(FORTUNE Magazine) – Even for prudent investors the pitch sounds achingly alluring: An all-star list of renowned jumbo-cap companies, from Microsoft to Coca-Cola to Cisco, are selling at anywhere from 40% to 85% below their peak prices. Not only are these players the best-in-class industry leaders, but their superior management has routinely produced returns for shareholders that wax the overall market. Sure, they've suffered for the past two years. But look beyond the craziness, and you'll spot a historic buying opportunity. The GEs and Dells, the argument goes, have again proved their pedigree by leading a recent 13% rally in the broader market. And the best news is that you're not too late. The stocks are still cheap. The time to board this deluxe train with a discount fare, the boosters say, is now.

But are these revered growth stocks really bargains? Or have they simply swollen into lumbering giants whose very girth means that rapid future earnings expansion of the kind that drives stock prices is beyond their means? To decide, we took a hard look at what matters--the numbers--for five large-cap powerhouses: the aforementioned Coca-Cola, Microsoft, and Cisco, plus Dell and General Electric. We don't know how much money the companies will earn. But we can forecast how much they have to earn to generate strong gains for their investors.

For 77 years the S&P 500 has handed investors, on average, total returns of 10.4% a year. That's a reasonable guide for the number we'd require to buy a Cisco or Coke right now. To achieve 10.4%, our tech candidates must double their share prices in seven years, or by 2009. Since GE and Coke pay dividends, their shares don't have to rise as fast to meet the target. GE can get there by boosting its stock price about 9% a year--from $27 today to $48.23 by 2009, an 80% increase.

But two crucial factors will look very different in 2009, and both will substantially raise the bar for future earnings growth. First, these companies will have far more shares outstanding because of growing stock options grants. The options hurdle is especially high for the tech contenders. Using Frederic W. Cook & Co.'s estimates on the portion of options that translate into shares, FORTUNE calculates that in seven years Dell will have an extra 250 million shares on the market, adding almost 10% to its total float. For Microsoft and Cisco, the number is more than 8%. The influx means that our tech candidates' market caps--and earnings--must be far higher to arrive at the 10.4% target than without the options flood.

Microsoft alone needs to increase its value by $50 billion in seven years just to counterbalance future dilution. Or it can spend a big chunk of its earnings, as in past years, on buying back shares. That's money it can't reinvest to grow the business. To keep it simple, we'll assume that Microsoft, Dell, and Cisco all see the 8% to 10% rise in their shares outstanding.

The second factor is that P/E ratios are far above their historical averages, and they're bound to fall. "From high levels, they always move back toward the mean," says Jeremy Siegel of the Wharton School. Over the past decade the S&P 500 has posted an average multiple of 26, far above the 77-year average of 16. It's reasonable to project that P/Es will fall to around that historical number (see "Waiting for the Sweet Spot" in this section.)

But great companies command far higher than average multiples. Over the past ten years Cisco's P/E has been more than double that of the S&P. For Coke the premium is 32%. So we'll leave those premiums in place. Hence, we'll grant Cisco a P/E of 33 in 2009 and allow Coke a multiple of 21. Surprisingly, GE's multiple has only matched the S&P average. So we'll forecast a ratio of 16 for 2009. But even that number may be optimistic. GE is now perceived chiefly as a financial services company (see "Can GE Light Up the Market Again?"). Because their businesses are so risky, financial services players merit multiples below the S&P average. Watch for GE's P/E to stay subdued (it's now at 16 based on the last four quarters). The big multiple for Cisco is also optimistic.

To return 10.4% a year--and fight the headwinds of growing dilution and declining multiples--how much do our contenders have to earn? The numbers are daunting. The hardest case is Dell. In its last fiscal year it earned less than $2 billion. By 2009 it needs to lift profits to no less than $7.9 billion, or 30% a year. To gauge how tough that is, consider that Dell has increased earnings just 7% a year since 1998. Even if it generates near-peak profit margins, to return 10.4% a year to shareholders Dell will need sales of $118 billion in seven years. That's more than 80% of the global PC market; today Dell's market share is 16%.

For Microsoft and Cisco the hurdles are lower but still formidable: Microsoft needs to lift profits 15% a year. For Cisco the target is 16%. Those numbers sound reasonable, given their huge growth in the 1990s. But Microsoft's and Cisco's earnings haven't risen since 1999. Generating 15%-plus profit increases would require vast improvements in margins and market share for what are now mature companies. The smart money says that won't happen.

For GE, the benchmark is lower, just 12% a year. But taken together, GE's industrial businesses--jet engines, turbines, medical systems--closely track the overall economy. Can GE grow at 12% while its market chugs along at 5%? It's conceivable but unlikely. Coke needs even steeper 16% growth in a soft-drink market that's expanding at around 7% a year. That means sales of $57 billion in 2009, almost three times that of 2001. To get there, Coke would have to capture virtually the entire global soft-drink market. No question, these are five great companies. But great investments? Not with these scary numbers.

Feedback? investing@fortunemail.com