quotes so you can track your investments throughout the day. The real-time tracker includes all exchanges fees and permits you to track more than 200 stocks at the same time. There's a 30-day free trial and it costs $9.99 a month. For more details, click

here.