One year ago, with the Dow Jones industrial average at 8,700, financial writer James Glassman and economist Kevin Hassett wrote in the Wall Street Journal that the price of stocks was still way too low. By their logic, they said, the Dow should be at 35,000--and would be, when investors caught on.
This week, as the Dow hit 10,000, Glassman and Hassett repeat their argument in the Journal. They admit that the "financial establishment" has reacted with guffaws. They do not mention that Jeremy Siegel, the Wharton finance professor whose research is central to their case, wrote to the Journal that "their analysis contains a serious flaw [and] vastly overstates the value of stocks." Slate also published a "Dialogue" between Glassman and the Economist's Clive Crook, in which Crook explained that Glassman was "wrong, plain wrong ... in the same way that it's wrong to say two plus two equals five." But Glassman and Hassett (henceforth, "Glassman") remain unbowed. In fact, they've upped the ante. This fall they're bringing out a book titled Dow 36,000.
It would make a wonderful tale if these two (both associated with the American Enterprise Institute, a conservative Washington think tank) had discovered what the entire finance establishment had missed, and written the next Wealth of Nations. The true tale, sadly, is somewhat less inspiring. Glassman's analysis suffers from a plain conceptual mistake, described below, which he simply refuses to admit, no matter how many times it is pointed out to him. Here is his argument, and why it is wrong.
Glassman begins by asking whether you'd rather have $100 in a savings account that pays 6 percent interest or a $100 share of a company that earned $1 per share last year. The answer is simple, right? The savings account pays you $6 per year, while the stock pays only $1--and even that only if the company's entire earnings are paid out in dividends.
But why would a stock that pays out only $1 be trading at $100, when investors can get $6 in a savings account? Because firms tend to grow over time and so do their earnings. Firms can grow for any number of reasons, but only one is mathematically certain, and that is the one Glassman's argument depends on: A firm that earns $1 can use it to expand, and thus increase its earnings, expand even more, and so on. For instance, a typical firm with $1 of earnings per share might earn $1.05 next year, $1.11 the following year, and so on. In 35 years, earnings per share will pass $6 and continue to increase. You can show mathematically that $6 a year forever and $1 this year plus $1.05 next year and so on are equal amounts. So it's not illogical that investors would be willing to pay the same $100 for either income stream.
Glassman points out that investors are not paying the same amount for these two income streams. To get $6 a year from a savings account costs you about $100 today, but you can buy $1 of corporate earnings on the stock market for an average of only $25! One reason is that stocks are considered riskier. Even if the average return on the stocks is the same as the certain return on the savings account, people will pay extra for the certainty itself. Glassman thinks this is the only reason for the difference--although it's not.
But Glassman argues--citing historical studies by Siegel--that stocks held for many years are not actually riskier. And when people realize this, they'll be willing to pay $100 instead of $25 for $1 of corporate earnings, and stock prices will quadruple. That's where he gets his 36,000 Dow: four times the current level. (Maybe it's not too late to call the book Dow 40,000.)
So where's the flaw? Assume that Glassman is right about the long-term risk of stocks, and assume he's right that average corporate earnings will grow at 5 percent a year. The problem is that he is double-counting. The $6 a year you get from a savings account is yours to spend on anything you please. The corporate earnings are yours to spend only if they are paid out in dividends. But if they are paid out in dividends, they aren't available to expand the firm, and so the delightful progression from $1 to $1.05 to $1.11 and so on won't occur.
Another way to see the flaw is to apply Glassman's logic to the savings account. Sure, a $100 investment today will only get you $6 this year. But those earnings will allow the savings account to grow, and next year's earnings will be $6.36, and then $6.74, and so on. So according to Glassman's theory, you should be willing to pay $400 for a $100 savings account.
The key point is that earnings cannot be simultaneously paid out in dividends and invested in future profits. Glassman would be right, however, if you could buy $1 of dividends for $25. But it turns out that to buy $1 of dividends costs you $72 (among Dow Jones industrial average stocks). Which suggests that if you really believe that stocks are actually no riskier than a savings account--or, rather, if you believe that everyone else will come to believe this--the Dow may still have 3,888 or so points to go. Dow 14,000 anyone?