Sometimes, It’s Bonds For the Long Run
When measured in three-decade increments, bonds did better than stocks as recently as 2011
ILLUSTRATION: ALEX NABAUM
64 COMMENTS
Maybe investors should question the dogma of “stocks for the long run.” History shows that a portfolio of bonds has outperformed stocks surprisingly often and for shockingly long periods.
That’s the intriguing argument in a new research paper by Edward McQuarrie, a retired business professor at Santa Clara University. Investors have long taken it as an article of faith that stocks have always beaten bonds—and always will—if you can just hang on long enough. Prof. McQuarrie’s research is a healthy reminder that this belief is wrong. His findings also show the limits and dangers of extrapolating from the past.
Stocks offer a stake in a business’s variable profits in the indefinite future. Bonds are contracts conferring rights to a fixed stream of income over a certain period. If stocks didn’t offer the prospect of higher return, investors wouldn’t want to brave the uncertainty of owning them. But whether stocks deliver that higher return depends largely on how they are priced relative to bonds.
The popular belief that there’s never been a 30-year period in which stocks had lower returns than bonds is false. As recently as 2011, bonds had earned higher returns than stocks over the prior 30 years (long-term Treasury bonds, 10.7% annually; U.S. stocks, 10.4%).
That’s no aberration, says Prof. McQuarrie. Using digitized antique newspapers to supplement an online database of U.S. stock and bond prices, he assembled an index of bonds back to 1793.
That has enabled him to calculate 30-year returns beginning in 1823. Between then and 2013, he shows, bonds earned higher returns than stocks in one-quarter of all 191 three-decade-long periods.
Most of those stretches were in the 19th century. But much of the data on which Jeremy Siegel, a finance professor at the University of Pennsylvania’s Wharton School, relied for his best-selling 1994 book “Stocks for the Long Run” also came from the same era.
The difference: Prof. McQuarrie constructed his early data with a wide variety of bonds that would have been available to investors at the time. Prof. Siegel chose the highest-quality and lowest-yielding bonds available—often a single U.S. Treasury bond or as few as two municipal bonds.
According to Prof. McQuarrie, the issues tracked in “Stocks for the Long Run” account for less than 5% of the total bond market in much of the 19th century.
Prof. Siegel used so small a sample in order to approximate what economists call the “risk-free rate,” or the return on a bond with the lowest possible danger of defaulting.
In academic theory, that makes perfect sense. In the real world, the early U.S. had no risk-free rate. Not only was the survival of the nation often in doubt, but from 1835 through 1841 the U.S. Treasury didn’t even have any debt outstanding. By the 1840s eight states had defaulted. Most bonds were risky, so they often had to offer yields of 5% or more.
Seen through that wider lens, says Prof. McQuarrie, stocks don’t overwhelmingly dominate bonds. “Sometimes bonds give you a better return; sometimes, stocks do.” Calculations of bond returns based on only a sliver of the market are a “heroic extrapolation,” he says.
That the bonds in “Stocks for the Long Run” might have been “a tiny part of the market does not bother me,” says Prof. Siegel. He points out that three-month Treasury bills, often used as today’s risk-free rate, are also a small fraction of the market.
Could bonds as a whole—as opposed to a handful of high-quality issues—have done better than stocks in the early U.S.?
“That’s possible, and it might be in the data,” says Prof. Siegel. “Clearly, in that early period, with bonds that had higher yields, it could well be that the broad bond market may have outperformed stocks.”
Prof. McQuarrie calculates that bonds did slightly better than stocks—an average of 5.9% annually versus 5.8%, after inflation—all the way from the beginning of 1793 through the end of 1877.
To come out ahead of bonds back then, you would have had to hold stocks continually for more than 85 years—probably a tad longer than what most investors have in mind when they think of the phrase “stocks for the long run.”
Are these returns from the days of steamboats and stovepipe hats relevant today?
The pattern identified by Prof. McQuarrie has held in several countries in the modern era. In France, Italy, Japan and Spain, among other nations, bonds have earned better returns than stocks—after inflation—for decades on end in the post-1900 era.
“Is it likely that stocks will outperform bonds?” asks Prof. Siegel. “Of course. But should we never expect to find periods when bonds outperform stocks? No, no, no. We should expect to find that, absolutely.”
No one should ever assume that the outperformance of stocks over bonds, even over extremely long periods, is “predestined or foreordained,” he says. That’s especially true when interest rates start at high levels.
Many investors have put blind faith in stocks, confident that history will repeat itself. Someday it might—in a way that investors who have all their money in stocks should hedge against before it’s too late.
