From the AAII Sentiment Survey: Cash Allocations is at 14.19%. Investors are holding the least cash in 4 years. This happened 3 times in the past 20 years: late-2017, Jan 2020, late-2021. Each case preceded significant market volatility and losses for $SPX https://t.co/T7G2LJTuJj
Source: Reuters
Research Brief
What our analysis found
The AAII Asset Allocation Survey for February 2026 shows that retail investors' cash holdings have fallen to just 14.2% of portfolios, a figure that sits well below the long-term average of 22.5% and marks the lowest reading since November 2021. Stock allocations, meanwhile, climbed to 69.4%, with bonds at 16.4%, painting a picture of aggressive risk appetite at a time when the S&P 500 breached 7,000 intraday for the first time on January 28, 2026. The data, confirmed by both AAII's own reporting and YCharts (which prints the precise figure at 14.19% for February 28), underscores a multi-year trend of shrinking cash buffers among self-directed investors.
A widely shared tweet claims that similar low-cash episodes occurred just three times in the past two decades — late 2017, January 2020, and late 2021 — and that each preceded significant market losses. The historical record does lend weight to this pattern: the late-2017 low was followed by the "Volmageddon" selloff of February 2018, the January 2020 low preceded the COVID crash that saw the S&P 500 plunge roughly 34%, and the November 2021 low came just before the 2022 bear market's 25.4% peak-to-trough decline. However, analysts caution that the tweet mislabels its source — calling it the "AAII Sentiment Survey" when the data actually comes from the distinct monthly Asset Allocation Survey — and that broader statistical analysis shows low-cash readings are not a reliable standalone sell signal.
Commentary from early March 2026 places the current cash allocation near the 4th percentile of all readings since 1987, framing it as a contrarian warning. Yet research from Schaeffer's Research found that when AAII cash sits in its lowest 10% of historical readings, the S&P 500 still averaged an 8.1% gain over the next 12 months with roughly 70% positive outcomes — only modestly below the typical 9.6% average return. The debate highlights the tension between anecdotal pattern recognition and rigorous backtesting in market forecasting.
Fact Check
Evidence from both sides
Supporting Evidence
February 2026 cash at a four-year low is confirmed
: AAII's own February 2026 report and YCharts both show cash at 14.2% (14.19% precisely), the lowest since a 14.2% reading in November 2021, corroborating the tweet's headline claim.
Late-2017 low cash preceded the February 2018 selloff
: Forbes reported November 2017 AAII cash at the lowest level since 1999; by January 2018 it was 13.3%. The S&P 500 then suffered a roughly 10–12% drop in early February 2018, and the BIS documented the VIX spike and XIV termination on February 5, 2018.
January 2020 low cash preceded the COVID crash
: AAII flagged a two-year low in cash exposure in January 2020. The S&P 500 subsequently fell approximately 34% from its February 19, 2020 high to the March 23, 2020 low.
November 2021 low cash preceded the 2022 bear market
: Cash hit 14.2% in AAII's November 2021 survey. The S&P 500 peaked on January 3, 2022 and declined roughly 25.4% to its October 12, 2022 trough.
Current reading sits at an extreme historical percentile
: Early March 2026 commentary places AAII cash near the 4th percentile of all readings since 1987, reinforcing the rarity and potential contrarian significance of the signal.
Contradicting Evidence
The tweet misidentifies the data source
: The 14.2% figure comes from AAII's monthly Asset Allocation Survey, not the weekly bull-bear Sentiment Survey. Multiple sources, including AAII itself, explicitly distinguish the two, meaning the tweet's framing is technically inaccurate.
Statistical backtesting shows low cash is not a reliable sell signal
: Schaeffer's Research found that when AAII cash is in its lowest 10% of readings since 1990, the S&P 500's next 12-month return averaged 8.1% with about 70% positive outcomes — only slightly below the typical 9.6% average and 80% hit rate. This suggests modest underperformance, not a consistent precursor to large losses.
Mechanical portfolio effects can distort the cash percentage
: Rising equity prices automatically increase the stock share of a portfolio and decrease the cash share without the investor actively reducing cash holdings. Schaeffer's highlighted this dynamic, meaning a low cash reading may reflect market appreciation rather than a deliberate decision to deploy reserves.
Small and non-representative sample size
: The AAII allocation survey typically polls roughly 600 self-selected members, according to academic reviews. This is not a comprehensive sample of all U.S. retail investors, limiting how broadly the findings can be generalized.
Three historical precedents is an extremely small sample
: Drawing a causal or predictive conclusion from only three prior episodes over 20 years does not meet a rigorous statistical threshold. The cited drawdowns (COVID pandemic, rate-hike cycle, volatility blowup) each had distinct macroeconomic catalysts unrelated to cash allocation levels.
Report an Issue
Found something wrong with this article? Let us know and we'll look into it.