What is Maximum Drawdown? Definition, Formula, and Example
Maximum drawdown is the largest peak-to-trough decline in a portfolio's value over a given period, expressed as a percentage, and serves as the standard measure of worst-case downside risk.
What is Maximum Drawdown?
Maximum drawdown (MDD) is the largest peak-to-trough decline in a portfolio's value over a specified period, expressed as a percentage of the peak. It measures the worst loss an investor would have suffered if they bought at the highest point and sold at the lowest subsequent point. MDD is the canonical measure of downside risk used by hedge funds, allocators, prop firms, and risk managers — because, unlike volatility, it captures the actual capital pain a strategy inflicts on the people running it.
How Maximum Drawdown Is Calculated
The formula is direct:
MDD = (Trough Value − Peak Value) / Peak Value
Constraints:
1. The peak must precede the trough chronologically
2. The trough is the lowest value reached *before* a new peak is established
3. MDD is always expressed as a negative number or absolute percentage
4. A new high resets the calculation — subsequent drawdowns are measured from the new peak
For a daily equity curve E(t), the running peak at time t is P(t) = max(E(0..t)), the drawdown at time t is D(t) = (E(t) − P(t)) / P(t), and MDD over the period is min(D(t)).
Related metrics built on MDD:
- Calmar Ratio = CAGR / |MDD| — risk-adjusted return using MDD as the risk denominator
- MAR Ratio — same as Calmar, named for the Managed Accounts Reports newsletter
- Recovery time — number of periods required to return from trough to prior peak
Worked Example
An SPY buy-and-hold portfolio peaked at $479 in early January 2022 and bottomed at $349 in October 2022. MDD = (349 − 479) / 479 = −27.1%. SPY did not reclaim its 2022 peak until February 2024 — a 25-month underwater period.
A 60/40 SPY/TLT portfolio during the same window peaked at $385 in January 2022 and troughed at $304 in October 2022. MDD = (304 − 385) / 385 = −21.0%. Same period, lower MDD, but the 60/40 also had a much slower recovery because long-duration bonds remained pinned.
By contrast, an all-cash BIL position over 2022 had MDD near 0% — earning a measly 1.5% but suffering no meaningful drawdown.
When Traders Use Maximum Drawdown
MDD is used across the risk-management stack:
1. Strategy backtesting — required alongside CAGR and Sharpe; a 30% CAGR strategy with 60% MDD is unrunnable in practice
2. Position sizing — Kelly-style sizing is often capped by MDD tolerance: "I can stomach 25%, so size accordingly"
3. Allocator due diligence — hedge funds report rolling 3-year MDD in pitch decks
4. Stop-out rules — prop firms set hard MDD limits (e.g., 5% peak-to-trough = account closure)
5. Personal risk tolerance — the honest question every investor should ask before deploying capital
Limitations and Common Misconceptions
MDD is informative but not bulletproof:
- Backward-looking — historical MDD does not bound future losses; black swans regularly deliver worse
- Path-dependent — small changes to the sample window can materially change the reported MDD
- Frequency-blind — a strategy with one 30% drawdown and one with five 30% drawdowns score the same on MDD
- Duration-blind — a 20% MDD that recovers in 3 months is very different from one that recovers in 5 years
- Sample-size sensitive — short backtest periods underestimate MDD; 10+ years of data including a recession is the minimum credible sample
A common mistake is comparing MDDs across strategies with different sample windows. A strategy backtested 2015-2019 (no major bear market) shows a misleadingly small MDD compared to one tested through 2008 or 2020.