What Rate of Return Can You Reasonably Expect by Peter Bernstein

Latest FAJ looks at retirement issues on its 70th anniversary.

Peter Bernstein took a look at stock and bond returns over a two hundred year period to try and figure out whether there was a "basic return" that could be expected for each asset class. He normalized his analysis to take account of starting and ending multiples and yields to ascertain what the return was over those periods.

His basic conclusion was that stock returns exhibited a mean nominal basic return of 9.6% with a standard deviation of 1.6% and some mean reversion elements.

With regard to bonds, the outcome was much less conclusive.  The mean nominal basic bond return came to 4.9% with a standard deviation of 2.3%, but there a wide band of uncertainty associated with the data and in fact he concluded we could have no confidence in expecting the nominal basic return because inflation would surprisingly pop up (investors were slow to respond to changing inflation environments) and there was no pattern to the returns (little credence can be attached to what anybody has to say about what the real what the real long term rate of interest should be, has been, or will be in the future over the long run). About the only thing we can say for bond holders is they can make con judgments about future returns beyond the duration of the particular instrument they happen to hold at any given moment.


Bottom-line: long run equity returns were more predictable than long run bond returns (with even greater application for real returns). With equities one can at least conjecture whether the market is high or low (based on recent returns), with the bond market there is no such opportunity.

He concludes that stocks are fundamentally less risky than bonds, not only because their returns have been consistently higher than those of bonds over the long run but also because less uncertainty surrounds the long term return investors can expect on the basis of past history.