Stocks Vs. Bonds -- A Radical New Look at an Old Question

Stocks always beat bonds over the long run? A report questions that long-held assumption.
Author:
Publish date:

Sometimes, you hear the most radical things in the most conventional of places.

That happened to me today when I heard a provocative analysis of the stock market at

The Plaza Hotel,

of all places. The occasion was a well-attended forum for institutional investors presented by Robert Arnott, managing partner at

First Quadrant

, an investment firm, and by Ronald Ryan, president of

Ryan Labs

, which advises pension funds. The forum was essentially an extended discussion of a research paper by Arnott and Ryan called, "The Death of the Risk Premium: Consequences of the 1990s".

The report asks an important and disturbing question for all investors: Will the return on stocks beat the return on bonds over the next 10 to 20 years? This may seem an alarming question to pose given that, in our lifetime, stocks always have beaten bonds. But, in light of this market's record high valuations and the relatively high real yield for bonds, it's surprising that more folks on Wall Street have not asked this question out loud. After all, if you are a stock investor -- and who isn't -- then the answer to this question may determine a good deal of your investment returns over the next two decades.

When Arnott and Ryan talk about "risk premium," they simply mean the superior return stocks must offer investors in order to attract money that would otherwise flow into bonds. The notion is that stocks must offer more because they are more volatile than bonds and because the stock investor is less assured of getting his or her money back than is the bond investor. "Stocks have outpaced bonds by about 5% per annum for a 74-year span and have produced real returns

i.e., adjusted for inflation north of 7% for an entire century."

"Is the party over?" they ask.

"It would be foolish to say that markets can go no higher," they write. "Of course, they can go higher. However, there is a tradeoff. The higher the markets go, without underlying fundamentals keeping pace, the lower the

future

rates of return must fall. This is a simple truism that has some rather alarming implications. Few would reject the notion that future real returns on stocks cannot, from current market levels, match the past. Interestingly, we

can

put a number on it. One path to estimating future returns is to examine the past."

They then proceed to dissect the 8.4% average annual real return on stocks since 1925.

They figure that 5.4 percentage points of the 8.4% real return come from the dividend yield in the period. Real dividend growth accounted for another 1 percentage point of the return on stocks. And they calculate that the final 2 percentage points of the total 8.4% has come "as a direct consequence of dividend yields falling to their lowest levels in U.S. history and P/E multiples

very recently at their highest levels in modern U.S. history. ... In 1925, investors paid 18 years' worth of current dividends to buy stocks; today's investors willingly pay 80 years worth of current dividends to buy stocks, more than quadruple the 1925 levels."

They then subtract from the 8.4% return the 3.7% bond yield that prevailed in 1925. Voila. They get the 5.1% long-term outperformance of stocks over bonds.

To estimate whether stocks today are priced to yield a comparable risk premium -- i.e., superior return -- over bonds in the future, Arnott and Ryan perform a similar calculation using contemporary figures. There are a few twists in the arithmetic, though.

First, they estimate that the real dividend growth rate will be 2% per year going forward. That seems extremely generous to me given that, historically, the dividend growth rate has been closer to 1%. Second, they assume no substantial increase in P/E ratios, which also seems sensible given that the broad market's P/E remains close to its all-time high. After all, how much more P/E expansion can an investor reasonably expect?

Bottom line, they estimate a 3.2% average annual real return on stocks going forward. They then compared that 3.2% to the riskless 4.1% yield on U.S. government inflation-adjusted bonds. In short, stocks could

underperform

bonds in the decades ahead by about 0.9% a year, according to Arnott and Ryan.

That doesn't necessarily mean that stocks won't offer a positive real return to investors, of course. And, the risk premium of stocks over bonds may simply shrink and not disappear entirely, as Arnott and Ryan suggest.

But their analysis does raise questions about the "perma-bull" or "New-Era" thesis that stocks

always

beat bonds and that you

must

overweight stocks at all times to achieve a superior long-term return on your capital.

Tomorrow, I'll present the alternate investing strategies and lessons that Arnott and Ryan derive from this "risk premium" analysis.