Predictions are difficult, especially about the future.
I was reminded of this famous quotation by a recent study of Wall Street research firms’ track records when projecting the longer-term returns of the various asset classes. The quote is from the Danish physicist Niels Bohr.
Those firms get lots of press attention whenever they update their projections, and they are fodder for much discussion. One such firm is GMO, the Boston-based investment firm.
According to its latest projections, for example, U.S. large-cap stocks are projected to produce a 2.2% annualized loss in inflation-adjusted terms over the next several years. Emerging market value stocks, in contrast, are projected to beat inflation by 8.5% annualized, while U.S. bonds’ inflation-adjusted return is projected to be minus 2.4% annualized. And so on.
I mention GMO because it famously has been way too bearish on U.S. equities over the past decade. And the firm has received its share of criticism and ridicule, though this year’s bear market has softened some of that criticism.
But, according to a just-released study, GMO is not alone in producing projections that don’t come close.
Titled “How Accurate are Capital Market Assumptions, and How Should We Use Them?,” the study was conducted by Mike Sebastian, an investment consultant who previously was chief investment officer at Aon (the British-American financial-services firm), and at NextCapital (the fintech firm that was acquired by Goldman Sachs in August).
Sebastian reached his conclusion by compiling the projected asset class returns from 17 firms in 2012, and then comparing those projections to what actually happened over the subsequent decade. The results are summarized in the accompanying chart, below.
Notice that for 14 of the 15 asset classes Sebastian analyzed, the class’ actual 10-year return was outside the minimum-maximum range of the firms in his sample. In other words, the asset class either produced a return that exceeded even the most optimistic firm’s 2012 projection, or was worse than that of the most pessimistic firm.
The only asset class for which the actual 10-year return fell within the minimum-maximum projected range (and even then only barely) was non-U.S. developed-country stocks, whose actual 10-year annualized return of 6.2% was slightly better than the most pessimistic projection in 2012 of 6.0% annualized.
It would be easy to conclude that those results are devastating, and at a minimum they are sobering. In an interview, however, Sebastian emphasized that capital market assumptions are not worthless. But his results show that they should not be taken as gospel.
One worthwhile approach he thinks would be appropriate would be to rarely deviate from your default asset allocation (your target weights for each of the asset classes) and not deviate very much when you do.
And on those rare occasions when you do deviate, you should do so only because you have a strongly held belief that the Wall Street consensus is not just a little bit off-base but hugely wrong. Low-conviction hunches are not reason enough. The average projections of various research firms, along with their minimum-maximum ranges, would help us know what that consensus is.
The lesson here isn’t that you need to find a better research firm than the prominent ones in Sebastian’s sample. Those firms’ analysts are some of the best and brightest people you’ll ever find in the investment arena, sporting impeccable educational and real-world backgrounds. And yet they still sometimes get it wrong, and not just by a little.
Humility is a virtue.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at firstname.lastname@example.org.