Win rate gets far more attention than it deserves in trading discussions. Drawdown — a less exciting, far more important number — gets far less than it deserves.
What drawdown actually measures
Drawdown is the decline from a peak in account equity to a subsequent low point, before a new peak is reached. If an account grows from $10,000 to $12,000, then falls back to $10,800 before recovering, that's a $1,200 drawdown — a 10% decline from the peak — regardless of the fact that the account is still up overall from its starting point.
Maximum drawdown specifically refers to the largest such decline over a given period — the worst peak-to-trough stretch a strategy has actually experienced, not an average or a typical case.
Why drawdown matters more than win rate
A strategy can have an excellent win rate and still have severe drawdown, if its occasional losses are large relative to its frequent small wins. Conversely, a strategy with a modest win rate can have very controlled drawdown, if losses are consistently kept small relative to account size. Win rate alone tells you almost nothing about how much pain — financial or psychological — you'd actually experience holding the strategy through a real losing stretch.
Drawdown is also a far better predictor of whether a trader will actually stick with a strategy long enough to realize its long-term edge. A strategy with a strong long-term return but severe, unpredictable drawdown is far more likely to be abandoned mid-drawdown — right before it recovers — than a strategy with a smoother, more controlled equity curve, even if the smoother strategy's total return is lower.
What genuinely controls drawdown
Position sizing, more than anything else. A strategy risking 1% of account equity per trade will, by mathematical necessity, produce dramatically smaller drawdowns than the same win/loss pattern risking 5% per trade — the underlying signal quality is identical, but the position sizing determines how much that signal quality translates into account-level pain.
A hard cap on consecutive losses. A circuit breaker that stops trading after a defined number of same-day losses directly limits how deep a single bad stretch can go, rather than allowing a losing streak to compound uninterrupted.
No martingale or loss-averaging. As covered in why martingale EAs blow accounts, any strategy that increases risk after a loss is specifically designed to produce a smooth equity curve most of the time at the cost of catastrophic drawdown eventually — the opposite of genuine drawdown control.
Diversification across genuinely independent setups, where applicable — though this matters less for a single-instrument strategy and more for a system trading multiple, meaningfully uncorrelated instruments.
How to actually evaluate drawdown before trusting a strategy
Look at maximum drawdown specifically, not just average performance — the worst historical stretch is a far more honest indicator of what you should expect to experience eventually than a smoothed average. Ask whether the drawdown figure comes from real trading data or an idealized backtest, since (as covered in can you trust backtests) a backtest's drawdown figure is exactly as manipulable as its return figure.
QMS Trading's homepage reports real maximum drawdown computed directly from actual, automatically-logged trade history — not a backtest assumption — specifically because an honest drawdown number matters more than an impressive-looking one.