A strategy that returns 20% a year sounds great until the year it loses 45% of your account in four months. Max drawdown is the number that tells you, in advance, how bad "bad" can actually get - and whether you'd survive it.
Every investment return comes with an unstated question attached: how much pain did it take to get there? Average annual return answers "how much did I make." Standard deviation and beta answer "how bumpy was the ride, on average." Neither one answers the question that actually determines whether an investor sticks with a strategy through a real crisis: what is the single worst peak-to-trough loss this thing has ever produced?
That is max drawdown. It's one of the simplest risk metrics to calculate and one of the most predictive of real-world investor behavior, because it measures the exact scenario that causes people to panic-sell at the bottom, abandon a fund manager, or blow through a risk budget: not the average day, but the worst stretch.
This guide covers the formula, the recovery-math asymmetry that makes deep drawdowns so dangerous, and real max drawdown figures for the S&P 500 across five modern downturns - computed directly from daily closing prices, not pulled from memory.
The Max Drawdown Formula
Max Drawdown = (Trough Value − Peak Value) ÷ Peak Value
The calculation walks through a price or portfolio-value series in chronological order and does two things at every step:
- Track the running peak - the highest value reached so far, updated any time a new high is hit.
- Measure the drawdown at that point - how far the current value sits below the running peak, as a percentage.
The max drawdown is simply the most negative of those measurements anywhere in the series. It is always zero or negative, and it is a single number describing a single episode - not an average of many small declines, the way standard deviation blends every up and down day into one blended figure.
A worked example: a portfolio grows from $100,000 to $140,000, falls to $91,000, recovers to $150,000, then falls again to $114,000 before finishing the period at $130,000. There are two distinct drawdowns in that path - the fall from $140,000 to $91,000 (-35%) and the fall from $150,000 to $114,000 (-24%). The max drawdown for the full period is the larger of the two: -35%, even though the portfolio went on to reach new highs afterward and ended up 30% above where it started.
Why the Recovery Math Makes This the Metric That Matters Most
The reason max drawdown gets so much attention from professional risk managers isn't abstract - it's arithmetic. Percentage losses and the gains needed to reverse them are not symmetric, because a loss shrinks the base that the recovery percentage has to be measured against.
| Loss | Gain Needed to Break Even |
|---|---|
| -10% | +11.1% |
| -20% | +25.0% |
| -25% | +33.3% |
| -30% | +42.9% |
| -40% | +66.7% |
| -50% | +100.0% |
| -60% | +150.0% |
| -70% | +233.3% |
| -80% | +400.0% |
The relationship is Gain Needed = Loss ÷ (1 − Loss). It stays roughly linear for small losses — a 10% decline is a manageable 11.1% climb back — but it steepens sharply once losses pass about 30%, and it becomes brutal beyond 50%. A portfolio that falls 50% doesn't need a 50% rebound to get back to even. It needs to double. A portfolio that falls 80% needs to increase fivefold.
This is why avoiding a deep drawdown is mathematically worth more than capturing extra upside during calm periods. A strategy that gives up a few points of return in exchange for a shallower worst-case drawdown can come out ahead over a full cycle, purely because it never has to claw its way back from a hole that deep.
S&P 500 Max Drawdown History: Five Modern Downturns, Computed
Historical max drawdown figures get repeated so often that the exact numbers tend to drift with each retelling. The figures below were computed directly from S&P 500 (^GSPC) daily closing prices - tracking the running peak through each period and finding the deepest peak-to-trough decline, the same method described above.
| Downturn | Peak Close | Trough Close | Max Drawdown | Decline Length | Recovery Length | Total Peak-to-Recovery |
|---|---|---|---|---|---|---|
| Dot-com bust | Mar 24, 2000 (1,527.46) | Oct 9, 2002 (776.76) | -49.2% | 30.5 months | 55.7 months | 7.2 years |
| Global Financial Crisis | Oct 9, 2007 (1,565.15) | Mar 9, 2009 (676.53) | -56.8% | 17.0 months | 48.6 months | 5.5 years |
| COVID crash | Feb 19, 2020 (3,386.15) | Mar 23, 2020 (2,237.40) | -33.9% | 1.1 months | 4.9 months | 0.5 years |
| 2022 bear market | Jan 3, 2022 (4,796.56) | Oct 12, 2022 (3,577.03) | -25.4% | 9.3 months | 15.2 months | 2.0 years |
| 2025 tariff selloff | Feb 19, 2025 (6,144.14) | Apr 8, 2025 (4,982.77) | -18.9% | 1.6 months | 2.6 months | 0.35 years |
"Recovery Length" is the time from the trough back to a new closing high above the prior peak - not back to breakeven from the investor's average cost basis, which would be faster, but back to the index actually making new all-time highs again.
A few patterns worth noticing:
- Depth and speed aren't correlated. The COVID crash was nearly as deep as the 2022 bear market (-33.9% vs -25.4%) but took roughly a ninth of the time to unfold - 33 days versus 282 days - and recovered in under five months. The dot-com bust was the slowest-forming and slowest-recovering by a wide margin.
- Recovery almost always takes longer than the decline. In four of the five episodes above, the climb back to a new high took substantially longer than the fall itself - the GFC took 17 months to fall and nearly four years to recover.
- The deepest drawdowns took the longest to fully recover. The two worst episodes by magnitude (GFC and dot-com) are also the two longest by total peak-to-recovery time, which is exactly what the recovery-math table above predicts.
As of this writing (September 11, 2026 close), the S&P 500 sat at 7,656.98, about 1.8% below its August 13, 2026 closing high of 7,798.99 - a routine pullback, not a drawdown event by the standard most of the episodes above cross.
Max Drawdown vs. Volatility, Sharpe, and Sortino
Max drawdown answers a narrower question than the more commonly cited risk metrics, and that narrowness is the point.
- Standard deviation / volatility measures how much returns bounce around on average, counting upside swings as "risk" right alongside downside ones. A stock that jumps 8% one week and drops 8% the next has the same standard deviation as one that grinds sideways - but very different lived experiences for the holder.
- Sharpe ratio divides excess return by standard deviation, so it inherits volatility's blindness to direction. A strategy with one catastrophic quarter buried inside years of small gains can post a respectable Sharpe ratio while carrying a max drawdown that would have wiped out an underprepared investor. For the full formula and how professionals use it, see Sharpe Ratio Explained.
- Sortino ratio improves on Sharpe by only penalizing downside deviation, but it's still an average of downside moves across the whole period - not the single worst stretch.
- Ulcer Index is the metric built specifically to weight depth and duration of drawdowns together, closer in spirit to max drawdown than to Sharpe or Sortino, but it produces a blended score across every drawdown in the period rather than isolating the single worst one. See Ulcer Index: Measuring Drawdown Stress for the formula and how it differs.
None of these metrics is wrong - they answer different questions. Max drawdown answers the one that matters most for whether an investor can actually stay invested: what is the worst single stretch I would have had to sit through, and could I have handled it?
How Max Drawdown Is Actually Used
Fund and strategy evaluation. Due diligence questionnaires for hedge funds and managed strategies almost always ask for max drawdown alongside annualized return, because two managers with identical returns can have wildly different risk profiles. A manager who returned 12% annually with a -15% max drawdown took a fundamentally different path than one who returned 12% annually with a -45% max drawdown, even if their average returns and even their standard deviations look similar.
Position sizing. Traders and portfolio managers often size positions partly around a strategy's historical max drawdown, working backward from "how much of my capital am I willing to see erased in the worst case" to a position size that keeps that worst case tolerable. A strategy with a documented 40% historical max drawdown gets sized very differently than one with a documented 12% max drawdown, even at identical expected returns.
Setting realistic expectations. A strategy's historical max drawdown is close to the best available estimate of its future worst case - not a guarantee, since a new max drawdown can always exceed the historical one, but a far better anchor than assuming the future will be gentler than the past. An investor who has genuinely internalized "this strategy has fallen 35% before and could do it again" is far less likely to panic-sell during the next decline than one who has only ever seen the average annual return quoted.
Managing Your Own Drawdown Risk
The practical takeaway from the recovery-math table isn't "avoid all risk" - it's that the deeper a drawdown gets, the more disproportionately valuable it becomes to cap it before it compounds further. A few concrete tools:
- Position sizing relative to worst-case scenarios, not just expected returns, so that no single holding can inflict a portfolio-level drawdown you couldn't stomach.
- Trailing stops or drawdown alerts on individual positions, so a decline gets flagged for review while it's still a manageable percentage rather than after it has compounded past the point where the recovery math turns ugly.
- Diversification across assets that don't all draw down at the same time - a portfolio concentrated in a single sector or factor will tend to experience its max drawdown all at once, while a genuinely diversified one smooths the worst-case path.
Setting a price alert at a pre-planned trailing-stop level in Stock Alarm Pro gives you a review window before a position's drawdown deepens - a notification the moment a holding crosses a threshold you set in advance, while you still have time to decide whether the decline is noise or the start of something that needs action. Pair that with the Screener to check a stock's trend state and volatility profile before sizing a position in the first place, so the worst-case scenario is one you've already thought through rather than one you discover in real time.
