Glossary
Maximum Drawdown
The largest peak-to-trough decline in an equity curve. The most behaviorally honest risk metric — what investors actually felt during the worst stretch.
Maximum drawdown (MDD) is the largest peak-to-trough percentage decline in an equity curve over a given period. It is the most behaviorally honest of the standard risk metrics — it measures what an investor actually lived through, not what a normal distribution implies.
Calculation
For an equity curve E_t:
Drawdown_t = (E_t − max(E_0..E_t)) / max(E_0..E_t)
MaxDrawdown = min(Drawdown_t) over the period
Always reported as a negative percentage (or its absolute value).
Why MDD matters more than σ
Volatility is a statistical abstraction; drawdown is a lived event. Two strategies with the same volatility can have very different drawdown profiles depending on the autocorrelation of returns. Trend-following strategies tend to have lower volatility but deeper drawdowns than their σ implies, because losses cluster.
See drawdown vs volatility for the full argument.
Drawdown components to track
- Magnitude — depth of the drop (the headline number)
- Duration — peak to recovery (often more painful than depth)
- Frequency — how often the strategy revisits worst-case territory
- Recovery factor — net profit ÷ |MDD|. Useful complement.
Common traps
- Single-point estimator. One observation; very sensitive to sample window.
- Closed-trade vs mark-to-market. They can differ by 2–3× for systematic futures strategies. Pick the more conservative.
- Conflating MDD with VaR. MDD is realized; VaR is forward-looking. They answer different questions.