Jim Jubak

I keep hearing that technology shares are cheap.

The numbers seem to back that up.

Computer stocks trade for just 9.3 times reported earnings before interest, taxes, depreciation and amortization, according to Bloomberg. That's just 1.3 times the multiple for the Standard & Poor's 500 Index ($INX) as a whole. And that's the smallest premium for computer stocks since Bloomberg's data began in 1998.

The price-to-earnings ratio for the companies in the information technology sector as compiled by Standard & Poor's is just 14.8. The multiple for the entire index is 14.7. The 14.8 multiple for the information technology sector is its lowest since December 2009.

And if you look at individual stocks, the numbers seem to make the argument even stronger. Cisco Systems (CSCO.O), one of the top technology names for decades, trades at a P/E ratio of just 11.5. The Intel (INTC.O) P/E is just 10.2.

But I just don't buy it. I think technology stocks as a whole are pretty accurately priced. The big companies that dominate the sector, the names we all recognize, may even be slightly overpriced. And the true growth companies in the sector may be attractive buys on their growth, but they sure aren't cheap on their price-to-earnings ratios.

I think the whole "technology stocks are cheap" argument fails to consider exactly how much the technology sector has changed from the good old days and how sweeping the revolution is that is now turning the sector upside down.

Let's take a look at the assumptions in the technology-is-cheap argument one at a time.

There's an assumption, for instance, that Cisco Systems shouldn't trade at a trailing 12-month P/E ratio.

Why in heavens not?

Here's a company that belatedly admitted that it pursued a profit-margin-killing strategy of heading off into the consumer market and that it needs to refocus. Then it presented a clearly inadequate reorganization plan.

Operating earnings -- a figure that doesn't include any one-time charges -- will grow, Standard & Poor's calculates, by just 6.7% in the fiscal year that ends in July and by just 8.4% in the fiscal year that ends in July 2012.

That 11.5 multiple doesn't look so out of line to me.

And lest you think this is simply the raving of someone who has recently lost money on an investment in Cisco (which I have), take a look at the other big names in the sector. Does Intel's 10.01 P/E look out of line with projected earnings growth of 7.3%? Does Hewlett-Packard (HPQ.N) look cheap at a price-to-earnings ratio of 7.9 against projected earnings growth of 9.5%?

Projected growth is low
Microsoft (MSFT.O) does look cheap against fiscal 2011 earnings growth of 25% at a trailing-12-month P/E of 9.7%, but not so cheap when you notice that earnings growth is projected to drop back to 10.3% in fiscal 2012. (Microsoft is the publisher of MSN Money.)

Compare these valuations with those of General Electric (GE.N) at a price-to-earnings ratio of 14.8 and a projected earnings-growth rate of 18%. Or consider an even stodgier Verizon Communications (VZ.N) at a P/E of 16.4 and a projected growth rate of 21%.

To get past the anemic rates of projected growth, you could argue (and I've heard analysts and investors make this case) that the consensus projected earnings-growth rate is too low. These technology companies are set to grow faster than this, you could say, and therefore they are cheap at these P/E ratios.

I'd find this more convincing if these technology companies were acting as if they agreed. Instead, I see one after the other of these giants repositioning its stock as a value instead of a growth stock by raising dividends to the levels that appeal to value investors. Intel now yields more than General Electric, while Microsoft pays only slightly less. Even Cisco Systems has added a minuscule dividend. Growth stocks don't have to pay dividends to lure investors, and they've got better things to do with their cash -- like invest in new business opportunities. Apple (AAPL.O) doesn't pay a dividend. Neither does Google (GOOG.O).