For a moment it looked like the old standbys were back. In the stock market rally that lasted from the low of July 22 to the high of Aug. 22, former market leaders such as

Cisco Systems

(CSCO) - Get Report

,

Nokia

(NOK) - Get Report

,

Oracle

(ORCL) - Get Report

,

EMC

(EMC)

and even

Intel

(INTC) - Get Report

roared ahead with some of their old ferocity. Cisco climbed 16% in the month, Nokia 19%, EMC 19%, Oracle 21% and Intel 5%.

It was enough to remind an investor of the glory days of 1999, before the bubble burst and the technology stars that had led the bull market higher began their collapse. In that one year, to refresh your memory, Cisco gained 131%, Nokia 219%, EMC 157%, Oracle 290% and Intel 39%.

But it was different this time. This time these former leaders didn't lead the general market; they followed it instead.

From July 22 to Aug. 22, the

Dow Jones Industrial Average

climbed 16%. And when the Dow went into reverse, falling 7% from Aug. 22 through Sept. 12, these stocks dropped with it. During that period, as the rally collapsed and the bear market decline resumed, Cisco fell 14%, Nokia 5%, EMC 21%, Oracle 15% and Intel 18%.

Leaders to Laggards

What explains the performance of these stocks? And why have these stocks changed roles from market leaders to market followers?

If you look purely at the estimates for future earnings and earnings growth, their decline makes no sense. The future, according to the estimates, is very bright. It's only when you consider the probability that these estimates may not be accurate that you begin to see why these stocks are struggling.

A quick glance shows why three of these stocks are struggling. Intel, EMC and Oracle all have growth problems in 2002. Intel is projected to grow earnings per share by just 5% in 2002, and Oracle is expected to show a 14% decline from 2001 levels. EMC is projected to show a 150% drop in earnings on a loss of 4 cents a share in 2002.

Given the still-declining earnings at Oracle and EMC and the anemic increase at Intel, the stocks are still expensive. At a PEG ratio (the price-to-earnings ratio divided by the earnings growth rate) of 5.8, Intel's growth is expensive enough to deter anyone right now. EMC and Oracle's PEG ratios can't be calculated.

Head Fake from Nokia and Cisco

But what about Nokia and Cisco? Using the estimated earnings per share growth for 2002, the stocks aren't any more expensive than two I consider potential market leaders. Nokia, at a PEG of 1.5, is still cheaper than the 1.6 PEG of Dell and the 1.75 PEG of food-distribution giant

Sysco

(SYY) - Get Report

-- and Cisco's PEG of just 0.17 makes the stock seem an astounding bargain.

Moreover, the projected growth rates over the next five years certainly would put those two stocks in the same class as Dell (at 17% projected growth annually over the next five years) and Sysco (at 14% growth per year).

What gives?

It's simple: overly optimistic growth projections by analysts.

That's why it's essential that you test the probability of growth projections for yourself before you buy anything in any market.

In the case of both Nokia and Cisco, investors have set a low price on the growth projections for these stocks. That, plus the difficulty these stocks face in climbing above technical resistance in any rally, is an indication that the stock market has assigned a very low probability to these estimates. These estimates may be the best numbers that any analyst can come up with, but the market doesn't have a lot of faith in the assumptions.

For example, Wall Street analysts projected that Cisco will grow earnings per share an additional 44% in the fiscal year that ends in July 2003, and that the company will show annual earnings growth of 19% on average over the next five years. If the networking business shows strong signs of recovery in 2003, and if companies start to increase their capital spending on information technology, and if telecommunications companies end their budget lockdowns and are willing to buy from Cisco instead of one of the traditional vendors, then those projections will be on the money.

A Matter of Timing

But projections from other equipment vendors put the recovery off into 2004. That means these analyst estimates are anything but certain. And the projections get even less certain if you consider Cisco's past strategy of sacrificing earnings for revenue growth. Because of that approach, the company's earnings per share climbed only 11% annually on average over the last five years.

The estimates at Nokia face uncertainties as well. Wireless phone companies -- especially those in Europe -- have cut capital spending even more drastically than the North American landline companies that buy from

Nortel Networks

(NT)

,

Lucent Technologies

(LU)

and Cisco (Cisco hopes). For example, in its most recent update for investors, Nokia lowered its projected growth for 2002's third quarter on weaker-than-expected sales of wireless networking equipment. Sales would fall 5% from the levels of the third quarter of 2001, itself a weak period. When will the decline finally end? Best estimates now say early 2004.

So with this much uncertainly surrounding the accuracy of these projections, investors are demanding a discount before they buy these stocks. If we could be certain that Cisco would grow earnings at 19% a year for the next five years, we'd all rush out to buy it. In that case, Cisco would indeed be one of the leaders of a resurgent bull market.

Thinking about this group of stocks in this way also marks out the path that these shares will have to follow in order to climb at a rate equal to, or better than, that of the market as a whole. Gains in stocks such as Cisco and Nokia -- and even in those of EMC, Intel and Oracle -- don't depend on increases in analyst estimates. Instead, lasting price appreciation hinges on improvements in the certainty of these projections.

When wireless phone companies start increasing their capital budgets, and when Cisco's sales growth is a clear result of growth in the size of its market, rather than from taking share from competitors, then these stocks will rise -- without analysts adding a single penny to their growth projections.

And until that happens, it really doesn't much matter whether this price-to-earnings ratio or that projected growth rate says the stock is expensive or cheap.

Jim Jubak owned or controlled shares in the following equities mentioned in this column: Lucent Technologies. He does not own short positions in any stock mentioned in this column.