Douglas Clement - Senior Writer
Published June 1, 2001 | June 2001 issue
This article and "The Stock Market: Too High? Too Low? Just Right." are based on work by Minneapolis Fed senior economist Ellen McGrattan, Fed adviser Edward Prescott, a Regents' Professor at the University of Minnesota, and Northwestern University economists Ravi Jagannathan and Anna Scherbina. Two of the papers appear in the fall 2000 Quarterly Review, published by the Minneapolis Fed's research department, and the third is currently a Minneapolis Fed research department working paper.
In The Declining U.S. Equity Premium, Northwestern University finance professor Ravi Jagannathan, Minneapolis Fed senior economist Ellen McGrattan and Northwestern instructor Anna Scherbina return to an issue that has vexed economists since the mid-1980s: the difference between returns to equity and returns to debt. Economic theory suggests that stocks should pay investors a bit more than bondsto reward them for enduring the higher risk of the stock marketbut that additional risk should give stock investors no more than a 1 percent premium. In fact, however, between 1926 and 1999, stocks returned nearly 7 percent more per year than bonds, according to the Center for Research in Security Prices. Why the extra 6 percent?
This conundrum, famously termed the "equity premium puzzle," was revealed in a seminal paper* in 1985 by Edward Prescott, a Minneapolis Fed adviser and Regents' professor at the University of Minnesota, and Rajnish Mehra of the University of California, Santa Barbara, and it quickly spawned a cottage industry of economic research that sought to verify, to quantify andmost elusivelyto explain the premium.
In their paper, Jagannathan, McGrattan and Scherbina find
that, in fact, the premium has declined dramatically in the last 30
years. As they define it, the premium is the difference between stock
yields and bond yields, where stock yields equal the sum of dividend
yields (per share dividend as a fraction of the share price) and expected
growth in dividends. (Mehra and Prescott computed the equity premium
by taking the difference between average stock returns and average bond
returns. Over long periods, average returns and average yields are roughly
the same.) The authors track measures of these yields over the last
70 years, looking at a variety of stock and bond portfolios (see Chart
For bonds, they concentrate on long-term (20-year) bonds, whose yields have varied from a low of 2.2 percent in the 1940s to a high of 10.4 percent in the 1980s, averaging 5.3 percent over the entire 1926-1999 period.
For stocks, the economists look at three different portfolios, ranging from the stocks in the Standard & Poor's 500 to a comprehensive list compiled by the Federal Reserve of all stocks of U.S. corporations. The yields of these stock portfolios differ from one another, and they also varied during the 1926-1999 time period. But the authors found that, over time, regardless of the stock portfolio looked at, the difference between stock and bond yields diminished significantly. The equity premium declined. For S&P stocks, for example, relative to long-term bonds, the average equity premium from 1926 to 1970 was 6.8 percentage points. In the three decades since then, the premium was just 0.7 percentage points. In 1999, in fact, the figure was slightly negativean equity concession, if you will (see Chart 2).
To see if their findings are robust, the authors tinker with measures of both bond and stock yields, using a variety of assumptions about dividend growth rates and bond returns. Their conclusion stands: The equity premium is gone, or nearly so. And thatsurprisingly or notis consistent with the theory that Mehra and Prescott used in their 1985 paper.
But what, then, accounts for the change? Why has the premium, so high in the past, apparently vanished? The authors don't try to solve this quandary. They note that their findings parallel recent estimates by other economists, and they theorize that the answers may lie in the declining costs of portfolio diversification and improved investor access to informationin essence, market imperfections have diminished.
But they also point out that not all analysts concede the death of the premium. While most recent economic analyses concur that the U.S. premium is lower than it has been historically, this finding is "in sharp contrast" to the view of numerous finance professors. A recent survey of 226 such professors found their equity premium forecasts to be quite high: an average one-year horizon forecast of 5.8 percentage points and an average five-year forecast of 6.7 points. "Apparently," note the authors, "finance professors do not expect the equity premium to shrink."
Reconciling the divergent views may be impossible, but Jagannathan, McGrattan and Scherbina suggest that their findings be taken seriously. Investors who rely on historical evidence to conclude that stocks will continue to pay better than bonds, they caution, "are likely to be disappointed."
* Mehra, Rajnish, and Prescott, Edward C. 1985. The equity premium: A puzzle. Journal of Monetary Economics 15 (March): 145-61.