Market Statistics
Numbers that show the market is stranger than it looks
Table of Contents
🧭 How to read this page
Market statistics are unusually easy to get wrong, and the usual failure is not an invented number. It is a real number attached to the wrong basis.
An intraday low quoted as an opening price. A peak-to-trough loss labelled with the wrong date range. An inflation-adjusted figure presented as if it were nominal. In every case the figure traces back to a reputable source, so checking the number finds nothing wrong. The framing is what broke.
So every figure here carries its basis and its date. Which index, closing or intraday, price or total return, nominal or inflation-adjusted, and over exactly what period. Where a statistic is routinely misused, there is a note saying what it does not mean. If a number could not be traced to a primary source, it is not on this page.
All figures checked against their sources in September 2026.
📉 How far markets fall, and how long they take to come back
The worst decline in US stock market history, and nothing since has come close.
Basis Dow Jones Industrial Average, daily closing price, nominal, price-only. Peak close 381.17 on 3 September 1929; trough close 41.22 on 8 July 1932. Two years and ten months.
Source: Federal Reserve History
How long it took to recover from 1929 depends entirely on what you measure — and both answers are true.
Basis On a nominal, price-only basis the Dow did not close above its 1929 peak until 23 November 1954, just over 25 years. On a real, total-return basis for the broad US market — dividends reinvested, adjusted for the deflation of the 1930s — a lump sum invested at the September 1929 peak was back to break-even by late 1936, about seven years.
Sources: Federal Reserve History and Morningstar
The financial crisis cut the S&P 500 by more than half in seventeen months.
Basis S&P 500, daily closing price, nominal, price return. Peak close 1,565.15 on 9 October 2007; trough close 676.53 on 9 March 2009.
Sources: Federal Reserve History, Yardeni Research
In 2020 the S&P 500 fell 33.9% in thirty-three calendar days, then set a new record six months later.
Basis S&P 500, daily closing price, nominal, price return. Peak close 3,386.15 on 19 February 2020; trough close 2,237.40 on 23 March 2020; new record close 3,389.78 on 18 August 2020, 181 days after the peak.
Source: FRED, data from S&P Dow Jones Indices
The largest single-day fall in the Dow’s history happened on an ordinary Monday with no news event to explain it.
Basis Dow Jones Industrial Average, 19 October 1987, daily close, nominal: down 508 points, −22.6%. The S&P 500 fell 20.4% the same day, from 282.70 to 224.84.
Sources: Federal Reserve History, Schwert, NBER
The Nasdaq fell roughly 78% after the dot-com peak and did not close above it again until 2015.
Basis Nasdaq Composite, daily closing price, nominal, price return. Peak close 5,048.62 on 10 March 2000; trough in October 2002; regained the peak on 23 April 2015 at 5,056.06 — fifteen years and one month.
Source: NPR
🌐 What “the market” actually is
Most people picture the stock market as a broad average that drifts upward. The research on individual stocks describes something much stranger.
Fewer than half of all US stocks have beaten a one-month Treasury bill over their own lifetime.
Basis 25,967 individual US common stocks in the CRSP database, July 1926 to December 2016. Lifetime buy-and-hold return with dividends reinvested, nominal, measured over each stock’s own listed life and compared with one-month Treasury bills over the matched period.
Source: Hendrik Bessembinder, Do Stocks Outperform Treasury Bills?, Journal of Financial Economics, 2018
All of the net wealth ever created by the US stock market came from 1,092 companies. Half of it came from 90.
Basis Same study and sample. Wealth creation measured in dollars rather than percentages, so it accounts for how much capital was actually invested. Total net wealth creation to December 2016: approximately $34.8 trillion. The other 96% of companies collectively matched Treasury bills.
Source: Bessembinder, 2018
Most US stocks that have ever listed ended with a negative cumulative return.
Basis 29,078 US publicly listed common stocks, December 1925 to December 2023, cumulative return over each stock’s listed life. The best performer over the period was Altria, at a cumulative 265 million percent.
Source: Bessembinder, Which U.S. Stocks Generated the Highest Long-Term Returns?, 2024
Outside the United States, the pattern is the same or slightly worse.
Basis Nearly 62,000 global common stocks, 1990 to 2018. 61% of non-US stocks and 56% of US stocks underperformed one-month US Treasury bills over their lifetimes. The top 1.3% of firms account for all global net wealth creation.
Source: Bessembinder, Chen, Choi and Wei, Do Global Stocks Outperform US Treasury Bills?, 2019
Ten holdings make up well over a third of the S&P 500.
Basis Float-adjusted market-cap weights of the SPDR S&P 500 ETF, which tracks the index by full replication, as at 11 September 2026. Because Alphabet appears as two share classes, those ten line items represent nine distinct companies.
Source: State Street Global Advisors
🎲 Why streaks fool us
The statistics above are surprising partly because human beings are poorly built for reading random sequences. These are not numbers about markets; they are findings about the people in them.
The gambler’s fallacy
Believing that an independent event becomes “due” because it has not happened recently. A coin that has landed heads six times is no more likely to land tails on the seventh throw, but almost everyone feels that it is.
The evidence — researchers analysed 18 hours of casino surveillance video from Reno, Nevada in July 1998: 904 roulette spins, 139 players, 24,131 individual bets. After streaks of one to four like outcomes, betting was indistinguishable from chance. After streaks of six or more, 85% of bets were placed against the streak. Source: Croson and Sundali, The Gambler’s Fallacy and the Hot Hand, Journal of Risk and Uncertainty, 2005.
It affects professionals making serious decisions
The fallacy is not confined to casinos. It shows up in the work of people who are trained, paid and accountable, in decisions that change lives.
The evidence — across 150,357 asylum decisions by 357 judges between 1985 and 2013, a judge was 0.5 percentage points less likely to grant asylum immediately after granting the previous case, with no relationship to the merits. Across roughly 900,000 called pitches by 127 baseball umpires, a called strike was 0.9 percentage points less likely immediately after a called strike, and the effect was ten to fifteen times larger on ambiguous pitches. Source: Chen, Moskowitz and Shue, Decision Making Under the Gambler’s Fallacy, Quarterly Journal of Economics, 2016.
People bet against numbers that just came up
A lottery draw is as close to independent as any event gets, and the payout structure means betting with the crowd is costly. People still avoid recent winners.
The evidence — in the Maryland daily numbers game, the amount wagered on a particular number falls sharply immediately after that number is drawn, then recovers to its previous level only gradually over several months. Source: Clotfelter and Cook, The Gambler’s Fallacy in Lottery Play, Management Science, 1993.
The hot hand, and the study that reversed
The mirror image of the gambler’s fallacy is believing a streak will continue. For thirty years the hot hand was the textbook example of an illusion. Then the statistics were re-examined.
The evidence — Gilovich, Vallone and Tversky found no evidence of streak shooting in basketball in 1985, and the hot hand became a standard example of a cognitive illusion. In 2018 Miller and Sanjurjo showed the original method contained a selection bias: under the original study design, genuinely random shooting would be expected to produce a negative streak effect of about 8 percentage points. Correcting for it turns the original result into a significant positive effect of roughly 13 points. Source: Miller and Sanjurjo, Econometrica, 2018.
Investors expect the recent past to continue
After a strong run, people expect more of the same. The uncomfortable part is when that optimism peaks.
The evidence — across six independent surveys of investor expectations from 1963 to 2011, expected returns were strongly correlated with past returns and with the level of the market. They were also negatively correlated with model-based expected returns. Investors were most optimistic precisely when the objective outlook was weakest. Source: Greenwood and Shleifer, Expectations of Returns and Expected Returns, Review of Financial Studies, 2014.
One caution on the hot hand. The 2018 correction is about basketball shooting. It says nothing about whether streaks exist in fund performance or stock returns — and the persistence data further down this page points firmly the other way.
⏱ What timing costs
The most quoted statistic in investing: miss a handful of days and most of the return disappears.
Basis S&P 500 total return, dividends reinvested, $10,000 invested for the 30 years 1996 to 2025, before fees and taxes. Fully invested: $192,167. Missing the ten best days: $85,490. Missing the twenty best days: $49,551.
Source: Hartford Funds, using Ned Davis Research and Morningstar data
Avoiding the worst days helps more than missing the best days hurts.
Basis Dow Jones Industrial Average, 1 January 1900 to 31 December 2006, 29,190 trading days, $100 invested, price return. Fully invested: $25,746. Missing the ten best days: $9,008, or 65% lower. Missing the ten worst days: $78,781, or 206% higher. Across fifteen international markets the same asymmetry held.
Source: Javier Estrada, Black Swans and Market Timing, Journal of Investing, 2008
The best days and the worst days come from the same weather.
Basis US stocks 1928 to 2010, plus fifteen foreign markets, with the market regime defined by the 200-day moving average. Roughly 60 to 80% of both the best and the worst individual days occurred while the market was already below its 200-day average. Across foreign markets, 76% of the worst days and 67% of the best days fell in that regime.
Source: Meb Faber, Where the Black Swans Hide, 2011
📊 Funds, investors, and the gap
Over ten years, most professional large-cap fund managers did not beat the index.
Basis Actively managed US large-cap funds measured against the S&P 500, returns net of fees excluding loads, for periods ending 31 December 2025. Over ten years 85.59% underperformed; over twenty years 92.89%. Corrected for survivorship: funds that closed or merged during the period are counted as underperformers.
Source: S&P Dow Jones Indices, SPIVA US Scorecard, Year-End 2025
Of the large-cap funds in the top quartile at the end of 2021, none were still there four years later.
Basis US large-cap funds ranked on trailing net-of-fee returns within their category. Of those in the top quartile at December 2021, 0.00% remained top quartile at December 2025. Pure chance would predict about 0.4%. Over the shorter two-year window, 28.9% of the December 2023 top quartile were still there at December 2025.
Source: S&P Dow Jones Indices, US Persistence Scorecard, Year-End 2025
Investors earn less than the funds they own — but how much less is genuinely disputed.
Basis The gap between what a fund returns and what the average dollar in it earns, in percentage points per year. Morningstar’s Mind the Gap 2026 puts it at 1.2 points for the ten years to December 2025. A 2026 paper in the Financial Analysts Journal, using Morningstar’s own sample, puts it at 0.10 points. Friesen and Sapp, studying 7,125 US equity funds from 1991 to 2004, found 1.56 points.
Sources: Morningstar, Financial Analysts Journal, Friesen and Sapp
Educational content only. Nothing on Market Organized is individualised investment, trading, tax or financial advice. Every figure on this page is linked to its source so you can check it, and past market behaviour tells you nothing reliable about what happens next.