As stock prices have fallen and Wall Street's ability to forecast earnings has deteriorated, however, using p-es as a shortcut has become much more dangerous for the average investor. "Anyone who is putting a disproportionate emphasis on the p-e is really setting themselves up for a problem," says Erick Maronak, director of research at investment firm NewBridge Partners.
WAY-OFF ESTIMATES. The root of the problem is with the "e" in p-e. Most often, the earnings number used in the calculation is a consensus of analysts' estimates for a company's operating earnings over the coming 12 months. Look no further than the Enron fiasco to see how easy it is for those estimates to be off by, say, several hundred million dollars. "In a lot of cases, earnings are hard to forecast, so it is hard to know how much meaning to give to those p-es," says Peter Cohan, an author and technology strategist.
Even when a company is completely forthright about its earnings prospects, analysts may still get their forecasts wrong in the current economic environment. While early indications show the economy is improving, few signs hint that corporate profits are picking up. With their myopic focus on a single industry and an incentive to pump stock prices higher, many equity analysts are likely to see an upturn no matter how scant the evidence.
This optimistic bias may be most affecting the technology sector, where brutal declines have followed years of torrid growth. "I don't think analysts doing their forecasts right now are really being realistic," says Cohan. "They are assuming there will be a turnaround sometime in 2002 that will find its way into the earnings of these companies. I think they are deluding themselves about how quickly the recovery will be here."
PAINTING PRETTY PICTURES. Making things even more difficult, companies have an incredible amount of leeway to pretty much offer up what they please to Wall Street as their operating earnings. Many are using a record number of one-time write-offs (for things like layoffs and inventory write-downs) to exclude expenses from operating earnings and make their results look better. Many companies also come up with their own "pro forma" method for calculating operating earnings when the official net income figure they report to the Securities & Exchange Commission casts them in a bad light.
What these companies are doing is perfectly legal. But add it all up, and you get lots of room for error when determining the "e" in the p-e. "It's the old garbage in, garbage out rule," says Chuck Hill, research director at Thomson Financial's First Call. "If your earnings number is no good, your p-e number is no good."
Solutions to the many problems with p-es mainly amount to doing more research (always a good idea). David Sowerby, a portfolio manager at Loomis Sayles & Co., says with all the questions about earnings these days, he's paying more attention to a company's price-to-sales ratio. Hill advises investors to look at the p-e not on 2002 earnings, but on 2003 earnings, when a more normal growth pattern for corporate earnings should resume.
PEER REVIEW. Maronak says he analyzes a company mainly by looking at its earnings growth rate compared with that of both its peers and the broader market. If the companies in the Standard & Poor's 500-stock index can increase earnings by only about 7% in the current environment, he thinks a stock like Forrest Labs (FRX
), which is increasing earnings at more than 25% a year when its industry is growing at 20%, is worth a current p-e of 50.
Such an approach brings into focus another drawback to putting too much emphasis on p-es today. Given the back-to-back declines in the market in 2000 and 2001, one would expect stocks to look a lot cheaper on a p-e basis than they did two years ago. Wrong.
The benchmark S&P 500 has about the same p-e it had during the peak of the bull market in the spring of 2000: 23. Even more startling is the p-e of the tech sector, which at 58 is more than twice as high as the broader market. How can this be? Simple: Earnings declines have been even sharper than stock-price declines.
WORTH THE PRICE? The high p-es are making some strategists fear that stocks -- especially the bellwether tech names that have run up the most since their September, 2001, lows -- are overvalued. But some analysts say a stock with a high p-e may be well worth the price. Maronak argues that companies with a lot of cash that are growing faster than their peers in this tough economy deserve high p-es. Investors have to dig much deeper into company fundamentals in a rocky market than they do in a bull market, when "a rising tide lifts all boats," he says.
In fact, Hill says the higher p-es shouldn't be all that shocking. "The p-e is sensitive to where we are in the business cycle," says Hill. "We should be putting a higher valuation on trough earnings and a lower valuation on peak earnings," he says. That means p-es were way too high at the peak of the economic cycle but are more appropriate now.
This deep into a recession in corporate profits, investors should be prepared to pay a higher p-e than they would at the peak of a bull market, Hill says. Investors should also be willing to pay a higher p-e when interest rates are as low as they are now, he adds. In theory, lower rates spur consumer and business demand for goods and services, which should boost earnings.
"Looking at the p-e is certainly a good starting point -- always has been, always will be," says Hill, who still believes analysts are forecasting unrealistically high earnings for the year. The bottom line: In a skittish market in which earnings quality is difficult to discern, investors need to make especially sure their glance at p-es is just the starting point for further research. Stone is an associate editor of BusinessWeek Online and covers the markets in our daily Street Wise column