Every S&P 500 Bear Market Since 1929: Duration and Recovery

What a Century of S&P 500 Bear Markets Teaches Investors

Bear markets are the rare, high-conviction reset events that define generational investor outcomes. They are also the most studied and most mythologized periods in financial history — and the data on them is more cleanly observable than almost any other return distribution in markets. This article is the complete reference table of every S&P 500 bear market since 1929, with each event's peak date, trough date, percentage decline, duration, and how long it took the index to claw back to a new high.

A bear market in this article means a peak-to-trough decline of 20% or more on a closing basis in the S&P 500 (and its pre-1957 predecessor, the S&P 90). That is the threshold used by Standard & Poor's, the U.S. Securities and Exchange Commission, Yardeni Research, and most mainstream brokerages. Intraday measures, total-return measures, or different index choices will shift specific numbers by a percentage point or two, but the structure of the table is stable across sources.

If you are tracking the index live, the S&P 500 ETF tracker updates the same series in real time. If you want to know how the current tape stacks up against past bears, the companion piece on the S&P 500 Risk Radar covers the AI-driven early-warning signals; the table below is the historical baseline those signals get compared against.

The Complete Table

Each row uses closing prices on the S&P 500. Duration measures peak day to trough day. Recovery measures the next day on which the index closed at or above its prior peak.

The 1990 event carries the asterisk every reference table needs: at its closing-basis trough it printed −19.9%, one tick below the conventional 20% threshold. It is included here because it is universally counted as a bear by Yardeni Research and S&P's own historical work, and because excluding it would distort the recovery statistics by removing the second-fastest event on the page.

What the Data Actually Says

With fifteen bear markets across 96 years, the dispersion is enormous — but the central tendencies are remarkably consistent.

• Average peak-to-trough decline: −36.7%
• Median decline: −33.5%
• Average duration (peak to trough): 15.6 months
• Median duration: 14.7 months
• Average recovery (trough to new high): ~50 months
• Median recovery: 14 months

The mean and the median are very far apart on recovery time. That is because the distribution has a small number of generational events — 1929, 1937, 1973, 2000, 2007 — that took roughly four to twenty-five years to print a new high, and a much larger cluster of ordinary bears that took one to two years. Inflation-adjusted recoveries are longer in every case, and the 1968–1982 window in nominal terms hides what is essentially one continuous, fifteen-year secular bear after inflation.

Excluding the Great Depression — a unique structural event tied to a banking collapse and a 25% unemployment shock — the average post-1937 decline narrows to roughly −33%, and the average recovery shortens to about 30 months.

Two patterns are worth keeping in mind.

1. The bear's depth and the recovery time are weakly correlated. A −33% bear (1987) recovered in 20 months. A −34% bear (2020) recovered in 5. The pace of policy response, the starting valuation, and the trajectory of earnings matter more than the headline drawdown.
2. Most bears end within 18 months from the peak. Eleven of the fifteen reached their trough inside that window. The exceptions — 1929, 1937, 1946, 2000 — were the secular bears, and all four featured either a banking shock, a sustained earnings collapse, or a multi-year valuation reset.

A Tour of the Bears, In Plain English

A reference table is only useful if you can put each row in context. Here is each event in one paragraph.

1929–1932 (−86.2%). Not a stock market story — a banking system collapse, a 25% unemployment shock, and a 30% contraction in real GDP. The S&P 90 lost 86% over 33 months, and on an inflation-adjusted basis the index did not make a new high until the late 1950s. This is the outlier in every chart and the reason the Federal Reserve's modern crisis playbook exists at all. Investopedia's Great Depression overview is the standard reference for the macro mechanics.

1937–1938 (−54.5%). A monetary-policy mistake. The Fed and Treasury tightened too soon, doubling reserve requirements and pulling fiscal stimulus while the recovery was still fragile. The index lost more than half its value in just over a year. Recovery took eight years and the help of the war economy.

1946–1949 (−29.6%). The post-war demobilization bear. Industrial production dropped sharply as wartime contracts ended, and inflation surged. The drawdown was relatively shallow by Depression-era standards but the bottom was choppy and took three years to print.

1956–1957 (−21.6%) and 1961–1962 (−28.0%). Two short, sharp cyclical bears around the 1957 recession and the Kennedy Slide of 1962. Both reset elevated valuations rather than reflecting a true earnings collapse.

1966 and 1968–1970 (−22.2% and −36.1%). The first wave of the Great Inflation. The 1966 mini-bear was triggered by a credit crunch. The 1968–1970 bear was longer and more painful, and it set the stage for the worst event of the post-war era.

1973–1974 (−48.2%). The 1973 oil embargo plus stagflation. The S&P lost nearly half its value over 21 months and did not make a new nominal high until July 1980. In real terms, the recovery did not arrive until the early 1990s. The companion piece on the 1973 oil shock covers the geopolitical mechanics in detail.

1980–1982 (−27.1%). Volcker's inflation war. Fed funds at 20% drove the deepest post-war recession to that point and crushed equity multiples. When inflation broke, so did the bear — the recovery to a new high took just three months from the trough.

1987 (−33.5%). Black Monday. The S&P lost more than 20% in a single day on October 19, 1987. The decline from August was 33.5% and lasted just three months. The interesting fact is the recovery: it took 20 months to make a new high, even though the entire decline took three. Markets digest crashes more slowly than the headlines suggest.

1990 (−19.9%). The Gulf War and Savings & Loan crisis bear. Borderline on the 20% threshold by closing prices, universally counted in the reference series. Three months down, four months back to a new high — the second-fastest cycle on the table after 2020.

2000–2002 (−49.1%). The dot-com unwind. The S&P took 30 months to bottom; the Nasdaq 100 lost 83%. The recovery to a new high in the S&P 500 took until May 2007 — and the next bear started five months later.

2007–2009 (−56.8%). The Global Financial Crisis. A bank balance-sheet event of a scale not seen since 1929, but met with a coordinated global monetary and fiscal response the 1930s never received. The trough at 676.5 on March 9, 2009, became the launchpad of the longest bull market on record. New high in March 2013.

2020 (−33.9%). The COVID crash. The fastest 30%+ drawdown in S&P 500 history — 33 calendar days from peak to trough — and the fastest recovery: a new high in five months, on the back of the most aggressive monetary and fiscal response in peacetime history. The standout entry in the table, and the data point that makes "this time is different" arguments harder to dismiss without evidence.

2022 (−25.4%). The post-COVID inflation bear, driven by the fastest rate-hike cycle since Volcker. Nine months down, two years back to a new high — slower than the COVID recovery, faster than the post-2008 grind.

Lessons the Table Forces You to Confront

A few facts in the data are uncomfortable for the standard narratives.

Time in market beats timing the market — but only because the secular bears are rare. If you missed the four secular events (1929, 1937, 1973, 2000), the recovery math from the other eleven looks gentle. The problem is that you cannot know in advance which bear is a 1987 (three months down, twenty months back) and which is a 2000 (thirty months down, almost five years back) until the bottom is already in. That is the entire reason systematic risk-off rules exist.

Diversification across asset classes only works in specific bears. In 1973 and 2022, both stocks and high-quality bonds lost real purchasing power at the same time. In 1987, 2000, 2008, and 2020, bonds rallied as stocks fell. The correlation between equity and bond returns has switched sign multiple times across the table.

The Fed matters more than valuation. Each of the four fastest recoveries (1982, 1990, 2020, 2022) was preceded by a decisive policy pivot. Each of the four slowest recoveries (1929, 1937, 1973, 2000) was characterized by either a policy mistake or a multi-year secular regime change that the Fed could not credibly counter.

Bear markets are not uniformly negative for everyone. A systematic strategy that re-balances into equities at fixed drawdown thresholds — say, every 10% — would have outperformed buy-and-hold in roughly two-thirds of the bears in the table. The SimianX autopilots implement exactly this kind of rules-based behavior; the comparable article on crypto bear playbooks walks through the analogous logic for digital assets.

Where the Current S&P 500 Sits

If you are reading this in real time, the relevant question is not what past bears did but what the current tape looks like compared to those bears. The S&P 500 ETF page covers the live drawdown from the most recent high, the rolling 200-day moving average, and the breadth indicators that tend to deteriorate before the index itself does. The AI model leaderboard publishes daily probability-of-drawdown signals across 30 frontier models from 6 providers — useful as a second opinion against pure technical screens. For investors who prefer rules-based execution, the SimianX autopilots translate those signals into systematic position changes.

For the broader context — current valuations, breadth, sentiment, and how those compare to readings at past bear-market peaks — the 2026 US stock outlook is the live counterpart to this historical reference.

FAQ

How long does the average S&P 500 bear market last? Roughly 15.6 months peak to trough, with a median of 14.7. Eleven of the fifteen bears in the table reached their trough within 18 months.

How long does the average recovery take? Mean recovery (trough to new high) is about 50 months. Median is 14 months. The mean is heavily skewed by the four secular events (1929, 1937, 1973, 2000); excluding those, the average recovery falls to about 16 months.

What was the worst bear market in history? The 1929–1932 collapse, with a peak-to-trough decline of −86.2% and a nominal-price recovery that took roughly 25 years. On a total-return, inflation-adjusted basis it took even longer.

What was the fastest bear market in history? The 2020 COVID crash. The S&P 500 lost 33.9% in 33 calendar days from its February 19 peak, and printed a new high in August 2020 — five months from the bottom.

Is the 1990 drawdown really a bear market? Closing-basis it bottomed at −19.9%, one tick below the conventional 20% threshold. Major reference series (Yardeni, S&P) include it as a bear; intraday measures put it slightly below −20%. It is included in the table here with that asterisk.

How is "bear market" measured in this article? Closing-basis S&P 500 (and pre-1957 S&P 90), peak-to-trough decline of 20% or more. Duration is measured in calendar months from peak close to trough close. Recovery is the first subsequent close at or above the prior peak.

Methodology Notes

All dates and percentages above are computed from the published S&P 500 closing series cross-referenced against CRSP, Bloomberg, and Yahoo Finance — the three diverge by at most a single trading day on a handful of pre-1990 events. The Great Depression decline is sometimes quoted as −89% using monthly averages or different index conventions; the figure used here (−86.2%) is the closing-basis peak to trough.

For working-investor purposes the exact numbers matter less than the structure: 15 bear markets in 96 years, an average decline of about a third, an average duration of about a year, and a distribution of recoveries that is genuinely bimodal — most bears resolve in 12 to 24 months, but a small minority become generational events that take half a decade or more to retrace. The job of the SimianX Risk Radar is to distinguish between the two cases in real time. The job of the table above is to remind you what is plausible — and what is not.

Related Reading

• Buy the Invasion 2026: 9 of 12 Wars Saw S&P 500 Rally
• Wall Street Drawdown Watch: 10% Warnings, 40% Tail Risk
• Pearl Harbor 1941: Dow -19.8% Crash, 307-Day Recovery

References

• Macrotrends — Dow Jones 100-Year History
• Investopedia — Bear Market
• FRED — S&P 500 Historical Data
• Bloomberg — Financial News & Market Data
• Nasdaq — Stock Exchange & Market Data