But this time there’s a twist: If investors are irrationally overvaluing hot young web startups, they also appear to be irrationally undervaluing not-so-hot, aging tech giants. Unlike the dot-com bubble of 1999, when giants like Microsoft (MSFT) and Cisco (CSCO) saw their stock valuations surge upward, big tech is languishing in 2011. These days, Microsoft, Cisco as well as other brand names like HP (HPQ) and Intel (INTC), are all trading below ten times their historical earnings. For the S&P 500, the average P/E is 17.3 times earnings.

In the 80s and 90s, the tech sector was the province of growth stocks. IT spending was increasing by 10%-15% a year, and investors who got in on the ground floor saw the stocks of Microsoft, Intel and Dell (DELL) rise as high as 50,000% over time. But with success comes slower growth. Microsoft and Intel are expected to see revenue and profit growth between 10% and 20% this year.

Normally, companies with moderate growth and P/E’s below 10 are considered value plays — stocks that have fallen so far below the value implied by their fundamentals they present investors a tempting bargain. Big tech currently has many hallmarks of value stocks: In addition to low valuations, they offer strong balance sheets and seasoned managers who take a conservative approach to risk.

In theory, the market will correct the imbalance, returning value stocks to their proper price. But this that hasn’t happened with big tech: Since the economy began to recover, their valuations have remained below average despite above-average profit growth. And although many analysts have tried, nobody has come up with a convincing answer to why this is happening, or whether big tech will rebound given time.

Last September, Bernstein Research analyst Toni Sacconaghi published a report that generated some discussion. Sacconaghi noted that tech stocks in the S&P 500 were trading on par with the broader index. That hadn’t happened since 1991. During the interim two decades, large-cap tech traded at a 30% premium to the rest of the market.

In early February, Scott Kessler, a seasoned stock analyst at Standard & Poor’s, calculated that P/E valuations of tech stocks in the S&P 500 still priced them at a 25% discount to the broader index. The aberration presented “considerable value in and appreciation potential for many large-cap technology names,” Kessler argued.
It made good sense through the lens of fundamental analysis, yet in the three and a half months since then the tech sector has continued to underperform the broader S&P 500 index.

And it’s not because of poor earnings. According to Bloomberg, the S&P 500’s information technology sector had a median operating margin of 21.6%, higher than the energy sector’s 18.1% margin and the financial sector’s 17.6% margin. Not only is infotech the index’s most profitable sector, profit margins for most companies have been steadily increasing. Yet the market is pricing them like yesterday’s fish.

Why is this happening? Sacconaghi and others have posited theories, but while they may account for part of the answer, none fully answer the mystery. One theory is that few tech giants report earnings according to GAAP, an accounting standard that takes a conservative approach to how companies record stock options as expenses.

This theory argues that that non-GAAP earnings of tech companies appear artificially high relative to industries that abide by GAAP. But this argument overlooks two things: First, that translating non-GAAP to GAAP is simple math for most analysts and institutional investors. And second, that many recent tech IPOs with obscenely high valuations also don’t follow GAAP. So why not punish them too?

Another theory is that big tech companies are hoarding too much cash: IBM (IBM) has $13 billion in cash on hand, Google (GOOG) $36 billion, Microsoft $50 billion. Investors are holding these stocks hostage — weighing down their prices (and the value of those stock options) until they pay a dividend.

This theory also fails to hold water. Such shareholder activism might work on one company, but it’s too lame to take on an entire sector. Microsoft, which has long paid a competitive dividend, still has a P/E below 10; while Apple (AAPL), which has never paid a dividend despite holding $29 billion in cash, has a P/E of 17. What’s more, Microsoft, IBM and HP have been returning more than $10 billion a year to investors by buying back shares. Big tech is returning cash to investors, and investors simply don’t care.

A third theory is that big tech is simply uncool — that not only is profit growth for these companies slowing, not only are their traditional markets are drying up, but their brands repel investors. This is certainly true for some tech giants. Nokia (NOK) has a vast customer base but lacks a smartphone platform to keep them loyal. Cisco is struggling to fight off competitors with lower-priced switches and routers.

But with other companies, it’s not necessarily the case. IBM and HP — which both struggle with below-average P/E ratios — are confidently projecting profits to grow 11% a year through 2014, thanks to new initiatives like cloud computing. Double-digit profit growth may not be Facebook-worthy, but it should appeal to rational investors.

And Microsoft, the king of operating software for desktop PCs, may be stuck in the post-PC age, but it’s adapting remarkably well: The Kinect sold 8 million units in its first two months, twice as many units as the iPad sold on its launch. The Xbox stands to benefit from Sony’s PR fiasco with the Playstation. And gutsy deals like the Nokia alliance and the Skype purchase suggest Microsoft has a future on the mobile web.

So while it’s tempting to cast off big tech as a bloated footnote to the Internet’s history, there remain some value investors who still believe the market will return to its senses. One of them is Bill Miller, the Legg Mason fund manager who made history by beating the S&P 500 every year between 1991 and 2005. Miller’s funds haven’t performed as well since then. But then again, value investing isn’t what it used to be.

In a recent commentary to his funds investors, Miller had a few words about Microsoft. After lamenting that nobody seems to want stocks “that are not ‘performing’, that have no momentum, and that are cheap,” he turned his focus on Microsoft. Although Microsoft has steadily posted earnings growth of 11% for the past decade, Miller noted, its stock trades at half its level in 1998, a year before the dot-com bubble went truly mad. That observation led to this wonky rant:

“[Microsoft] now trades at a similar price-to-earnings ratio as Pfizer (PFE), which is expected to grow at 3% to 4% vs. MSFT’s 15%, according to Value Line. Microsoft earns 44% on equity, generates almost $2 billion of free cash flow every month, yields significantly more than the market, last raised its dividend 23%, and has reduced shares outstanding by 2 billion in the past 5 years. The current valuation is completely irrational, but as Keynes so correctly noted, the market can stay irrational longer than you can remain solvent.”

Maybe that flurry of statistics means nothing to you, but to a value investor it’s a cri de coeur — a mathematical prayer that culminates in a plea to John Maynard Keynes, the economist whose cool mind in a previous era returned rationality to a mad market. But that era is long gone. And, for now, tech investors have gone mad again.

That is, math is still here, it’s just busy with Mark Zuckerberg’s algorithms. No one knows how long it will take for simple math to return to stock fundamentals. When it does return, investing in big tech will seem like a good idea in practice, and not just in theory.