In 1999, James Glassman and Kevin Hassett created a sensation with their book Dow 36,000: The New Strategy for Profiting from the Coming Rise in the Stock Market. The authors, who viewed stocks as undervalued, predicted the Dow Jones Industrial Average would quadruple in three to five years. Reviewing it at the time, I wrote, “The only thing missing from this halftime speech of an investment book is the exhortation to buy stocks for the Gipper.”

Their headline assertion, of course, never came to pass; 1999 was the peak of the biggest bubble in U.S. stock market history. Stocks fell 44% in real terms over the next few years, then rose to almost 1999 real levels in 2007 before falling 50% in the financial crisis. The Dow has only recently hit 29,000, 25% short of 36,000 after 20 years.

Although the book became a symbol of late 1990s mania,1 its primary message—stocks return more than bonds and are less risky over long periods—is worth exploring. From the perspective of 20 years, it doesn’t look as silly as it did at the market bottoms in 2002 and 2009, but it still doesn’t fully make its case.

Consider this hypothetical: Suppose at the end of 1999 an eccentric uncle gave a $36,000 portfolio of government bonds2 to his nephew Fred and $11,497—the level of the Dow at the time—of stocks to his niece Ginger. If the true value of the Dow were $36,000, as Glassman and Hassett essentially asserted, these should have been equal gifts.

Over the subsequent 20 years, Fred received $26,000 of interest while Ginger received only $5,000 in dividends. But Ginger also received the indirect benefit of $7,250 of retained earnings, which the market thinks were reinvested to create an additional $7,250 of value. Given that interest income is fully taxable while dividends and capital gains have favored tax rates for some investors, Fred’s $26,000 might not be much different from Ginger’s $19,500 after taxes.

If Dow 36,000 is viewed as a prediction of investor opinions, it was decisively wrong. But if you take it as an assertion of long-term economic value, it may have been in the ballpark: $5,000 of Ginger’s return is hard cash; $7,250 is accountants’ opinions of her economic gain; and $7,250 is investor opinion about the returns earned on reinvested equity. That’s the least solid part of the estimate, but even if the market is off by 50% in either direction on that $7,250, that $36,000 in government bonds and $11,497 of stock might have had roughly similar after-tax results over 20 years.3

Cash-to-cash comparison of $36,000 in government bonds in 1999 and $11,497 of stock depends on the choice of securities, the reinvestment of cash flows, taxes and 2020 stock market valuations. Cash-flow analysis makes Dow 36,000 seem optimistic but not crazy.

Nevertheless, the core thesis of Dow 36,000 is wrong for three reasons. First, there are almost no pure long-term investors. Nearly everyone feels the pain when their portfolios crash 50% or more, and most people—including at least one author of Dow 36,000—sell at the wrong time as a result. Nearly everyone needs money in the short or medium term sometimes. Even if you are 100% sure stocks will provide more total after-tax real cash flow over 20 years than bonds, there is psychological and financial value to portfolio stability.

Second, investors can’t be 100% assured that stocks will provide more real return than bonds, even over very long periods. That idea arose studying the U.S. in the 20th century, an exceptional one for stocks. (Even then, government bonds outperformed stocks over some periods, such as 1901 to 1932.) And although data from the 19th and 21st centuries, and non-U.S. stocks markets (including those markets wiped out by World War I, World War II or revolution), suggest that stocks are probably the best bet for the long run, there is no guarantee.


Finally, while first-generation quantitative valuation models tended to assume investors demanded high expected returns on stocks to compensate for their volatility, subsequent research suggests that the return premium is mainly for the exceptionally bad performance that stocks give at the worst financial times. Because this seems likely to be true for stocks in the future, there seems little reason for investors to change their valuation principles over the next few years, or even the next few decades. The fundamental economic value of stocks based on cash flows may be triple their market prices, but investors would have to hold them over decades through thick and thin to realize that economic value. They won’t collect by selling to other investors anytime soon.

Dow 36,000 is not as comically wrong as it looked in 2002 or 2009. It makes one important point: Even buying at the all-time peak valuations in December 1999, the increasing stream of cash flow from holding stocks pays off over a long-enough period. Low-cost, well-diversified index funds of equities remain the main tool for most individual investors to earn their way to financial security. But the book misses one equally important point: Equity risk is real.

  1. Glassman reversed himself and wrote “Safety Net,” advising investors to sell their stocks, move their money outside the U.S., buy bonds and short the market. Timing was against him again as the stock market proceeded to stage the longest bull market in history.
  2. Equally weighted among six-month, one-year, three-year, five-year, seven-year, 10-year and 20-year securities; with maturing principal reinvested at the same maturity
  3. This took place over 20 years in which stocks underperformed history (4.7% average annual real return compared with 8.5% for 1871 – 1999) and bonds outperformed (3.3% compared with2.9%). In an average 20 years for the two asset classes, Ginger might be significantly better off than Fred despite getting a gift worth less than a third as much in 1999.

This article was provided by Bloomberg News.