As measured by the S&P 500, large-cap domestic stocks have produced an annualized return of 9.8% since 1926, or 6.8% on a real basis (for the period ending 12/31/2011; Source: Ibbotson and Standard and Poor's Index: SBBI S&P 500 Index). Decent numbers to be sure, but the average hides a lot of history. When it comes to performance, indexes, like individual portfolios, are a collection of uneven events.
Breaking that 85-year stretch into just two periods, for example—from 1926 to 1981 when interest rates peaked, and from 1982 to 2011—we get very different outcomes. For the period 1926 until the beginning of 1982, the S&P produced a gain of 6.1% real terms. In the declining rate environment from 1982 to 2011, however, the real return for large-cap stocks jumped to 8.0%. Breaking down performance even further, say, by the decade, the numbers swing even wider.
Seeing the numbers raises an interesting question: How often do the returns by decade match the long-term average? Answer: Never. On a real basis, the only decade for the S&P that was even come close to the long-term average was the 1960s, and then there was a 150 basis point differential. Fact is, the performance of large-cap stocks by the decade bears little resemblance to the long-term average. The same is true if we look at small-cap stocks, long-term government bonds or long-term corporate bonds. Which raises another question: Does any of this matter?
Actually it does, and a lot. That's because many investment advisors and financial planners use historic long-term returns for their forward-looking return expectations. Using historic numbers in either the asset allocation process or arriving at wealth and spending forecasts introduces two major problems.
First, there is no fundamental reason why stocks should provide anything close to their long-term returns in any given decade. And yet investors undoubtedly view any 10-year period as "the long run." We would suggest that the misalignment of expectations and reality rarely leads to happy clients.
Second, the extrapolation of past returns–damaging enough in the realm of equities—can often be downright reckless for investment-grade fixed income, as is the case right now. At this point in the cycle, a 30-plus year of declining interest rates virtually ensures that the past won't be prologue. The average real returns produced by long-term government bonds since 1982—an astoundingly high 8.2% according to Ibbotson and Barclays Capital—will not be duplicated anytime soon. Nor will the nearly-as-astounding 7.7% produced by long-term corporate bonds.