The stock market’s best forecasting tool says returns over the next 10 years will be dismal
The “Single Greatest Predictor of Future Stock Market Returns” has never been more bearish.
I’m referring to the average U.S. household’s portfolio allocation to stocks. It is an inverse indicator, with higher equity allocations correlated with lower returns over the subsequent 10 years, and vice versa. Credit for discovering the indicator goes to the anonymous author of the Philosophical Economics blog.
Most Read from MarketWatch
The below chart plots this indicator since 1951, which is how far back data for it are available. As you can see, its correlation with the S&P 500’s SPX subsequent 10-year real total return is impressive. The r-squared of the correlation — a measure of its explanatory and predictive power — is a highly statistically significant 58%.
I know of no other valuation indicator that comes close to having as much explanatory power. The “Cyclically-Adjusted Price Earnings” ratio (or CAPE), the indicator made famous by Yale University finance professor and Nobel laureate Robert Shiller, has an r-squared of 30%, which — though still statistically quite significant — is barely half as much as that of the “Single Greatest Predictor.” The “Buffett Indicator” (the ratio of the stock market’s total market cap to GDP, which Warren Buffett once declared is “probably the best single measure of where [stock market] valuations stand at any given time”) has an r-squared of 28%.
The average household’s latest equity allocation reading is 56.4%, an all-time high. Its average since 1951 is 36.4% — 20 percentage points lower. Based on the historical relationship between the “Single Greatest Predictor” and the stock market’s subsequent 10-year return, this latest reading implies that the S&P 500 will produce an inflation-adjusted total return through 2036 of minus-6.0% annualized. Assuming inflation over the next 10 years is 2.7% annualized (which is the latest from the Cleveland Federal Reserve’s Inflation Expectations model), the average household’s equity allocation implies a nominal return of minus-8.7% annualized.
Some object to this bearish projection, arguing that the average household’s equity allocation has grown since the 2008-’09 financial crisis for reasons having little to do with the stock market’s future potential. That’s because, in the wake of real estate’s devastating losses during the financial crisis, investors began allocating more of their assets to equities. To that extent, some therefore believe, recent years’ higher indicator readings don’t have the bearish significance they otherwise would have.
This argument cuts both ways, however. Higher equity allocations, regardless of the sources of the increase, mean that households are that much more vulnerable to any hiccups in the stock market. This is because of the so-called “wealth effect,” which supports economic growth when the stock market is rising. But the wealth effect also means that when the stock market falls, households with greater equity exposures will feel that much poorer and consequently cut back on their spending — furthermore compounding the downturn.
The usual qualifications apply. Correlation is not causation, and no indicator — no matter how good — carries a money-back guarantee. Still, given the statistical record of the “Single Greatest Predictor,” its current all-time-high reading should give us all pause.
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