Chip valuations have seen frantic double-digit gains in just the last week, outpacing the broader market, though there's no evidence that end markets have strengthened. In light of that, investors could be ready to take some money off the table.

The Philadelphia Stock Exchange Semiconductor Index is up 18% since March 11, easily besting gains of 10% by the

Nasdaq

and 8% by the

S&P 500

.

An elite handful of semiconductor stocks have done even better than the chip index. In the same period,

Linear Tech

(LLTC)

is up 20%,

Xilinx

(XLNX) - Get Report

is up 19%,

Altera

(ALTR) - Get Report

26% and

Maxim

(MXIM) - Get Report

18%.

While stocks seem to have risen partly in hopes the Iraqi conflict will be resolved, there aren't enough fundamental reasons to justify their gains. Pacific Growth Equities analyst Jim Liang, who downgraded all the above four to underweight from equal weight, says that over the past couple of months, a big disconnect has developed between the robust performance of semi stocks and the lackluster outlook for their end markets.

Granted, three of the companies he downgraded have offered upbeat comments on business in the past couple of weeks. As a quick recap, Xilinx said its revenue will be at the high end of a previously outlined range, while Altera raised its first-quarter revenue guidance to 4% sequential growth, from its previous forecast for sales to stay flat, plus or minus 2%. And Maxim's CEO made note of favorable business trends at a recent investor conference. Pacific Growth has no banking relationship with any of the companies Liang downgraded.

But that's not cause for much excitement: Liang says the analog and PLD companies typically see seasonal strength at this time of year. "Investors should not equate seasonal strength with cyclical strength," he maintains. "I don't believe this is a cyclical recovery."

In fact, he says, the March quarter was seasonally strong in 2002 for the sector, only to see business fall off in the ensuing June and September quarters, mostly because of end-market weakness. "Given the data points so far, despite the seasonal strength at the chip level, the end-market data points in our view are signaling more weakness ahead," he says. "As such, we think going into the next two quarters, there's a level of risk that positive demand trends may not be sustained."

Among the worrying signs:

Gateway's preannouncement reflects soft demand for PCs.

Intel's flash shortfall reflects badly on cell-phone demand.

3Com's ( COMS) preannouncement and the recent commentary from Foundry Networks ( FDRY) underscore weakness in enterprise networking.

Agrees one buy-side analyst, "My guess is the run-up in stocks is probably a little premature in the near term." The question, he adds, is whether the run-up is "discounting something six months out, nine months out or 12 months out."

Typically, chip stocks may start to move up about half a year in advance of a recovery. But the buy-side analyst believes IT spending won't start to accelerate until next year, assuming corporate profitability improves. In short, even assuming recovery is in the offing, chip stocks are clearly getting ahead of themselves.

Right now, the four downgraded by Pacific Growth trade at an average of 33 times calendar year 2004 EPS estimates. That's above even the high end of the historical range for the group. Since 1992 (excluding the bubble year of 2000), their forward P/Es have averaged between 15 and 30.

"In the best case, in the event of a cyclical recovery ahead, forward 12-months EPS should be around 25 rather than 33," says Liang.

Could sentiment related to a looming war explain the rise in chips? Lately the group has staged several strong rallies on signs that war is imminent. But most analysts say a semi recovery hinges primarily on economic recovery. A resolution of the conflict in Iraq is likely to offer only a marginal benefit.

Surveys of corporate CIOs show they're buying (or not buying) new equipment mostly on the basis of their assessment of the economy rather than the conflict in Iraq, the buy-side analyst points out.

"We think the resolution

would be great in terms of removing some of the uncertainties, but we are in a school of thought that the end-market weakness can only partially be blamed on the uncertainties due to pending war with Iraq. To a larger extent, we think it's more due to the overall sluggish economy," says Liang.

"Assuming we achieve a quick victory and the uncertainty is resolved -- and that's a big if, right -- then people will start to focus more on fundamentals. That's when I think we'll see a real test for the market," says Liang. "And if our thesis is right and the overall economy is still sluggish and a postwar strong rebound is unlikely, then

chip stocks will be subject to downside risk."