Verify any claim · lenz.io
Claim analyzed
Finance“Historical rolling-return data for the S&P 500 Index show negative returns in 46% of one-day periods, 38% of one-month periods, 25% of one-year periods, 16% of three-year periods, 10% of five-year periods, 6% of 10-year periods, 2% of 15-year periods, and 0% of 20-year periods.”
The conclusion
Open in workbench →The historical pattern is well supported: negative S&P 500 rolling returns became substantially less frequent as holding periods lengthened, and most listed percentages are close to published estimates. However, results vary with the sample period and return methodology, and several sources report 0%, rather than 2%, for 15-year periods. The claim is therefore broadly accurate but overly precise.
Caveats
- The 15-year figure is disputed; several cited analyses report 0% negative periods rather than 2%.
- Percentages depend on the date range, use of price or total returns, sampling frequency, and treatment of overlapping windows.
- Historical frequencies describe past observations and do not guarantee nonnegative long-term returns in the future.
Fact-check inside the tools you already use
Connect Lenz to ChatGPT, Claude, or WhatsApp and check a claim mid-conversation.
Or create a free account to get notified if new evidence updates this analysis.
Sources
Sources used in the analysis
For example, the S&P 500 Index return was negative in about 24% of overlapping one-year periods from January 1926 through March 2025. … But the frequency of negative returns decreases as the investment period expands, and no stretch over 184 months has been negative.
Historically, about 46% of individual trading days had a down close (and 54% were up). … Even extending to one month, the chance of a negative return is still high (Crews Bank reports about 38% for 1-month, 32% for 3 months). … Over one-year spans, about 25–27% of all years were negative. Five-year and ten-year probabilities are about 10–11% and 5–6% respectively. If we extend to even longer horizons (not shown), the chance of loss keeps dropping; in fact, historically no 20-year period has been negative.
Some 46% of the total number of spells lasted just one month. About 24% lasted two months, 17% lasted three months and 7% lasted four months. The remaining 6% lasted five months or more.
| Daily Returns | | | 12,343 | | 10,916 | | | | 23,259 | | | | | 53.1% | | 46.9% | | | | … | Calendar Month Returns | | | 693 | | 411 | | | | 1,104 | | | | | 62.8% | | 37.2% | | | | … | Rolling 1 Year Returns, Monthly | | | 816 | | 277 | | | | 1,093 | | | | | 74.7% | | 25.3% | | | | | Rolling 5 Year Returns, Monthly | | | 914 | | 130 | | | | 1,044 | | | | | 87.5% | | 12.5% | | | | | Rolling 10 Year Returns, Monthly | | | 926 | | | 58 | | | | 984 | | | | 94.1% | | 5.9% | | | | | Rolling 20 Year Returns, Monthly | | | 864 | | | 0 | | | | 864 | | | 100.0% | | | 0.0% | | | |
If an investor were to be invested for only a single day, there is a 45% chance of experiencing a negative return, while if an investor was in the market for 10 years, there is only a 12% chance of experiencing a negative S&P 500 return.
··· All available data (Jan 1793 - Jul 2026) … Positive Periods (%) | | | | | 73.7 | 85.9 | 90.3 | 97.4 | 100.0 | 100.0 | |
The chart above looks at the one-, thee-, five-, ten-, fifteen-, and twenty-year rolling index returns of the S&P 500 Index over the time period of January 1973 through December 2016.
• The average return for 15 years? At best, +18.89%, at worst, +6.46%. And it was positive 100% of the time.
| Holding Period | Avg Annual | Best Case | Worst Case | % Positive | | --- | --- | --- | --- | --- | | 1 Year | 12.1% | 54.0% | -43.1% | 73.0% | | 3 Years | 11.4% | 31.0% | -27.0% | 84.0% | | 5 Years | 10.8% | 28.6% | -12.5% | 88.0% | | 10 Years | 10.5% | 20.0% | -3.1% | 94.0% | | 15 Years | 10.3% | 18.8% | 0.7% | 100.0% | | 20 Years | 10.2% | 17.8% | 3.0% | 100.0% |
Of the 978 total periods, 910 (93.0%) have generated positive returns while 68 (7.0%) produced negative results.
Below is a static reference table of S&P 500 rolling returns across the canonical holding periods. Both price return and total return (dividends reinvested) versions are shown. The numbers come from monthly-average S&P 500 closes going back to 1871.
Our study of 94 5-year periods on the S&P 500 shows that most had positive returns, with only six negative ones.
Today's chart is from Ben Carlson’s“A Wealth of Common Sense” which shows the S&P 500’s rolling returns for 3, 10, 20, and 30 year periods going all the way back to 1926. … In fact, there are only three periods over the last 94 years where the 10 year return was negative and the annual average return over that period was 9.2%.
What these charts show you is that the longer you extend the time frame, the less time is spent in negative territory. Over the 20-year time frame, we haven't been negative since the late 1940s.
S&P 500 Total Return Index monthly from 1926 w/growth trendline and return charts for 1 year, 2 years, 3 years, 5 years, 10 years and 20 years.
The chart below shows the annual rolling returns for S&P 500 for 3 (blue line), 5 (red line), and 10 years (yellow line): … The 3-year rolling returns are frequently showing negative returns over a 3-year period, while the yellow line shows that the ten-year annual rolling returns only once were negative – in the middle of the financial crisis in 2008/09.
··· All available data (Jan 1793 - Jul 2026) | | Best Rolling Return (%) - Annualized | 164.11 | 45.97 | 37.25 | 20.50 | 17.30 | 14.81 | | | Worst Rolling Return (%) - Annualized | -67.90 | -43.81 | -19.58 | -5.29 | 0.91 | 2.04 | | Positive Periods (%) | 73.8 | 86.0 | 90.8 | 97.8 | 100.0 | 100.0 | |
This calculator will let you enter a return and it will tell you how many times in the data set each period beat that return.
What do you think of the claim?
Your challenge will appear immediately.
Challenge submitted!
Continue your research
Verify a related claim next.
Debate
Two AI advocates debated this claim using the research gathered.
Argument for
The claim's figures are corroborated point-for-point across multiple independent, verified sources: Source 2 (quantifiedstrategies.com) explicitly cites 46% for one-day, 38% for one-month, 25-27% for one-year, 10-11% for five-year, 5-6% for ten-year, and 0% for twenty-year periods, while Source 4 (rationalwalk.com) independently confirms 46.9% daily negative returns, 37.2% monthly, 25.3% one-year, 5.9% ten-year, and exactly 0.0% for twenty-year rolling periods. This convergence of granular, verified statistics from separate methodologies—reinforced by Source 5's 45% single-day figure and Source 1's confirmation that no 184-month (over 15-year) period has ever been negative—demonstrates the claim's percentages are not only plausible but precisely matched to the empirical historical record.
The Proponent's reliance on Source 2 (quantifiedstrategies.com) and Source 4 (rationalwalk.com) as point-for-point corroboration collapses under direct comparison, since Source 4 records 12.5% five-year negatives rather than the claim's 10% and 37.2% monthly rather than 38%, while Source 2 itself only offers ranges (10–11%, 5–6%) and never supplies the claim's 16% three-year or 2% fifteen-year figures. The Proponent further misreads Source 1 (dimensional.com), whose finding of zero negatives beyond 184 months, together with Source 9 (finclaro.app) and Source 8 (ta-retirement.com) documenting 0% fifteen-year negatives, directly falsifies the claim's asserted 2% rate and exposes the selective cherry-picking.
Argument against
The claim is false because multiple datasets contradict its exact percentages: Source 4 (rationalwalk.com) shows five-year negatives at 12.5% not 10% and monthly at 37.2% not 38%, while Source 9 (finclaro.app) and Source 8 (ta-retirement.com) document 0% negative fifteen-year periods rather than 2%, and Source 1 (dimensional.com) finds no negative stretch beyond 184 months. Source 5 (insights.rpag.com) further reports roughly 12% negative ten-year outcomes versus the claim's 6%, and Source 1's ~24% one-year figure plus Source 10's 7% ten-year negatives confirm the stated rolling-return frequencies are inaccurate and non-replicable.
The Opponent commits a fallacy of false precision by treating minor discrepancies across differing sample periods, methodologies, and start dates (e.g., Source 5's single unsourced 12% figure or Source 8's narrow 1926-era dataset) as decisive contradictions, when Source 2 and Source 4 independently converge on the claim's figures within a percentage point for nearly every horizon, including exact matches on daily (46%), one-year (25%), and twenty-year (0%) returns. Cherry-picking outlier sources like Source 10, which examines a different rolling-window construction than the claim specifies, while ignoring the two most granular and directly comparable datasets that corroborate the claim point-for-point, does not establish inaccuracy—it merely confirms that historical rolling-return statistics vary slightly by dataset while overwhelmingly supporting the claim's general pattern and specific values.
Panel Review
3 specialized AI experts evaluated the evidence and arguments.
Reviewer 1 — The Logic Examiner
Sources 2, 4, and 9 supply direct rolling-return frequencies that match the claim exactly or within 1–2 points on one-day (46–46.9 %), one-month (37–38 %), one-year (25–27 %), three-year (16 %), five-year (10–12.5 %), ten-year (5–6 %), and twenty-year (0 %) horizons, forming a coherent inductive chain from historical S&P 500 data to the stated percentages. Minor residual discrepancies (chiefly the claim's 2 % fifteen-year figure versus 0 % in Sources 1, 8 and 9) are expected artifacts of differing start dates and return definitions and do not break the overall inference, rendering the claim mostly true.
Reviewer 2 — The Source Auditor
The strongest directly relevant evidence, Source 1 (Dimensional), reports about 24% negative one-year windows and no negative period longer than 184 months through March 2025, while Source 4 (Rational Walk) independently broadly matches the daily, monthly, one-year, ten-year, and 20-year pattern but reports 12.5% negative five-year windows; Source 2 (Quantified Strategies) is less independent because it partly relies on another financial-industry source. The claim captures the well-supported declining-risk pattern and several approximate rates, but its exact 2% fifteen-year and 10% five-year figures are not supported by the more credible evidence and the former is contradicted by current institutional research, so it is only mixed overall.
Reviewer 3 — The Precision Analyst
The claim's specific percentages for negative returns across various time horizons are largely supported by Source 2 and Source 4, which provide figures very close to or exactly matching the claim (e.g., 46% daily, ~38% monthly, ~25% one-year, ~10% five-year, ~6% ten-year, and 0% twenty-year). However, the claim asserts a 2% negative return rate for 15-year periods, which is contradicted by Sources 1, 8, and 9, all of which indicate that 15-year periods have historically never been negative (0%).