How Often Do 10%, 20%, 50% Drops Happen?
How Often Do 10%, 20%, 50% Drops Happen?
Knowing that crashes are inevitable is one thing. Knowing how often they occur is another. Investors who understand frequency can calibrate realistic expectations and avoid panic when a normal drawdown arrives. The data is surprisingly consistent across decades.
Quick definition: Correction frequency refers to how often a market experiences drawdowns of specific severity levels (10%, 20%, 50%+). Historical frequency data helps predict future occurrence likelihood.
Key Takeaways
- 10%+ corrections occur roughly every 1–2 years on average
- 20%+ bear markets occur roughly every 5–7 years on average
- 30%+ severe drawdowns occur every 10 years on average
- 50%+ crashes occur every 20–30 years on average
- Frequency varies by decade and by asset class, but the pattern repeats consistently
The S&P 500: 125 Years of Data
The most comprehensive dataset is the U.S. stock market (S&P 500 or its predecessor index) going back to 1900. This gives us 125+ years of market behavior across booms, depressions, world wars, and crises.
10%+ Corrections: The S&P 500 experiences a 10% decline from recent peak roughly every 12–18 months. Some years have zero 10% corrections (rare). Others have multiple. Over the long run, expect 6–10 per decade.
20%+ Bear Markets (Technical Bear): A 20% decline is conventionally called a bear market. The S&P 500 has experienced approximately 25–30 bear markets since 1900. That's roughly one every 4–5 years—closer to every 5–7 years when you account for clustering.
30%+ Severe Drawdowns: A third of the portfolio gone. The S&P 500 has experienced roughly 12 declines of 30% or more in 125 years. That's one roughly every 10 years on average.
50%+ Crashes: These are the worst-case scenarios that test investor resolve completely. The S&P 500 has experienced roughly 7 declines of 50% or more since 1900. That's one every 17–20 years on average.
| Severity | Typical Frequency | Average Time Between |
|---|---|---|
| 10%+ | Every 12–18 months | ~18 months |
| 20%+ | Every 5–7 years | ~6 years |
| 30%+ | Every 8–12 years | ~10 years |
| 50%+ | Every 17–25 years | ~20 years |
Decade-by-Decade Breakdown
Frequency isn't perfectly stable—some decades are more crash-heavy than others.
1900s–1920s: Highly volatile. Multiple crashes, including the 1907 Panic (49% decline). Market structure was less regulated, and leverage was high.
1930s: The Great Depression dominated. The 1929–1932 crash was 89%, and recovery took 25 years to exceed the 1929 peak.
1940s–1950s: Relatively stable. One major drawdown (1946–47, around 30%) but strong growth otherwise.
1960s: Stable bull market. Few significant drawdowns.
1970s: Stagflation era. Multiple 20%+ drawdowns and high volatility.
1980s: Strong bull market after the 1981–82 recession. The 1987 Crash (22% in one day) was severe but recovered quickly.
1990s: Powerful bull market with few significant drawdowns until the 1998 Long-Term Capital Management crisis (a 20% correction recovered quickly).
2000s: High volatility. The dot-com crash (2000–2003) was 49%. The financial crisis (2007–2009) was 57%.
2010s: Strong bull market with few severe drawdowns. The 2015 "flash crash" was brief and shallow.
2020s (so far): The 2020 COVID crash was 34% (recovered in 5 months). The 2022 decline was 27%.
What This Means: Probability Expectations
If you're a 40-year-old planning to invest until age 80, you have a 40-year horizon. Based on historical frequency:
- 10%+ corrections: You'll experience ~25–30 of them
- 20%+ bear markets: You'll experience 5–8 of them
- 30%+ severe declines: You'll experience 3–5 of them
- 50%+ crashes: You'll likely experience 1–2 of them
This isn't speculation. This is what the data shows. If you're uncomfortable with experiencing multiple 20%+ drawdowns and at least one 50%+ crash, your asset allocation is too aggressive for your actual risk tolerance.
Behavioral Data: How People React
Dalbar Inc. studies investor behavior during drawdowns. The data shows:
- Most investors sell during drawdowns (locking in losses near the bottom)
- The deeper the drawdown, the more selling accelerates
- Average investors underperform buy-and-hold investors by 3–5% per year, primarily due to emotional trading around crashes
The 2008 financial crisis is instructive. Investors who held recovered gains by 2013. Investors who sold in late 2008 and re-entered in 2012 captured only half the subsequent recovery. The emotional cost of the drawdown (the realized loss) exceeded the mathematical benefit of "missing" the decline.
Asset Class Variations
Different asset classes have different drawdown frequencies.
Technology Stocks: Higher volatility. 20%+ declines every 2–3 years on average. 50%+ declines every 8–10 years.
Broad Market Index (S&P 500): 20%+ declines every 5–7 years. 50%+ every 20 years.
Bonds (U.S. Treasuries): Less frequent drawdowns. 10%+ declines roughly every 8–10 years. 20%+ declines rare (every 30+ years).
Dividend Stocks (Dividend Aristocrats): Moderate volatility. 20%+ declines every 7–10 years. Less severe than broad market.
Emerging Markets: More volatile. 20%+ declines every 3–5 years. 50%+ declines every 12–15 years.
This is why diversification works: not all asset classes crash simultaneously at the same severity. Bonds often rally during equity crashes. International stocks don't always move with U.S. stocks.
The Distribution Isn't Uniform
Drawdowns don't spread evenly across years. They cluster. Long expansions with minimal drawdowns (like 1995–1999) are followed by clusters of drawdowns in short periods (2000–2003).
This clustering is important psychologically. An investor who experiences three 15% declines across 8 years might endure them fine. An investor who experiences two 30% declines in 18 months often panics. The psychological impact of rapid losses is higher than identical losses spread over decades.
Statistical Outliers
Some drawdowns exceed historical frequency predictions. The 2008 crisis was worse than one might predict from statistical models. The 1929 crash was an extreme outlier. These tail events are why:
- Diversification is important (it reduces the worst-case scenario)
- Position sizing matters (don't bet the farm on one idea)
- Emergency funds are essential (don't force selling into crashes for living expenses)
Statistical models based on historical data underestimate the severity of tail events. This is why Nassim Taleb emphasizes "tail risk"—the possibility that markets behave worse than historical patterns predict.
Real-World Examples
The Nifty Fifty (1970s): A group of 50 high-quality growth stocks fell 50%+ in the 1973–74 bear market, far exceeding their historical volatility. Investors shocked by "high-quality" stocks behaving so badly.
Japanese Stocks (Nikkei): The Nikkei 225 fell 63% from 1989–1992. Japanese investors expected lower volatility given Japan's stable economy. They were wrong about frequency.
Cryptocurrency: Bitcoin has experienced 65–80% drawdowns multiple times in 15 years. The frequency of severe drawdowns in emerging asset classes is higher than in mature markets.
Dividend Aristocrats: Companies with 50+ years of dividend increases still experienced 30–40% declines in 2008 and 2020. A history of consistency doesn't prevent drawdowns.
Common Mistakes
Overweighting recent periods: If you've only been investing for 5 years in a bull market, you've never experienced a bear market. Assuming the future will look like the recent past is the mistake Buffett warns about.
Conflating probability with timing: A 20% drawdown happens on average every 5–7 years. That's a probability statement, not a prediction. You can't use it to time entry/exit.
Ignoring personal frequency tolerance: The data says you'll experience 5–8 bear markets in a 40-year horizon. Your personal tolerance might be for only 1–2. In that case, reduce your stock allocation.
Assuming diversification eliminates drawdowns: Bonds reduce drawdowns but don't eliminate them. A 60/40 portfolio still experiences 20%+ declines every 10–15 years.
FAQ
Q: Does frequency stay the same going forward? A: Probably. The pattern of crashes every 5–7 years has held for 125 years despite massive changes in market structure, regulation, and technology. But there's no guarantee.
Q: Will algorithmic trading increase crash frequency? A: Possibly in magnitude (crashes might be faster) but probably not in frequency. The fundamentals of sentiment swings and information surprises haven't changed.
Q: Can you use frequency data to predict the next crash? A: No. Just because a crash happens every 5–7 years doesn't tell you if the next one is 3 months away or 3 years away. This is why market timers fail despite knowing historical frequency.
Q: Should I reduce stock allocation because a crash is "due"? A: No. Crashes are always "due" in probability but not in timing. This reasoning costs investors billions annually in opportunity costs from holding too much cash.
Q: Is a crash more likely after a long bull market? A: Not necessarily. The longest bull market in history (2009–2021, 12 years) contradicted the "due for a crash" narrative. Valuation matters more than duration.
Q: What's the worst frequency we should plan for? A: Plan for a 30%+ drawdown every decade and a 50%+ drawdown every 20 years. If you can't stomach a 50% loss over 40 years, your asset allocation is misaligned with your risk tolerance.
Related Concepts
Sequence of Returns Risk: The timing of returns matters. Two investors with identical average returns but different sequences can end up with very different wealth—because crashes at the wrong time are costlier.
Black Swan Events: Draws that exceed statistical predictions. The 2008 crisis was worse than financial models predicted. Always maintain some extra margin of safety.
Volatility Term Structure: Short-term volatility (crashes) is higher than long-term volatility (smooths out). One 50% drawdown over 20 years translates to much lower volatility when annualized.
Recession Frequency: Recessions are the economic equivalent of crashes and follow similar frequency patterns (roughly every 4–6 years).
Summary
The historical data is clear: crashes aren't anomalies. They're regular features of equity markets. A 20% drawdown every 5–7 years is normal. A 30% decline every 10 years is expected. If your financial plan breaks because the market experiences a 30% drawdown in any given decade, your plan is unrealistic.
Accepting this frequency changes how you plan. Instead of trying to avoid the inevitable, you engineer portfolios and mindsets to survive them. In the next article, we'll examine the anatomy of a bear market—how it typically unfolds and why the middle is always the scariest part.