Maximum drawdown — usually written max DD in trading vocabulary — measures the worst losing run a strategy has experienced. If equity peaks at $10,000 and subsequently falls to $8,500 before reaching a new high, the max drawdown over that window is 15%.
Why it matters more than return
A strategy that returns 30% per year with a -10% max DD is a fundamentally different product from one that returns 30% per year with a -50% max DD, even though the headline number is identical. The first is tolerable through real-world losing streaks; the second wipes out most retail accounts because the trader either runs out of capital or quits at the bottom.
For automated trading specifically, max DD answers the operational question: how much account capital must I keep available? A common heuristic is to size the account at 2-3× the published max DD — so a strategy with -20% max DD requires a $6,000-9,000 account to deploy responsibly with a $2,000 risk budget.
How max DD is reported
Three forms appear in EA listings:
- Backtest max DD — measured over the historical data window. Honest reports state the start and end dates plus modelling quality.
- Out-of-sample max DD — measured on data the strategy was not optimised on. Almost always larger than in-sample max DD; the difference is a marker of curve-fitting.
- Live trading max DD — measured on real money since deployment. Most relevant but typically smallest sample size.
mt5depot’s EA pages publish backtest max DD with explicit data source and modelling quality. The next-most-honest variant is to publish all three side-by-side; this catalogue is moving toward that as the live trading sample lengthens.
Common pitfall
Max DD is path-dependent: it cares about the deepest drop, not the average drop. A strategy that has a single -25% drawdown in 5 years and never another loss looks worse on max DD than one that has 12 separate -8% drawdowns. The two have very different risk characters. Use max DD alongside Sharpe ratio, win-rate, and worst-streak metrics — not on its own.