AAII allocation signal looks less useful after a strong stock decade

A once-compelling stock-market signal has run into a problem retail investors know well: the last decade did not behave like the old charts said it should. An analysis from Of Dollars and Data found that high investor exposure to stocks, which historically pointed to weaker future returns, did not stop U.S. equities from posting strong gains from 2015 to 2025.

The signal comes from the American Association of Individual Investors, or AAII, which has tracked how individual investors split their portfolios among stocks, bonds and cash since 1987. According to the analysis, investors put an average of 62% of their portfolios in equities over that period.

That number is a rough demand gauge. If households already have a large share of their money in stocks, the thinking goes, there may be less room for additional buying to push prices higher. If stock allocations are low, future demand may be stronger, especially after a selloff.

The old relationship was simple

Of Dollars and Data said AAII’s equity-allocation series tended to fall during major market breaks, including the dot-com bust, the global financial crisis and the Covid-era crash. That makes sense mechanically: when stock prices drop, stocks become a smaller slice of a portfolio unless investors add enough money to offset the decline.

The more useful finding was about future returns. From 1987 through the period before the global financial crisis, higher average equity allocations generally lined up with lower annualized total real returns for the S&P 500 over the following 10 years. “Real return” means the return after adjusting for inflation, and “total return” includes dividends.

That relationship has weakened, according to the analysis. The key example is 2015: AAII equity allocations were high, but U.S. stocks still delivered strong real returns over the following decade through 2025.

A timing model caught two crises, then missed the rally

Of Dollars and Data previously tested a tactical model using the AAII allocation data. The rule started fully invested in an S&P 500 index fund, moved fully into five-year Treasuries when average equity allocation rose above 70%, and returned to stocks only after the allocation dropped below 50%.

In the backtest, that rule moved out of stocks in September 1996 and did not re-enter until October 2002, covering the dot-com bubble and crash. It also sold in May 2006 and bought again in November 2008, around the global financial crisis. The analysis said the model had no false positives across those two major episodes.

Then came the problem. The same model signaled an exit from stocks in January 2018. Of Dollars and Data said following that signal would have left an investor in bonds and missed a 196% rise in the S&P 500 since then.

The analysis points to a possible reason: the type of demand for stocks has changed. Automatic 401(k) contributions and passive index funds send money into equities regularly, often without regard to valuation, according to Of Dollars and Data. If more buying is automatic and less price-sensitive, high equity allocations may carry less warning power than they once did.

Anton Howes, writing about 14th-century England, described how labor markets found ways around wage controls after the Black Death through better non-cash compensation. Of Dollars and Data uses that history as a reminder that supply and demand can be powerful, but its conclusion is more cautious for modern stocks: the relationship may still matter, but recent evidence shows it is no longer a clean timing tool.

This story draws on original reporting from Of Dollars and Data.