The authors wish to thank John Campbell, Barry Eichengreen, Ken Froot, James Hamilton, Bruce Lehmann, Greg Mankiw, Jeff Miron, Peter Lindert, Robert Shiller, Andrei Shleifer, and Arnold Zellner for helpful discussions.Google Scholar

¹ Nominal stock index values are deflated by the producer price index. The index used is a smoothed monthly version of series 6 reported in the appendix to Shiller, Robert, Market Volatility (Cambridge, MA, 1989). The underlying nominal stock index prices are monthly average values from various issues of the Security Price Index Record, published by the Standard & Poor's Corporation. The S&P price index was begun in 1926; the Cowles Commission, however, extended the index before 1926 back to 1870.Google Scholar See Cowles, Alfred et al. , Common Stock Indices (2nd edn.New York, 1939). We focus on twentieth-century stock prices because the pre-1900 railroad-dominated market appears to us qualitatively different from the post-1900 industrials-dominated market. There is little reason to believe that price/dividend ratios and expected dividend growth rates should bear the same relation to each other as the market shifts from being two-thirds railroads in 1890 to being two-thirds industrials in 1910.Google ScholarFor the industrial composition of stock market indices, see Wilson, Jack and Jones, Charles, “A Comparison of Annual Common Stock Returns: 1871–1925 with 1926–1985,” Journal of Business, 60 (04. 1987), pp. 239–58.CrossRefGoogle Scholar For the rise of a market in industrial securities, see Davis, Lance, “Capital Immobilities and Finance Capitalism: A Study of Economic Evolution in the United States,” Explorations in Entrepreneurial History, I (Fall 1963), pp. 88–105Google Scholar; and Carosso, Vincent, The Morgans: Private International Bankers (Cambridge, MA, 1987).Google Scholar

² The real GNP series used is spliced together from the revised estimates of GNP in different periods made by Christina Romer and is discussed in De Long, J.Bradford and Summers, Lawrence H., “How Does Macroeconomic Policy Affect Output?” Brookings Papers on Economic Activity (Fall 1988), pp. 433–80.Google Scholar

³ Present values are calculated using a constant real discount rate of 6 percent. Figure 2 follows Shiller's, “Comovements in Prices and Comovements in Dividends,” in his Market Volatility. In Figure 2 it is assumed that the stock market's value in June 1989 accurately forecasts the present value of dividends to be paid in years after 1989.Google Scholar

⁴ See Galbraith, John Kenneth, The Great Crash (Cambridge, MA, 1954)Google Scholar; Sobel, Robert, The Big Board (New York, 1965)Google Scholar; Keynes, John Maynard, The General Theory of Employment, Interest and Money (London, 1936), especially chap. 12Google Scholar; and Shiller, Market Volatility, especially the chapters previously published as “Stock Prices and Social Dynamics,” Brookings Papers on Economic Activity, 1984 (Fall 1984), pp. 457–98Google Scholarand as “Do Stock Prices Move too Much to Be Justified by Subsequent Changes in Dividends?” American Economic Review, 71 (June 1981), pp. 42136.Google Scholar

⁵ See Sobel, Robert, N.Y.S.E. (New York, 1975), pp. 166–79.Google Scholar

⁶ A third view, popular among finance economists, is that stock prices always and everywhere equal the best available estimate of present fundamentals. The belief that the stock market was “the stage whereon is focused the world's most intelligent and best informed judgment of the values of … enterprises” and that no one has a better estimate of values than the stock market as a whole was popular in the 1920s until the crash of 1929. For example, see Lawrence, Joseph S., Wall Street and Washington (Princeton, 1929); Galbraith notes that after 1929 “Mr. Lawrence disappeared from Princeton. Among economists his voice was not heard again.” The efficient market hypothesis did not become popular again until the 1960s.Google Scholar

⁷ Campbell, John, “Estimating the Persistence of Expected Returns” (unpublished manuscript, London School of Economics, 1989) suggests that in the United States since 1926 fads and fashions have had an expected lifetime on the order of six months, not the five to ten years required if they are to account for major bull and bear swings.Google Scholar

⁸ Even if long swings in the stock market were primarily driven by fundamentals, sufficiently large short-run swings in stock prices could still lead to the conclusion that the economic performance of the stock market has been poor if marginal investors and firms follow short-run buy-and-sell rather than long-run buy-and-hold investment strategies. See De Long, J.Bradford et al. , “Noise Trader Risk in Financial Markets,” Journal of Political Economy, 98 (forthcoming 1990)CrossRefGoogle Scholar; and De Long, J.Bradford et al. , “The Size and Incidence of Losses from Noise Trading,” Journal of Finance, 44 (07 1989), pp. 681–96.CrossRefGoogle Scholar

⁹ Shiller, Robert, “Comment on Miller and on Kleidon,” in Hogarth, Robin and Reder, Melvin, eds., Rational Choice: The Contrast Between Economics and Psychology (Chicago, 1987), pp. 285–315.Google Scholar

¹⁰ See Williams, John Burr, The Theory of Investment Value (Cambridge, MA, 1937).Google Scholar

¹¹ Muth, John, in his essay “Optimal Properties of Exponentially Weighted Forecasts,” Journal of the American Statistical Association, 55 (03 1960), pp. 299–306, showed that the forecast rule given in equation 3 would be the best that could be done forecasting from past rates of dividend growth alone as long as the rate of dividend growth is subject to independent and serially uncorrelated permanent and transitory shocks that have Gaussian normal distributions.CrossRefGoogle Scholar

¹² Equation 2 is used instead of equation 1 so that equal percentage deviations of prices from fitted values are counted equally in determining the fit of the model.Google Scholar

¹³ On this point, see also Cutler, David et al. , “What Moves Stock Prices?” Journal of Portfolio Management, 15 (Spring 1989), pp. 4–12.CrossRefGoogle Scholar

¹⁴ Campbell, “Estimating the Persistence of Expected Returns,” estimates that news about future cash flows accounts for about only 30 percent of the variation in month-to-month stock returns.

¹⁵ The 1980s have also seen a swing in stock prices larger than can be accounted for in our model. We suspect that our model's failure to fit the Reagan bull market is due to the changing role of dividends. Before 1980 cash payments to shareholders besides dividends were trivial, since 1980 they have made up a substantial part of shareholders' income. Our dividend series does not include these payments. See Poterba, James, “Tax Policy and Corporate Saving,” Brookings Papers on Economic Activity (Fall 1987), pp. 455–516.CrossRefGoogle Scholar

¹⁶ Galbraith is more cautious. He sees the rise in stock prices through the end of 1927 as largely justified by fundamentals, as does Sirkin, Gerald, “The Stock Market of 1929 Revisited,” Business History Review, 49 (Summer 1975), pp. 233–41.CrossRefGoogle ScholarSeligman, Joel also agrees that “overspeculation” can only be said to have begun in the spring of 1928.Google ScholarSee Seligman, Joel, The Transformation of Wall Street (Boston, 1982)Google Scholar; and White, Eugene, “When the Ticker Ran Late” (unpublished manuscript, Rutgers University).Google Scholar

¹⁷ See Sobel, The Big BoardGoogle Scholar; Galbraith, The Great CrashGoogle Scholar; Dice, Charles, New Levels in the Stock Market (New York, 1929)Google Scholar; and Lawrence, Wall Street and Washington.Google Scholar

¹⁸ Fisher, Irving, The Stock Market Crash—and After (New York, 1930).Google Scholar

¹⁹ An additional cause of rapid stock market growth in the 1920s may have been a general recognition that, in a turbulent time like the interwar period in which the long-run drift of the price level was uncertain, stocks were no more risky than bonds and yet promised higher expected returns. See Smith, Edgar L., Common Stocks as Long Term Investments (New York, 1924).Google ScholarThe late 1920s also see the appearance of handbooks for investors that place great stress on valuing compound growth. For example, see Gould, Samuel, Stock Growth and Discount Tables (Boston, 1931).Google Scholar

²⁰ A few other voices, chosen more or less at random, reaching the same conclusion are Smith, Common Stocks as Long Term InvestmentsGoogle Scholar; von Strum, K. S., Investing in Purchasing Power (New York, 1925)Google Scholar; Ripley, William, Main Street and Wall Street (New York, 1927)Google Scholar; and Russell, Kimball, and Co., The Investment Problem—Its Solution (Boston, 1928).Google Scholar After 1929 the view that the post-World War I boom would continue for a long time was easy to critique: see, for example, Miller, A. T., What to Consider When Buying Securities Today (New York, 1932)Google Scholar; and Allen, Frederick, Only Yesterday (New York, 1931). But the view found few critics before 1929.Google Scholar

²¹ Woodruff, George P., Investment and Speculation (New York, 1939). The lesson Woodruff drew from the ex ante reasonableness of confidence in the “New Era” and the subsequent catastrophe of the Depression was that one could have little confidence in any fundamental-based valuation of prices, for “the rationalizing of a price level, whether it is a high one or a low one, is one of the most insidious lures … whenever the papers are full of a rationalizing or explaining of high prices, the investor may take it as one cue to begin selling.”Google Scholar

²² Galbraith, The Great Crash, pp. 20, 90Google Scholar; Babson, Roger, Business Barometers (Wellesley, 1930)Google Scholar; and Babson, Roger, Investment Fundamentals (New York, 1930).Google Scholar

²³ See Graham, Benjamin and Dodd, David, Security Analysis (New York, 1st edn., 1934; 2nd edn., 1941)Google Scholar; and Graham, Benjamin, The Intelligent Investor (New York, 1954).Google ScholarAlso see Warren Buffett's appendix to Benjamin Graham, The Intelligent Investor (5th edn., New York, 1987)Google Scholar; and Train, John, The Money Masters (2nd edn., New York, 1987).Google Scholar

²⁴ See Graham, The Intelligent InvestorGoogle Scholar; and Graham, Benjamin, Dodd, David, and Cottle, Sidney, Security Analysis (4th edn., New York, 1962).Google Scholar

²⁵ Graham, Benjamin and Dodd, David, with the assistance of Charles Tatham, Security Analysis (3rd edn., New York, 1951).Google Scholar

²⁶ Graham was actually on the prudent end of commentary on the market in the early 1960s. Others urged investors to pick “stocks for the surging '60s.” See Cobleigh, Ira, A Guide to Success in the Stock Market (New York, 1961).Google Scholar

²⁷ Bean, Louis, How to Predict the Stock Market (Washington, DC, 1962).Google Scholar

²⁸ Babson, Thomas and Babson, David, Investing for a Successful Future (New York, 1959).Google Scholar

²⁹ The most recent expression of this view is Herschel Grossman, “The Political Economy of War Debts and Inflation” (NBER Working Paper 2743, 1988).Google Scholar

³⁰ See Temin, Peter, Did Monetary Forces Cause the Great Depression? (New York, 1976)Google Scholar; Temin, Peter, Lessons of the Great Depression (Cambridge, MA, 1989)Google Scholar; and Allen, Only Yesterday.Google Scholar

³¹ Miller, What To Consider When Buying Securities Today.Google Scholar

³² Post-1929 U.S. economic growth does not show any noticeable slackening of pace.Google Scholar

³³ For example, Smith, Common Stocks as Long Term Investments.Google Scholar

³⁴ Drew, Garfield, New Methods for Profit in the Stock Market (Boston, 1948).Google Scholar

³⁵ See Hansen, Alvin, Economic Policy and Full Employment (New York, 1947)Google Scholar; and Haberler, Gottfried, Prosperity and Depression (New York, 1958).Google Scholar

³⁶ Tobias, Andrew, The Only Investment Guide You'll Ever Need (New York, 1978).Google Scholar

³⁷ See Heller, Walter, New Dimensions of Political Economy (Cambridge, MA, 1965)Google Scholar; and Tobin, James and Weidenbaum, Murray, eds., Two Revolutions in Economic Policy (Cambridge, MA, 1987).Google Scholar

³⁸ On technical analysis, see Goodman, George, “Adam Smith,” The Money Game (New York, 1968).Google Scholar

³⁹ Friedman, Milton and Schwartz, Anna J., A Monetary History of the United States, 1867–1960 (Princeton, 1963).Google Scholar

⁴⁰ Goodman, George, “Adam Smith,” Paper Money (New York, 1981). Goodman, however, also accurately forecast that if the doomsday scenarios failed to come true, the 1980s “could be quite a party.”Google Scholar

⁴¹ The data for Figure 6 are drawn from Summers, Robert and Heston, Alan, “Improved Comparisons of Real Product and Its Composition, 1950–1980,” Review of Income and Wealth, 20 (03. 1984), pp. 207–61.CrossRefGoogle Scholar