PALO Alto, Calif. (CNN/Money) - This article originally appeared Sept. 4 and was updated Sept. 6. You've heard of the value trap -- stocks with rock-bottom P/Es lure in value investors only to crash even further.

Now consider the quality trap, the wrong-headed theory that it's always a good idea to invest in well-known established companies when they're down.

It's probably the single biggest justification for buying shares of Intel (INTC: up $1.28 to $16.39, Research, Estimates), today. Results for the chip maker came in light in the second quarter, and this week, the company trimmed sales forecasts for the third quarter. (see more) The stock is more than 50 percent off its 52-week high.

And yet some optimists would say that now is precisely the time to buy. There's no question Intel is a quality company, with solid management, killer intellectual property and solid control of its markets.

For all its problems, Intel is far and away the leading maker of microprocessors for personal computers. What's more, Intel's far-thinking managers were well aware three years ago that the PC market was becoming a stinker and that change was needed. That's why Intel has invested massively -- both through acquisitions and through R&D -- in the communications chip market.

But heavy investment does not a new monopoly make. Through shrewdness, savvy marketing and superior manufacturing Intel ended up with a monopoly on chips for PCs. In communications, it's going up against the likes of Texas Instruments (TXN: up $1.49 to $20.11, Research, Estimates), and a host of smaller concerns. No monopoly there. Oh, and the market temporarily stinks.

Which brings us inevitably to price. Back in June, when Intel's shares were trading above $22 (already down from the lows $30s just a few months earlier), I walked through the company's historical valuations with a goal of understanding what "cheap" meant.

Standing on the shoulders of my favorite chip-industry source, I pointed out that in 1996, when Intel was beginning a run that would put it in the business history books, its shares ranged in value from 9.5 times to 21.6 times that year's actual earnings. In June, Wall Street expected Intel to earn about 60 cents this year, making the shares worth about 37 times earnings -- and way overvalued.

Recently by Adam Lashinsky       Cisco: Above the fray Crapped out In defense (sort of) of Jack

Even today, however, you can't call the shares cheap. Estimates have come down to the 50-cent range, with a financial update expected from Intel on Thursday. So at around $16, Intel trades for roughly 32 times likely 2002 earnings. In other words, it's not even down to the peak of its pre-bull market valuation. "My price target is $7," says my longtime Intel source.

Even this chip maven agrees Intel is well positioned. But it's all relative. He doesn't think it will regain its former glory. And this is the problem investors face when looking at any "quality" company: You just don't know if it will be a quality company three years from now, especially in the technology world.

Remember, once upon a time AT&T, Polaroid and Xerox were high-quality companies too. Holding your nose and buying them when their stocks were down would have caused only hyperventilation.

Adam Lashinsky is a senior writer for Fortune magazine. Send e-mail to Adam at lashinskysbottomline@yahoo.com.

Sign up to receive The Bottom Line by e-mail.