Investors are an arbitrary bunch, but chip-equipment investors are perhaps the most arbitrary of all. In the case of

Applied Materials

(AMAT) - Get Report

, for example, sometimes fundamentals matter, and sometimes they don't.

Last Wednesday, industry leader Applied Materials finally uttered the magic words investors have been longing to hear for three-plus years: "clear signs of an upturn."

Yet since then, shares of the equipment industry bellwether have taken a hit, bringing its peers down with it. Besides the natural "sell the news" reaction and the broader market's weakness, investors may finally be acknowledging that this upturn may be humbler than the last.

Shares of Applied closed at $23.74 Monday, down from the 52-week high of $25.61 they notched two weeks ago. Similarly under pressure,

KLA-Tencor

(KLAC) - Get Report

has lost $3.64 of its value since Applied's Nov. 12 report, ending Monday at $56.44.

Novellus

(NVLS)

is off $1.85 to $42.60 in the same time frame.

The recent stock-price moves suggest that once investors are through pricing in the upside -- which already may have begun to happen -- fundamentals, even common sense, suddenly may start to matter again.

Having bid up Applied shares 82% year to date, investors are suddenly showing a newfound uneasiness about the equipment industry's prospects, especially against the backdrop of its heady valuation. That's rare in a sector that typically flies by the seat of its pants. Chip-equipment stocks are widely acknowledged to trade more on momentum than business fundamentals.

Plenty of skeptics have long been arguing, futilely until just recently, that it's time for a reality check. Capital spending typically doubles from peak to peak in the tech cycle. But many expect this cycle's peak to fall below the prior one, suggesting that Applied should sport a suitably dressed-down valuation.

Semi-equipment spending topped out at $48 billion in 2000 and toppled steeply downward thereafter for two years. By 2002, spending stood at a mere $20 billion, with the industry expected to eke out 5% growth in calendar-year 2003.

Until recently, investors have been revving up AMAT shares as if they expected a redux of the late 1990s. On the surface, recent evidence of a rebound does suggest some parallels.

As Needham semiconductor analyst Cristina Osmena observed in a recent note, chipmakers are currently using around 90% of factory capacity, similar to the 94% utilization rates seen back in November 1999. Both periods were characterized by sequential orders improvement and an upswing in the U.S. economy.

The difference is that Applied now serves a smaller base of customers than it did then, as many chipmakers migrated to a fabless model to save costs in the downturn. Plus, companies that still have fabs are broadly expected to be somewhat more cautious in their spending.

The bottom line: Applied is expected to offer up growth of around 25% next year, compared with 40% growth in 2000, notes Osmena, who has a hold rating on the shares.

In light of the more relatively muted rebound outlook, AMAT's valuation looks outsized even after the recent selloff.

Based on Monday's close of $23.74, Applied trades at 47 times expected earnings for the fiscal year 2004 ending in October and 26 times projected earnings for fiscal year 2005. At 9.8 times trailing sales, valuation is approaching levels seen only during the 2000 peak, Osmena observed.

With that being the case, the analyst is doubtful that AMAT could eke out much more of a gain, given that it's now pricier than it was four years ago, when its growth outlook was better. She has a hold rating on the shares; Needham hasn't done investment banking for Applied.

On a similar note, Banc of America Securities' Mark FitzGerald predicts AMAT's valuation suddenly will start to matter when momentum stalls, which he believes could happen by the middle of 2004 -- barring what he calls a "1990s-type end-market recovery." He has a neutral rating on Applied; his firm hasn't done investment banking for Applied.

Until that happens, Applied probably will just coast on the ebb and flow of investor sentiment, confounding anyone who tries to impose logic on what analysts privately refer to as a trading vehicle. The stock's volatility only endears it to Wall Street, which finds abundant profit potential in its high beta. As proof of that affection, look no further than its heavy analyst following.

According to Thomson First Call, 27 brokers follow Applied, which claims a $39 billion market cap. In other words, it's managed to attract one analyst per $1.4 billion of market cap value.

To put that in perspective,

Intel

(INTC) - Get Report

has a market cap about five times as big as Applied, at $209 billion, but only four more analysts cover the company. It claims one analyst per $6.7 billion in market cap.

IBM

(IBM) - Get Report

, with a market cap of $154 billion -- about four times that of Applied -- has about two-thirds the analyst coverage, with a mere 19 analysts. Big Blue sustains only one analyst per $8.1 billion in market value.

The takeaway for investors: Don't assume Wall Street's obsession with chip-equipment stocks means that analysts can predict price moves with any reliability. As a case in point, Applied Materials' stock has consistently flouted conventional wisdom on valuation.