Most risk metrics — standard deviation, Value at Risk — summarize risk as a statistical property of returns. But investors experience risk as watching their portfolio decline from a peak and wondering how much further it will fall. Maximum drawdown captures exactly that: the worst peak-to-trough loss over a given period. This guide covers what maximum drawdown measures, how to calculate it, how it compares to other risk metrics, and why recovery time matters just as much as the magnitude of the decline.
What is Maximum Drawdown?
Maximum drawdown (MDD) is the largest peak-to-trough percentage decline in an investment’s value before a new peak is established. It measures the worst cumulative loss an investor would have experienced over a specific period — the deepest hole the portfolio fell into before recovering.
Maximum drawdown answers the question: “If I had invested at the worst possible moment and sold at the worst possible moment within this period, how much would I have lost?” It captures the single worst loss experience — something that standard deviation and single-period return metrics cannot convey.
Unlike a single-day loss or a monthly return, maximum drawdown is a cumulative, path-dependent measure. A drawdown can span days, months, or even years. It begins when the portfolio starts declining from a peak and ends only when the portfolio surpasses that previous peak. The maximum drawdown is the deepest point within any such drawdown episode across the entire observation period.
MDD is widely used in fund evaluation, risk budgeting, strategy comparison, and investor suitability assessments. It complements volatility-based metrics like beta and standard deviation by focusing exclusively on downside risk as investors actually experience it.
The Maximum Drawdown Formula
Where:
- Vpeak — the highest portfolio value before the subsequent trough
- Vtrough — the lowest portfolio value after the peak, before the portfolio recovers to a new high
The result is a negative number (e.g., -0.339 for a 33.9% drawdown). Some sources express MDD as a positive absolute value — this article uses the negative convention throughout, which keeps the direction of loss explicit.
MDD can be calculated even if the portfolio has not yet recovered to a new peak. If the current drawdown from the most recent high is the deepest on record, that is the maximum drawdown — it simply hasn’t ended yet.
One of the most important consequences of a large drawdown is the asymmetric gain required to recover:
The asymmetry of losses and gains is why drawdowns are so damaging. A -50% loss requires a +100% gain to recover. A -33% loss requires a +49.3% gain. This compounding penalty makes avoiding large drawdowns far more valuable than chasing equivalent upside — a core principle in risk management and portfolio construction.
MDD values depend on data choices: price index vs total return (with dividends reinvested), daily vs monthly observation frequency, and the lookback window. Daily data captures intra-month troughs that monthly data misses. Always note these choices when reporting or comparing drawdowns.
How to Calculate Maximum Drawdown
Computing maximum drawdown is a four-step process:
- Build the equity curve — track the portfolio or index value over time (not just periodic returns)
- Track the running maximum — at each observation point, record the highest value seen so far
- Calculate the drawdown at each point — compute (Current Value – Running Maximum) / Running Maximum
- Find the minimum — the most negative drawdown value is the maximum drawdown; record the peak date and trough date
Here is a step-by-step calculation using selected S&P 500 daily closing prices around the COVID-19 crash:
| Date | S&P 500 Close | Running Peak | Drawdown |
|---|---|---|---|
| Jan 17, 2020 | 3,329.62 | 3,329.62 | 0.0% |
| Feb 19, 2020 | 3,386.15 | 3,386.15 | 0.0% (new peak) |
| Mar 12, 2020 | 2,480.64 | 3,386.15 | -26.7% |
| Mar 23, 2020 | 2,237.40 | 3,386.15 | -33.9% (trough) |
| Apr 29, 2020 | 2,939.51 | 3,386.15 | -13.2% |
| Aug 18, 2020 | 3,389.78 | 3,389.78 | 0.0% (new peak) |
Calculation:
MDD = (2,237.40 – 3,386.15) / 3,386.15 = -1,148.75 / 3,386.15 = -33.9%
The S&P 500 fell 33.9% from its February 19 peak to its March 23 trough — a peak-to-trough duration of approximately 33 calendar days. The index recovered to a new all-time high by August 18, 2020, making the total underwater period roughly 6 months.
Historical Maximum Drawdowns
Examining major S&P 500 drawdowns provides perspective on the range of outcomes investors have faced and — critically — how long recovery took:
| Event | Peak → Trough | Max Drawdown | Peak-to-Trough | Trough-to-Recovery | Total Underwater |
|---|---|---|---|---|---|
| 2007–09 Financial Crisis | Oct 2007 → Mar 2009 | -56.8% | ~17 months | ~4 years | ~5.5 years |
| 2000–02 Dot-Com Bust | Mar 2000 → Oct 2002 | -49.1% | ~31 months | ~4.5 years | ~7 years |
| 2020 COVID-19 Crash | Feb 2020 → Mar 2020 | -33.9% | ~1 month | ~5 months | ~6 months |
| 1987 Black Monday | Aug 1987 → Dec 1987 | -33.5% | ~3 months | ~20 months | ~23 months |
S&P 500 price index, daily closing values. Durations are approximate calendar time.
These episodes illustrate a crucial pattern: drawdowns of similar magnitude can have vastly different recovery profiles. The 2020 COVID crash and 1987 Black Monday had nearly identical drawdowns (~-34%), but COVID recovered in 6 months while Black Monday took nearly 2 years. The 2008 financial crisis was underwater for over 5.5 years from peak to new peak — an entire market cycle. Diversification across asset classes with low correlation is the primary tool for reducing portfolio-level drawdowns.
Maximum Drawdown vs Standard Deviation vs Value at Risk
Maximum drawdown, standard deviation, and Value at Risk (VaR) each capture a different dimension of risk. Understanding what each measures — and what it misses — is essential for comprehensive risk assessment:
Maximum Drawdown
- Measures: worst peak-to-trough cumulative loss
- Time scope: spans multiple periods (path-dependent)
- Risk focus: downside only — worst actual loss experience
- Investor relevance: directly reflects what an investor would have endured
- Limitation: single worst episode; may not represent typical risk
Standard Deviation
- Measures: dispersion of periodic returns around the mean
- Time scope: summarizes return variability (often annualized)
- Risk focus: symmetric — penalizes upside and downside equally
- Investor relevance: core input for portfolio optimization
- Limitation: doesn’t distinguish upside from downside volatility
Value at Risk
- Measures: threshold loss at a confidence level (e.g., “5% chance of losing more than X%”)
- Time scope: single-period (1-day or 10-day typical)
- Risk focus: downside threshold — doesn’t tell you how bad it gets beyond VaR
- Investor relevance: widely used in regulatory capital requirements
- Limitation: silent about losses beyond the VaR threshold
These metrics answer fundamentally different questions. Standard deviation tells you how much returns vary in a typical period. VaR gives you a quantile-based loss threshold. Maximum drawdown tells you what actually happened at the worst point — the cumulative loss an investor endured from peak to trough. A portfolio with low standard deviation can still suffer a deep drawdown if small losses cluster sequentially. VaR gives the threshold, Expected Shortfall (CVaR) gives the average loss beyond that threshold, and MDD gives the worst realized path loss. Evaluating risk-adjusted returns with the Sharpe ratio alongside drawdown analysis helps determine whether a portfolio’s returns justify the losses it has experienced.
Recovery Time and Drawdown Duration
The magnitude of a drawdown is only half the story. How long it takes to recover is equally important — and often more relevant to investor behavior and financial planning. Three time metrics define the full drawdown experience: peak-to-trough (how long the decline lasted), trough-to-recovery (how long it took to reach a new peak), and total underwater time (the sum of both). The asymmetry between losses and gains makes recovery disproportionately difficult:
| Drawdown | Required Gain to Recover |
|---|---|
| -10% | +11.1% |
| -20% | +25.0% |
| -30% | +42.9% |
| -50% | +100.0% |
| -60% | +150.0% |
Recovery time is often more important than drawdown magnitude for real-world investor behavior. A -30% drawdown that recovers in 6 months is far more tolerable than a -30% drawdown that takes 4 years to recover. The 2020 COVID crash (-33.9%) recovered in roughly 6 months; the 2008 financial crisis (-56.8%) took over 5 years. For stress-testing how portfolio drawdowns might unfold under different scenarios, Monte Carlo simulation can model the probability distribution of drawdown paths and recovery times.
Common Mistakes
1. Ignoring recovery time. Focusing only on drawdown magnitude without considering how long it took to recover misses half the picture. A -25% drawdown that recovers in 3 months is fundamentally different from one that takes 3 years — yet both report the same MDD. Always report underwater duration alongside drawdown depth.
2. Using only max drawdown without context. MDD is a single data point. Two portfolios with the same MDD can have very different risk profiles — one might have experienced frequent moderate drawdowns while the other had a single sharp drop followed by steady gains. Supplement MDD with standard deviation and drawdown frequency for a more complete picture.
3. Confusing max drawdown with a single-day loss. A single-day drop of -7% is not a maximum drawdown. MDD measures the cumulative decline from a peak, which may span days, months, or years. The S&P 500’s -33.9% COVID drawdown accumulated over approximately one month of trading, not a single session.
4. Computing MDD from periodic returns instead of the equity curve. You cannot calculate maximum drawdown by scanning individual monthly or daily returns for the largest negative number. You must first build the cumulative value series (equity curve) from those returns, then track the running maximum and compute drawdowns from it. The largest single-period return is not the same as the largest peak-to-trough decline.
5. Assuming past max drawdown is the worst that can happen. MDD is backward-looking. Future drawdowns can exceed historical ones. The 2008 financial crisis (-56.8%) exceeded any post-WWII precedent in U.S. equities at the time. Never treat historical MDD as a floor for future losses.
6. Comparing drawdowns across different time periods. A fund with 5 years of track record will almost certainly show a smaller MDD than one with 20 years, simply because it has encountered fewer market cycles. Always note the observation period and start date when comparing drawdowns across funds or strategies.
Limitations of Maximum Drawdown
Maximum drawdown is inherently backward-looking and captures only the single worst historical episode. It provides no probabilistic framework for future losses and can be misleading if the observation window does not include a major market stress event.
1. Single worst episode. MDD reduces an entire return history to one number. It tells you nothing about the frequency, duration, or severity of other drawdowns. A portfolio that experienced one sharp -30% drop and then smooth returns looks identical (by MDD alone) to one that endured three separate -28% drawdowns.
2. Backward-looking. Like all historical risk measures, MDD only reflects what has already happened. A fund that launched in 2010 and was measured through 2019 would show a relatively mild MDD — completely missing the 2008 crisis. The observation window fundamentally determines the result.
3. Path-dependent. The same set of monthly returns, reordered, can produce very different maximum drawdowns. Consider two sequences of annual returns: {+20%, -15%, -15%, +20%} produces a deeper drawdown than {-15%, +20%, +20%, -15%}, despite identical average returns. This makes MDD sensitive to the specific sequence of returns, not just their distribution.
4. No probabilistic framework. Unlike VaR, which provides a quantile-based loss threshold, MDD is a single observation with no confidence interval. You cannot say “there is a 5% chance of a drawdown exceeding X%” from MDD alone. For probabilistic drawdown analysis, combine MDD with Monte Carlo simulation or bootstrap resampling.
Maximum drawdown is one of the most intuitive and investor-relevant risk metrics — it directly answers “how bad did it actually get?” But it should be combined with standard deviation (for typical return variability), VaR/CVaR (for probabilistic loss estimation), and the Sharpe ratio (for risk-adjusted return assessment) for a complete risk evaluation. For a deeper dive into risk metrics and portfolio analysis techniques, see our Portfolio Analytics & Risk Management course.
Frequently Asked Questions
Disclaimer
This article is for educational and informational purposes only and does not constitute investment advice. Maximum drawdown values cited are based on the S&P 500 price index using daily closing data and may differ based on the data source, observation frequency, total-return vs price-return methodology, and time period used. Past drawdowns do not predict future drawdowns. Always conduct your own research and consult a qualified financial advisor before making investment decisions.
