Win Rate and Expectancy: Why 40% Can Beat 70%

Win rate is the share of trades that made money. Expectancy is what an average trade earned, wins and losses combined. Traders obsess over the first number; the second one pays the bills. The two are connected by one identity:

Expectancy per trade = (average win × win rate) − (average loss × loss rate)

Why win rate alone means nothing

A strategy that wins 70% of the time with $50 winners and $200 losers has an expectancy of (50 × 0.70) − (200 × 0.30) = 35 − 60 = −$25 per trade. It loses money while winning most of its trades — the classic profile of strategies with tight targets and wide stops. Meanwhile a trend-following system that wins 40% with $300 winners and $100 losers earns (300 × 0.40) − (100 × 0.60) = 120 − 60 = +$60 per trade. The market pays expectancy, not accuracy.

The break-even trade problem

Here is a subtlety most tools hide: what happens to trades that close at exactly $0 — break-even stops, scratched entries? If they count in the denominator, they deflate the win rate while costing nothing. A strategy with 30 winners, 30 losers, and 40 break-evens has a 30% "win rate" that behaves like 50%. Any honest backtest report should show you the full split — winners / break-evens / losers — and state which denominator its win rate uses. (ForexEdge shows the three-way split beside every win rate for exactly this reason, in backtests and on live results alike.)

What to check in any backtest

• Sample size. A 60% win rate over 20 trades is noise; over 500 trades it is evidence.
• Expectancy after costs. A +2-pip expectancy is wiped out by a 2-pip spread. Costs must be in the simulation.
• Consistency across stages. If the backtest says 65% but the same strategy runs 51% on an independent tester, the gap is your real finding — see why engines disagree.
• The companions. Read win rate with profit factor and maximum drawdown; each covers the others' blind spots.