Are stocks overvalued? Federal Reserve Board Chairman Alan Greenspan raised the issue most famously in December 1996 when he said he was worried about “irrational exuberance” in the stock market.1 At that time, the Dow Jones Industrial Average,2 the oldest and most popular index of stock prices, was about 6500. It is now over 9000, and Greenspan has backed off from making comments while other government officials have chimed in. Treasury Secretary Robert Rubin suggested in early May that investors use more “rigor” in evaluating their investments. A day later, Alice Rivlin, vice-chairwoman of the Fed, said that she saw “cause for concern about the values in the stock market,” but added, “That doesn’t necessarily mean it’s a bubble.” Between the comments by Greenspan and Rivlin, the financial press has kept the question of overvaluation alive, particularly as stock prices have reached new heights—7000, 8000, 9000.

It is hardly surprising that people are worried about stock prices because they have been rising so rapidly for such a long time. Despite setbacks in 1987 and 1990, stock prices have increased more than ninefold since 1982, when the Dow was less than 1000. Just the current phase of the upswing, which began in October 1990 with the index slightly above 2000, is considered the longest “bull market” in history (see Figure 1 on page 24, which also shows that when the long-term record of the Dow is drawn on a ratio scale which gives the same weight to equal proportional changes in the index, then the current rise in stock prices is less dramatic when it is compared to the pattern during several earlier periods). As a result of the sustained run-up in stock prices, total returns (i.e., dividends plus capital gains) on the stocks in the S&P 500, a more representative market index than the Dow, have averaged about 18 percent a year in the 1980s and 1990s, compared to about 11 percent between 1926 and 1980.

The rise in stock prices has been driven largely by improvements in the economy, principally rising corporate profits, falling inflation, and falling interest rates. Profits of the companies in the S&P 500 index more than doubled between 1982 and 1996, growing more than 6 percent a year on average. But because share prices rose even more rapidly, the price of stocks relative to a dollar of earnings (the price/ earnings, or P/E, ratio) also rose, from about 8 times earnings to more than 20; it is now about 24, roughly 50 percent above its long-term average of approximately 16 (see Figure 2 on page 24). Earnings growth and the rise of the P/E ratio to only its average level account for about two thirds of the rise in stock prices from 1982 to 1996. Investors have bid up stock prices even further because they anticipate growth in earnings to continue, and because interest rates have fallen sharply since the early 1980s, making investments in stocks even more attractive than in bonds.3

So far investors have been rewarded for assuming that corporate profits will continue to grow and that interest rates and inflation will remain low. The worry is that their expectations have become so high that when the economy slows down, or inflation accelerates, they will be disappointed and will dump shares, possibly producing a market crash. And because Americans now have about 28 percent of their assets in stocks, more than in any other asset, and more than at any other time in the last fifty years, there is a further fear that any sizabledecline in stock prices will have a much greater effect on the economy than in the past, creating a vicious cycle. When the market crashed in October 1987, for example, stocks accounted for only 13 percent of assets. The crash had little effect on the economy and was relatively short-lived.4

Anxiety about stock prices is thus to a great extent concern over whether the economy can continue to grow at a steady pace without a rise in inflation. But neither the path of the economy nor changes in stock prices can be predicted with much confidence. Seventy years ago, not long before the great stock market crash of 1929, the Federal Reserve wrote: “There is no way of knowing beyond question how far this recent rise in stock prices represents excessive speculation and how far a readjustment of values to increased industrial efficiency…and larger profits.”5 Because that observation is still true, and always will be true, we can’t really know whether stocks are overvalued. But other things are known about financial markets that can guide investors.

Perhaps the most important fact about investing in the stock market is that, over the long term, stocks in general have been a vastly superior investment. Nobody develops this theme more convincingly than Jeremy Siegel, a professor at the Wharton School of the University of Pennsylvania. As can be seen in Figure 3 (on page 25), which is reproduced from his book Stocks for the Long Run, a dollar invested in a representative group of stocks in 1802 would have grown to $559,000 in 1997 after adjustment for inflation, which reduced the value of the dollar to seven cents over this period. By comparison, in the same period a dollar invested in long-term government bonds, short-term bills, or gold, would have grown, after inflation, only to $803, $275, and $0.84, respectively. In other words, the real return on stocks over almost two hundred years was 7 percent a year, compounded, compared to 3.5 percent for bonds and 2.9 percent for bills.6


Moreover, the superiority of stocks grows, and their riskiness falls, the longer they are held. For example, stocks outperform bonds and bills about 60 percent of the time over any single year, but about 70 percent of the time if they are held for five years, 80 percent over ten years, and more than 90 percent of the time if held for twenty years. Unlike bonds or bills, stocks as a group have not had losses over any twenty-year period since 1802. It is true that it would have taken about fifteen years for stocks bought just before the 1929 crash to recover their value. Since World War II, however, the longest it has taken for an investment in a representative sample of stocks to show a profit has been three and a half years, between December 1972 and June 1976.

Of course, the long-run superiority of stocks as a group doesn’t guarantee results in any particular period, or for any particular stock. It is naturally tempting, at a time like this, to believe one can avoid a possible downturn in the market. But the second important thing we know about the stock market is that it is difficult to predict short-term changes in stock prices, and the consequences of mistiming the movements of the stock market as a whole can be so severe that it doesn’t make much sense to try. In general, an investor who tried to predict short-term changes in share prices would have to be right about 70 percent of the time to beat the market. During the 1980s, for example, the average real return on the S&P 500 index was about 12.5 percent a year. But an investor in that index who missed just the ten best days of the decade would have realized only 7.5 percent after inflation, about the same as on ten-year government bonds. Similarly, New York University’s endowment has reportedly languished because the trustees have been leery of stocks for most of the last two decades. It appears that the fund would be at least one third larger had they invested an additional $100 million in the stock market in 1982.7

The third thing we know about investing in the stock market is that it is possible, without taking bigger risks, to do better than the market averages through strategies such as “value investing,” the approach associated most closely with Warren Buffett and his teacher, the investor and former Columbia professor Benjamin Graham. Buffett says his huge success in picking stocks is owing to his ability to find companies whose shares are undervalued. Most economists, however, don’t believe that there can be such bargains in the stock market. To them, the stock market is “efficient”; that is, stock prices are set by a straightforward process in which self-interested choices by competing investors will almost automatically price shares correctly. The Nobel laureate economist Paul Samuelson explained to Congress thirty years ago that large numbers of highly motivated investors, constantly “selling those stocks they think will turn out to be overvalued and buying those they expect are now undervalued,” drive share prices to their true values. In this view, undervalued (or overvalued) shares simply cannot exist for more than a brief moment; if they did, investors would almost instantaneously rush to buy (or sell) them and their prices would rise (or fall) until they were valued correctly.

Buffett disagrees. He has described value investing as a “search for discrepancies between the value of a business and the price of small pieces of that business in the market.” There clearly would not be bargains of that kind if share prices reflected companies’ values. But at least as practiced by him and other like-minded investors, value investing has produced spectacular results. Ten thousand dollars invested with Buffett in 1956 would be worth about $200 million today, a return of about 27 percent a year compounded, more than double the annual return on the shares in the S&P 500 over the same period.

Buffett attributes his success to the fact that “market prices are frequently nonsensical.” Sometimes this is obvious: “The Washington Post Company in 1973 was selling for $80 million in the market. At the time, that day, you could have sold the assets to any one of ten buyers for not less than $400 million, probably appreciably more. The company owned the Post, News-week, plus several television stations in major markets.” Buffet perceived what most other investors did not. He invested about $10 million in the company, an investment that is worth approximately $930 million today.


In most cases, the relationship between price and value is less clear. In part this is because more easily measured assets such as plant and equipment have declined in importance, while less tangible assets such as brand names, technology, software, and the skills and commitment of the work force have become more important. But it is also because making even general judgments about the “intrinsic value” of a business is very difficult. Like many other investors, Buffett defines this as the value today of “the cash that can be taken out of the business during its remaining life.” Thus he is not concerned simply with a company’s growth or even with its annual earnings, but with how much money will be available for shareholders—over and above the amount the company must invest in the business—over a considerable period of years.8 Because making even rough estimates of these future cash flows is not easy to do, Buffett writes, “you do not cut it close. That is what Ben Graham meant by having a margin of safety. You don’t try and buy businesses worth $83 million for $80 million.”

Even so, mistakes are likely. Buffett, for example, underestimated the problems of USAir’s relatively high operating costs when he invested in the airline in 1989. As a result, its cash flows, and hence its value, were well below the amount he had expected and he essentially gave up on the investment. Although the airline and the investment have recovered spectacularly in the last few years, he considers his record in making the investment “unblemished by success. I was wrong in originally purchasing the stock, and I was wrong later, in repeatedly trying to unload our holdings at fifty cents on the dollar.”

Overall, he and his investors have prospered because he appreciated the lasting value of both strong franchises—e.g., the dominance of The Washington Post in Washington, D.C.—and good managers in companies such as the Washington Post Company, Coca-Cola, Gillette, and GEICO, and was willing to bet heavily on them when he could buy their shares at prices well below their “intrinsic values.” Unlike most economists and other investors, Buffett believes that a concentrated portfolio acquired at bargain prices is inherently less risky than a diversified one bought with little “margin of safety,” i.e., with unrealistically high prices relative to earnings or other measures of value that will badly endanger the stock’s price if the company does not meet these lofty expectations or if market conditions change. Maintaining a “margin of safety” is the standard Graham used to distinguish “investing” from “speculating.” Remarkably, as we shall see, even crude applications of the value approach outperform the market averages and are less risky.9

During the last fifteen years, research by academics and investors has contested the idea of automatic stock market efficiency favored by econ-omists such as Burton Malkiel of Princeton University, one of its leading popular exponents. Since 1973, when his book A Random Walk Down Wall Street was first published, he has argued that stock prices are so efficiently determined by the millions of individual decisions by investors—and thus are so unpredictable—that it is only a slight exaggeration to say that “a blindfolded chimpanzee throwing darts at the Wall Street Journal can select a portfolio that performs as well as those managed by the experts.”

Those opposing this view have conducted much research to try to show that there are patterns in stock prices suggesting that some trading strategies work much better than others. Some of the important research papers from the 1980s and early 1990s are collected in the volume edited by Richard Thaler, Advances in Behavioral Finance. Many of the successful strategies are based on, and confirm, the principles of value investing laid out in the 1930s by Benjamin Graham and David Dodd in Security Analysis, and carried forward by Buffett and other disciples. In What Works on Wall Street, James O’Shaughnessy, a financial writer and investment adviser, presents evidence on the returns and risks of more than fifty investment strategies over the forty-five-year period between 1951 and 1996. Among his findings are the following:

  • Between 1951 and 1996 you could have done four times as well as the S&P 500 by concentrating on large, well-known stocks with high dividend yields (i.e., with low prices relative to dividends).
  • Buying Wall Street’s current favorites with the highest price/earnings ratio is one of the worst things you can do.10

When investments are made by following rules such as these, the strategies sacrifice subtlety and nuanced judgment in favor of simplicity and system. But O’Shaughnessy and many others show statistically that so long as they are used systematically, even crude measures of value can identify stocks that will yield uncommon returns. For example, one important study showed that, between 1968 and 1990, portfolios of “value” stocks—defined simply as stocks whose prices are low relative to very rough measures of their worth such as book value, earnings, or cash flow—that were held for five years outperformed portfolios of “glamour” stocks by about eight to ten percent a year. “Glamour” stocks, as defined in this study and others like it, are polar opposites of value stocks. They are fashionable, have been bid up, and thus sell at prices that are high relative to variables like book value, earnings, or cash flow. (Portfolios of value stocks were defined, for example, as the 10 percent of all stocks in the sample with the lowest price/earnings ratios, while glamour stocks were the 10 percent with the highest price/earnings ratio.)11

The particular danger of glamour stocks is that they leave little margin for error when they don’t meet expectations. Oxford Health Plans, one of the fastest-growing HMOs, is a recent example. Oxford provides health care coverage to almost two million people, mostly in the Northeast, where it has a large roster of doctors and hospitals. On October 24, 1997, its shares were trading at about $70, roughly 250 times its earnings during the preceding twelve months, and more than ten times higher than typical price/earnings ratios at that time. Investors apparently thought that the rising demand for managed health care and Oxford’s attractive program would support the company’s growth and earnings. But the company’s price/earnings ratio, relatively high in any case, should have been cautionary. Managed care companies were at the time under pressure to improve service and not scrimp on medical care, which could well have meant lower profits. Oxford shares fell more than 60 percent during the market’s break the following Monday, October 27, when the company announced that its revenues and earnings would be lower than expected because it had wrongly estimated its costs and receipts. A share of Oxford stock is now less than $18.

Internet stocks like, the on-line bookseller, are currently very hot. The Amazon company went public last year at a price of $18 a share. It still has no earnings, yet its share price is about $82 (down from a high of $200), giving the company a total market value of approximately $2 billion. And although its revenues have grown sharply, from $511,000 in 1995, its first year of operations, to $147 million last year, its losses have risen from $303,000 to $27.6 million. Amazon, moreover, will be facing increased on-line competition from book chains such as Barnes and Noble and publishers such as Bertelsmann.

Amazon and Oxford are part of the Nasdaq, or over-the-counter market, in which apparently inflated valuations such as theirs are quite common. The Nasdaq market includes about 5,400 companies, including Microsoft, Intel, and Cisco Systems, as well as a great number of small, unproven ones. As a group, the Nasdaq companies have a P/E ratio of about seventy, almost three times that of the S&P 500. But suppose we exclude from that group Microsoft, Intel, and Cisco, which account for about 25 percent of the market value of all Nasdaq companies and for nearly 60 percent of their total earnings. These three stocks currently trade at P/E ratios of about fifty-five, eighteen, and fifty-two, respectively. Without them, the P/E ratio for the remaining companies jumps to more than ninety.12

The statistical studies of value investing suggest that just as popular stocks like Oxford frequently cannot live up to investors’ inflated hopes, out-of-favor stocks like the Washington Post Company in 1973 tend to rebound from excessively depressed levels. Equally important, value stocks seem to be no more risky than glamour stocks, whether risk is measured by the variability of their returns (as it most commonly is) or, more appropriately, by how well the returns hold up in bad times such as recessions or falling stock markets. This conclusion crucially undermines the conventional wisdom of economists. True believers in stock market efficiency tend, tautologically, to attribute the success of any strategies that beat the market to their presumed riskiness.

According to Peter Bernstein’s Against the Gods, risk became identified with the variability of investment returns in part because volatility can be measured relatively easily.13 But despite this advantage, volatility may not be the best measure of risk if shares are undervalued and if investors are willing to be patient. Warren Buffett, for example, prefers the dictionary definition of risk, “the possibility of loss or injury,” which is often unrelated to volatility. If, he points out, the stock of the Washington Post Company in 1973 had suddenly fallen even further, then its volatility would have been greater. By this measure, Buffett wrote, “the cheaper price would make it look riskier. This is truly Alice in Wonderland. I have never been able to figure out why it is riskier to buy $400 million worth of properties for $40 million than $80 million.”

Why do such opportunities for value investors continue to exist? Keynes hinted at the reasons in the General Theory. First, he observed, many investors do not look for long-term values, but try simply “‘to beat the gun,’…to outwit the crowd, and to pass the bad, or depreciating, half-crown to the other fellow,” much like “a game of Snap, of Old Maid, of Musical Chairs.” As a result, the stock market can be

likened to those newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those faces which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the other competitors, all of whom are looking at the problem from the same point of view.14

This way of thinking creates discrepancies between value and glamour stocks. Because it may also cause undervalued shares to become even more undervalued in the short term, value investors must be able to withstand losses.15 For example, throughout 1973 shares of the Washington Post Company sold in the stock market at a small fraction of the company’s value. But the shares were twice as cheap at the end of the year as at the beginning. Although Buffett acquired his stake in the company near the low end of the range, he still had unrealized losses that year and at times during the following year.

Second, because it involves betting against popular opinion, value investing requires greater conviction than simply following the crowd. If these factors, financial and social, limit the willingness of investors to take advantage of market prices that can be far out of line, they can also explain the persistence of stock market bargains. “An investor who proposes to ignore near-term market fluctuations needs greater resources for safety and…will in practice come in for most criticism,” Keynes wrote. “For it is in the essence of his behaviour that he should be eccentric, unconventional and rash in the eyes of average opinion.”

American experience during the last fifteen years makes it easy to forget that stock prices have not always gone up and that investing in the stock market is risky. During the 1970s, for example, investors in both stocks and bonds lost money after adjusting the values of their holdings for inflation. In the first volume of Securities Markets in the 1980s, Barrie Wigmore, a limited partner of Goldman Sachs, describes in detail the abrupt transformation in financial markets that took place in the early 1980s. Between 1979 and 1984, returns on the S&P 500 averaged almost 10 percent a year after inflation compared to losses of 1.5 percent annually during the preceding decade. At the same time, the markets for junk bonds, mergers and acquisitions, mortgage securities, and government bond futures began to grow dynamically. They have hardly paused.16

Unless we are now in a “new era” of permanent prosperity, a view that also was popular during the late 1920s and seems equally unlikely today,17 then the economy eventually will worsen and share prices will adjust, perhaps excessively, if only because hopes, and thus prices, are currently so high. For example, a recent survey of mutual fund investors found that they expected returns of 34 percent a year for the next ten years. That should be a warning sign.18 But because attempting to predict when the market will fall is such a poor gamble, and the rewards and risks of sound stock investments so attractive, investors who can afford it would be better off maintaining a relatively steady presence in the stock market. The extent of their stock investments should depend on how long they expect to hold their investments and on how much money they would be prepared to lose in the short term. J.P. Morgan suggested one way to estimate how much money to commit to stocks when he reputedly advised a friend who was losing sleep over his shareholdings, “Sell down to the sleeping point.”19

By this standard Americans seem, at least for the moment, to be sound sleepers. Over the last ten years they have increasingly invested in the stock market, largely through mutual funds, which now have assets of about $2.7 trillion and hold approximately 16 percent of all corporate shares compared to roughly 5 percent in the mid-1980s. Despite the success of Buffett and other value investors, mutual funds do not appear to have followed their lead. Neither the Investment Company Institute nor Lipper Analytical Services has a “value” category for their surveys of mutual funds because the label is used in such different ways by fund managers. Instead, the most popular mutual funds are “growth funds,” which tend to buy stocks with relatively high price/earnings ratios. Such funds account for more than 50 percent of all mutual fund assets.20

This helps to explain why even in these exceptional times, mutual funds have not matched market averages. For example, between 1984 and 1994 the average mutual fund returned about 12 percent a year compared to about 14 percent for the S&P 500. During that ten-year period, some 26 percent of all actively managed mutual funds beat the S&P, more than in any other ten-year span that ended in the 1990s. During the last ten years, however, only 14 percent did so. Because of results like these, and because those funds that beat the market averages do not do so consistently, “index” funds, whose portfolios simply mimic a broad stock market index such as the S&P 500, have better results than most active investment strategies.

The assets of funds based on the S&P 500 and other indexes have grown even more rapidly than mutual funds generally, increasing from about 1 percent of the assets of all equity funds in 1985 to 7 percent today; but they remain relatively small. The small size of index and value funds, combined with the popularity of growth funds, is yet another indication of the optimism of stock market investors and the attraction of fashionable investment strategies. If index and value funds were to account for a much more significant proportion of all shares, then the market would more closely resemble the economists’ efficient ideal—there would be far fewer undervalued or overvalued stocks. But the S&P 500 is after all merely an index: one can buy a fund that mimics it but, unlike most other Wall Street investments, the index cannot be promoted as benefiting from the expertise of stock analysts and money managers; and those who invest in it have no choice over the stocks that make it up. Passively investing in a market index is just not as exciting as investing in a growth fund or in the stocks that many are talking about.

“Worldly wisdom,” Keynes wrote, “teaches that it is better for reputation to fail conventionally than to succeed unconventionally.” For this reason, there will always be overlooked value stocks, and we will never know for certain if the market’s exuberance is rational or irrational. Experience suggests, moreover, that economic and stock market growth cannot be sustained indefinitely. If the unemployment rate continues to fall, for example, wages are likely to rise more rapidly, increasing inflationary pressures and making higher interest rates more probable. If this happens, (and it may not), the economy and the stock market can be expected to falter, and those who have chased fashion will suffer most.—May 26, 1998

This Issue

June 25, 1998