(FORTUNE Magazine) – Five years after the NASDAQ began its free fall, the case for jumping back into tech stocks can appear almost sober and more than a little convincing. While the Dow has regained much of the ground it lost after the bubble burst, the Nasdaq is still languishing 60% below its March 2000 peak. Moreover, after a relatively strong showing in 2004, it's down 10% so far this year. Why wouldn't investors leap at the chance to buy great names like Microsoft, Dell, and Cisco at steep discounts to their late '90s prices? It's time to let reason prevail, the talking heads on CNBC and Fox often argue, and stop wantonly punishing the Nasdaq for past sins.

If investors require more coaxing, they need only look at where many of America's most prestigious mutual funds are putting their cash. Fidelity's famed Magellan Fund has 17% of its portfolio in technology right now. Vanguard's $8 billion Capital Opportunity Fund is betting 31% of its money on tech. Plus, virtually every one of the 250 growth funds tracked by Lipper has tech as its largest sector holding, with an average weighting of about 25%.

Well, we're here to tell you that the highly paid investment pros are making a bad bet and that tech is, in fact, an absolutely disastrous place to invest right now. Why? The big and small players in telecom, software, the Internet, and chips are still trading at Brobdingnagian valuations that can't be supported by reasonable math. Their earnings are still puny compared with their stock prices. And given their modest prospects for growth, it's virtually impossible for them to expand profits fast enough to reward investors. "The problem is that the bubble never fully deflated and investors never really ended their overdone love affair with tech stocks," says Cliff Asness, co-founder of hedge fund AQR Capital. "When you have a category that's as expensive as tech stocks appear now, the future returns are typically low."

For the purposes of this argument, we'll use the Nasdaq 100, an index of the Nasdaq stocks with the largest market capitalization, as a proxy for the technology sector. The Nasdaq 100--which has a corresponding exchange-traded fund with the symbol QQQQ--encompasses most of the big tech names, from Intel and Oracle to Yahoo and eBay. To simplify, let's treat the Nasdaq 100 as if it were one big company and see whether you'd want to buy its stock.

Today the Nasdaq 100 companies are earning $55 billion a year, based on their cumulative profits over the past 12 months. Their combined market cap is $1.82 trillion. So the price/earnings ratio is 33. That's already a daunting number. But it's far from the whole story. Companies don't have to subtract stock-option expense from earnings, even though it's a real cost to shareholders. So let's adjust for the options factor, which imposes an extra expense in the form of stock dilution.

For the Nasdaq 100, the hidden cost of options is staggering. In the most recent fiscal year, option expense hit $1.3 billion at Intel (17% of reported earnings), $1.2 billion at Cisco (28%), and $800 million at Dell (27%). At a dozen companies, including Broadcom and Mercury Interactive, option expense was bigger than reported earnings. All told, option costs not included in official net earnings reached $12 billion for the most recent fiscal year. That reduces earnings for the Nasdaq 100 from $55 billion to $43 billion. Hence the real P/E ratio, the one that matters to investors, isn't 33 but 43.

If you're buying in at a P/E of 43, how fast must earnings grow to hand you a decent return? Let's assume shareholders want a five-percentage-point premium over the rate on the ten-year Treasury note to compensate for the risk of owning stocks. That's the spread that equity investors have been pocketing for the past 50 years, and it represents a minimum, because they might demand even more to wade into the volatile tech sector. Since the Treasury yield is currently 4.5%, investors should require a minimum return of 9.5%. The Nasdaq's dividend yield is a mere 0.3%, so it will provide only a tiny part of that return. The rest must flow from capital gains. That means that the Nasdaq 100 companies must lift their share prices by 9.2% a year. In ten years the market cap of the Nasdaq 100, compounding at 9.2%, would have to rise from $1.8 trillion to $4.4 trillion, or 140%.

The rub is that for stock prices to climb 9.2% a year, earnings will have to grow much faster. That's because the Nasdaq 100 can't possibly sustain its P/E of 43. No multiple for a major index stays that high for long. We asked Steve O'Byrne of Shareholder Value Advisors, an economic research firm, to calculate how rapidly earnings must rise to propel stock prices 9.2% a year. O'Byrne reckons that profits must climb 16.1% annually over the next ten years, to a total of $190 billion, or 4.4 times the figure today. That would still leave the Nasdaq 100 with a formidable P/E of 23, meaning that investors a decade hence would still be flush with optimism, expecting tech earnings to rise in double digits for years to come.

Unfortunately, the idea that profits can grow at 16% a year for any sustained period is a pipe dream. First, the profits for the Nasdaq 100 are extremely erratic; they typically drop sharply in recessions. So they need to generate huge gains in good times to achieve the 16% average. In 2000, for example, the ten biggest market-cap companies in the Nasdaq 100, a group that includes Microsoft, Intel, Oracle and Amgen, earned a total of $36 billion. In 2001, profits dropped to $9 billion. They only climbed back to $34 billion in the latest fiscal year. "Recessions come around every four to five years," says Jim Stack, founder of InvesTech, an asset-management firm in Montana. "Achieving 16% growth across the cycles doesn't add up."

Second, the Nasdaq 100 is dominated not by startups built for rapid growth, but by yesterday's jackrabbits that have evolved into big, maturing players. Analysts polled by IBES predict that Microsoft will increase earnings by 10% in fiscal 2006 and that Cisco will grow at 12%--and that's with a favorable economy at their backs.

So how far must the Nasdaq 100 sink before it becomes a good buy? Our chart, provided by Reuters StockVal, gives some tantalizing clues. (The chart shows a current P/E of 30 but that isn't adjusted for stock-option expense or for nonrecurring charges.) Taking out the huge run-ups before the 1987 crash and the start of the late-1990s bubble, the P/E of the Nasdaq 100 has averaged around 23. The best bet is that it falls back to around that number or even lower. That implies that the Nasdaq as a whole must fall to around 1000, or by 50%, before investors should get back in. So ignore the happy talk and hoard your cash until the inevitable, wrenching correction makes tech a good buy again. It's bound to happen. The math says so.

Feedback? investing@fortunemail.com