Maximum drawdown (MDD) is the largest peak-to-trough percentage decline of a portfolio over a period. If a portfolio grew to $120k and then fell to $84k, the drawdown was −30% — regardless of whether it recovered later.
Why it's the most useful risk measure
Average return and volatility are abstract. MDD answers the question you actually feel: "how much would I be down if I had entered at the worst moment?" It's also a psychological measure — most investors abandon a strategy near the bottom of a drawdown, converting a paper loss into a permanent one.
The arithmetic of recovery is brutal
- −10% needs +11% to break even,
- −30% needs +43%,
- −50% needs +100%,
- −80% needs +400%.
That's why defensive designs (cash-stepping dual momentum, or volatility-inverse weighting as in Inverse-Volatility) can win long-term despite lagging in bull markets — they lose less, so they have less to claw back.
Caveats when comparing MDD
MDD depends on the window (a strategy with no bear market in its history looks deceptively safe), on data frequency (daily data shows deeper troughs than monthly), and on backtest honesty. Every track record here shows MDD computed on daily data, costs included — methodology here, side-by-side comparison here.
Educational material — not investment advice.