The Bargain Bin

When the economy was going full tilt, investors extrapolated rising corporate earnings further into the future than any expansion had ever lasted. Dumb. Booms always bust. And now that the bust is here, shortsighted investors are acting as if there will never be another boom. Dumb squared. Early in a downturn, as now--not late in an expansion--is when you should take the long view in determining if a stock is fairly priced.

And the long view says that, assuming a recovery in 2002, some battered tech giants are getting discreetly attractive.

Much has been made of the coincident slide in tech shares and tech earnings, a brutal one-two punch that has left some stocks down 80%--but no less expensive than during the bubble. How can that be? Stocks are most often valued relative to future earnings. If earnings prospects fall faster than a stock's price, the stock gets more expensive no matter how low it goes.

Consider telecom equipment maker Nortel. Last week it warned yet again that earnings would fall short. In fact it will lose money in the first quarter. Already hammered, the stock fell again, to $14, from around $40 in December. The consensus today is that Nortel will earn 14[cents] a share--down from expectations of 98[cents] four months ago, according to earnings tracker FirstCall. The revised earnings picture gives the stock a price/earning ratio of 100--more than twice its lofty P/E of 41 when the stock was much higher.

That kind of math plagues tech stocks across the board. Yet it's a mistake to think that way. It assumes that the nasty tech-spending slowdown dragging the economy into what feels like a recession will never end. It will. And when it does, the earnings prospects of market leaders like Nortel will improve rapidly. Factoring in a recovery, analysts estimate that Nortel will earn 58[cents] a share in 2002. On that estimate, its P/E is 24. Not bad for a company expected to grow 25% annually. That translates to a 0.96 PEG ratio, which is a measure popular with growth-stock investors. The PEG is the P/E divided by the growth rate. Under 1 is cheap. Put another way, they like the P/E to be lower than the long-term growth rate.

For profitable market leaders, a PEG up to 1.5 is fair, and by that standard a bunch of big names--Cisco, Oracle, Nokia, Verizon, Intel--are in the zone, even based on this year's depressed earnings. The risk in looking at things this way is that the earnings picture can sour further, and even long-term growth rates erode. So some money managers lop 10% off consensus earnings estimates and 20% off the generally accepted growth rates.

That discount still yields some attractive PEGS. Take Cisco. It hit $15 last week, giving it a P/E of 26 based on earnings this year of 57[cents]. It's growing at 27% a year for a PEG of 0.96. If earnings come in at 51[cents] and growth slows to 22%, it still has a reasonable PEG of 1.3. On 2002 numbers, the PEG becomes 1.2.

These stocks can still go down. Warren Buffett, for one, says tech remains expensive. But there's an old saw about cyclical companies sure to survive a recession: buy them when P/Es are high; often that means earnings are temporarily depressed, not that the stocks are overpriced. And yes, tech stocks are cyclical, but you knew that.

See time.com/personal for more on tech stocks. Dan joins The Money Gang each Tuesday at 2:15 p.m. E.T. on CNNfn