TCS Daily

Stocks Are Not Overpriced

By James K. Glassman - March 12, 2002 12:00 AM

The market has risen smartly in recent weeks, but the average stock is still down by more than one-fifth from its high of two years ago. Does that mean shares are still cheap? On the contrary, say analysts like James Grant. They're dangerously expensive.

In a recent op-ed piece in the New York Times, Grant pointed to research by the Leuthold Group of Minneapolis that shows that the market "would have to fall by 41 percent to reach the median valuation prevailing since 1957." In other words, for stocks to get back to reasonable levels (at least according to history), the Dow Jones industrial average would have to drop to about 6200. The Dow closed Friday at 10,572. It's a long way down.

Let me quickly say that I disagree with Grant and the other pessimists -- and I'll tell you why a bit later. But they offer a powerful message. You need to hear them out.
Their argument revolves around the word "valuation." All by itself, the price of a share of stock is meaningless. What counts is its relationship to something else -- such as the profits (or earnings) of the company that issued it, or the assets on the firm's balance sheet, or the dividends it pays. The most popular measure of such a relationship is the price-to-earnings, or P/E, ratio. Calculating a P/E is the main method of determining a company's valuation.

Compare two technology stocks: Oracle, which was recently trading at about $14, and IBM, trading at about $104. IBM has the higher price, but Oracle has the higher valuation. If you divide Oracle's earnings over the past 12 months by the number of shares the company has outstanding, you get about 43 cents. That's earnings per share, or "E." Now divide $14 (the price, or P) by 43 cents (E), and you get a P/E ratio of about 33. Perform the same exercise with IBM, and you get a P/E of 24. So Oracle's stock has a higher valuation; that is, it's more expensive.

What bothers Grant and other analysts is that the P/E of the market as a whole seems to be exceptionally high right now. How high? On Dec. 5, 1996, Alan Greenspan, the Federal Reserve chairman, warned of "irrational exuberance" in the stock market. At the time, the P/E ratio of the benchmark Standard & Poor's 500-stock index stood at 19. A year ago it was 24. Currently it is 29. The P/E of the Dow has shot up in a year from 21 to 28.

Beware, that "E" part of the P/E ratio is figured in different ways. Usually (as above) it's the earnings of the past four quarters. Sometimes it's the earnings that are estimated for the current calendar year. The Value Line Investment Survey has its own unique method: adding the actual earnings of the past six months to the estimated earnings for the next six months. This technique last month produced a P/E for the market of 20.1. That compares with just 18.0 at the last market top (May 2001) and 10.6 at the last market bottom (October 1987).

During recessions, like the one that appears to have just ended, earnings fall unusually fast, so PEs don't reflect normal business conditions. "But even measured against the peak operating earnings of 2000 and potential earnings when the economy recovers, stocks are high," writes H. Bradlee Perry, a veteran analyst for David L. Babson & Co., the Cambridge, Mass., investment firm. Perry estimates that, based on likely "normalized" earnings for 2002, the P/E of the S&P is at least 26. That looks like a big number. The average P/E ratio of the S&P between 1950 and 1999 was about 14.

M. David Testa, chief investment officer of T. Rowe Price Group Inc., the Baltimore mutual fund house, worries that stocks as a whole did not go through the usual wringing-out process during this bear market. Yes, the S&P fell a painful 37 percent from top to bottom. But what explains the price decline almost entirely is a drop in earnings -- not a drop in valuation. That's surprising in a bear market -- and disturbing to Testa. "On normalized earnings," he says, "the market never got nearly as cheap as it has even in an average market environment, much less a bear market." And, now that stocks have rallied -- the Dow is up 28 percent from its low the week after the terror attacks of Sept. 11, 2001, and the tech-heavy Nasdaq composite index is up 33 percent -- P/Es are way, way up there.

The question today, says Testa, is "When do company performance and fundamentals catch up" with valuations? Grant doubts they will.

My own view is that P/Es have a role in investing, but you shouldn't make a fetish of them. High P/Es and other valuation indicators are, emphatically, not predictors of market declines, as two economists at UCLA, Amit Goyal and Ivo Welch, have shown in an academic paper. In fact, Robert Hall of Stanford found that high P/E ratios tend to precede periods of unusually high earnings growth, and low P/E ratios foreshadow sluggish earnings.

In the early 1990s, P/E ratios were double the average of the preceding decade. But reported earnings during the 1990s soared, and so did stock prices. On July 1, 1991, the Dow stood at 2907, which represented a P/E of 22. That certainly would have scared off the fetishists. For the rest of the decade, P/Es averaged 25 and never got below the long-term average of 14. Yet the Dow kept rising. It ended 1999 at 11,497. Investors in the 30 boring Dow stocks (not racy high-techs) more than quadrupled their money, including dividends.

But doesn't history count for something? Yes, but market valuations can also change. In our book, "Dow 36,000," published in 1999, economist Kevin Hassett and I put forward the theory that, as investors came to realize (correctly) that stocks, held for the long run, weren't as risky as they had previously thought, they did the logical thing. They bid up valuations.

Kevin and I have taken some criticism over the past few years, but it seems the market is backing us up. P/E ratios have remained high -- indeed, they've gone higher. Yes, stock prices have dropped -- because "E" has dropped. Earnings in this recession have fallen more sharply than in any economic slowdown in at least 50 years. But "E" will recover. And if P/Es remain in the mid-20 range, then stock prices will recover, too.

Need more reassurance? Take the normalized P/E ratio of 25 and invert it; that is, turn it into an E/P ratio. It becomes 1/25, or 4 percent. That's the "earnings yield," or the percentage of your investment that the company produces in earnings -- that is, your share of the annual profits. Now compare the 4 percent earnings yield with the yield on a 10-year U.S. Treasury bond, which recently was 5 percent.

The Treasury bond wins, right? Not at all. Say you invest $1,000 in both the stock and the bond. The first year, the stock earns $40, but assume that earnings rise at a rate of 10 percent annually. After 10 years, thanks to compounding, the stock will be earning $104. The bond will still be earning $50. (Even at a sluggish growth rate of 6 percent, the stock's earnings rise to $72.) The stock is probably more risky than the bond, but if you own a diversified portfolio of stocks or a mutual fund, the difference in risk could be tiny.

Let me be clear. I am not saying that P/Es don't matter. Comparing two similar companies, I would much rather own the one with the lower P/E. It is, after all, cheaper.

What I am saying is that a high P/E ratio for the market as a whole is not a cause for alarm. If you believe that the U.S. economy will continue growing as it has -- at about 3 percent annually -- then P/E ratios at current levels, or higher, are justified.

Still, there's nothing wrong with being on the safe side. You can choose value-oriented mutual funds, whose managers search for bargains. For example, Dodge & Cox Stock fund has traditionally owned stocks with P/Es about 20 percent lower than the market as a whole. For the 10 years ended in 2001, it returned an annual average of 16.6 percent, nearly four points better than the S&P. Top holdings at last report included Golden West Financial (with a P/E of 13), Phillips Petroleum (11), May Department Stores (17) and Union Pacific (17). Dodge & Cox receives a five-star (highest) rating from Morningstar and charges an expense ratio of less than 0.6 percent -- or less than half that of the average fund.

Other value funds with excellent track records include Vanguard Growth & Income, Fidelity Low-Priced Stock (for small-caps), Legg Mason Value Trust, Weitz Mutual (mid-caps) and American Funds Washington Mutual.

Value Line's subscription service each week lists the companies with the lowest P/Es. Among those with the highest rankings for what the firm's analysts call "timeliness" and "safety" are Holly Corp., energy; Ryland Group, home building; Nissan Motor, the Japanese automaker; Ameron International, building materials; and Universal Corp., tobacco. All have P/Es (based on the Value Line method) of under 10.
Always search for low valuations in individual stocks and funds, but don't let a high market valuation scare you away from intelligent long-term investing.

This article first appeared in the Washington Post.

TCS Daily Archives