By Jack Hough
U.S. stocks are 37% cheaper than usual, based on one comparison of share prices to company earnings. By another measure, they are 25% overpriced.
Which one is right?
There are more than a dozen modern versions of that cornerstone of value investing: the price/earnings ratio. But newer isn't always better. Here's how to cut through some of the contradictory numbers and decide whether now is the time to load up on stocks or to sell.
Review: a P/E ratio is a company's stock price divided by a year's worth of its per-share earnings. The measure matters because stocks aren't abstract trading instruments; they represent part ownership in businesses. Earnings are one measure of the surplus these businesses generate on behalf of their owners -- money that can be used to pay dividends, buy back shares, expand into new markets or other pursuits. The lower a P/E ratio, the less an investor must pay for these things.
Using the most traditional P/E calculation, U.S. stocks are trading at 14.3 times earnings. That is, the Standard & Poor's 500-stock index is at about 1200, and the per-share earnings of the companies in the index totaled $83.87 over past four calendar quarters, according to Standard & Poor's.
That valuation is spot-on with the historical average: U.S. stocks traded at an average of 14.5 times earnings between 1872 and 2000, according to a study by the Federal Reserve Bank of Kansas City. Based on this measure, investors should sit tight.
One modern twist on P/E math is to use "operating" earnings rather than earnings as defined by regulators. The basic idea is to ignore charges and credits that don't stem from core operations. "There's no legal definition for operating earnings," says Howard Silverblatt, senior index analyst at Standard and Poor's. "It's more principle. I want to know how much I made from widgets last quarter, not how I financed them or whether I incurred a cost to fire people."
Based on operating earnings -- which are 8% higher than regulatory earnings -- the S&P 500 has a P/E of 13.2. But there's no telling whether that's low, because the record for operating earnings starts in 1988. That's too short and too bubbly a history to serve as a reliable benchmark. So investors probably shouldn't pay too much heed to operating earnings multiples.
Another "improvement": the use of polls of earnings forecasts rather than recorded earnings. Wall Street analysts expect S&P 500 members to report record operating earnings this year and then boost earnings by 14% next year. But how can such distant forecasts be reliable?
It's at least suspicious that Europe is flirting with financial crisis, America's economic recovery has begun wobbling and stock markets have roiled since summer, yet earnings estimates have barely budged.
"I think these forecasts are very high," says Mr. Silverblatt. "Current-quarter estimates might be safe, because companies are careful to manage short-term expectations. But after Thanksgiving, analysts will revisit their models and meet with companies, and 2012 numbers could move significantly lower."
The plain old trailing P/E--the one that suggests stock prices are normal--is at least based on known earnings, so it seems the best place to start in deciding whether stocks are broadly affordable. But it has a flaw. It doesn't adjust for swings in the business cycle. If the past year's earnings were unsustainably high, today's P/E might be deceivingly low.
One fix for that is to take an average of many years of trailing earnings. Yale economist Robert Shiller advocates using 10 years, adjusted for inflation. His "cyclically adjusted P/E" is just over 20, versus an average since 1881 of closer to 16.
There's another reason to believe, as Mr. Shiller's measure suggests, that stocks are more expensive than they appear. According to government statistics, after-tax corporate earnings are near a record high relative to worker wages. In the past, similar readings have foretold sharp earnings declines. If workers aren't participating in the boom, after all, the boom might not last, because workers are also customers. Today, it isn't clear whether the same will happen, because companies make more of their money overseas than in the past.
"The government data might be a warning," says Mr. Shiller. "It's difficult to quantify these things. This is history in the making."
Where does all this leave investors? With more opportunity than they might think. Even though U.S. stocks might not be broadly cheap, they're priced to deliver long-term returns of 4% a year after inflation, reckons Mr. Shiller. And that's about double what the 10-year Treasury bond pays.
What's more, the valuation gap between cheap stocks and expensive ones has rarely been wider, according to research by Brandes Investment Partners, a San Diego money manager. In the past, that has signaled that value stocks, including low-P/E ones, are due to outperform.
One way to get exposure to the cheap end of the market is with an exchange-traded fund like the Rydex S&P 500 Pure Value ETF. It's based on an index of stocks that are priced low relative to book value, cash flow, sales and dividends. They also tend to be priced low relative to earnings.
Another is through individual stocks. As of Thursday, at least 81 companies across a variety of sectors had share prices trading at single-digit ratios to trailing regulatory earnings, based on an analysis of Reuters Research data. Among them: Eli Lilly (LLY),