Expectancy, explained

By the XAUUSDLiveChart Research Desk · 5 min read

Two traders argue: one wins 70% of trades, the other 35%. Who earns more? Unanswerable, because win rate is half a number. Expectancy completes it: (win rate x average win) minus (loss rate x average loss), the expected profit of one trade. Positive, and time works for you; negative, and no discipline can save the system. Everything else in strategy evaluation is decoration around this.

The formula in R terms

Measure results in R, multiples of the amount risked, and expectancy reads cleanly: a 40% win rate with average 2R winners and 1R losers yields 0.4x2 - 0.6x1 = +0.2R per trade. Forty small numbers into one honest one. Our own engine backtests report exactly this figure because it is the only summary that cannot hide behind a pretty win rate.

The two levers

Expectancy improves by winning more often or winning bigger relative to losses, and the levers trade against each other: tighter targets raise win rate while shrinking wins, wider stops raise win rate while growing losses. Chasing either lever in isolation, usually win rate, because losing feels bad, is how traders optimise their way into negative expectancy with excellent statistics.

Sample size decides trust

Expectancy from twelve trades is noise wearing a suit; streak luck dominates small samples completely. Meaningful estimates start around a hundred comparable trades, and even then a confidence band belongs around the number. This is why journaling exists: it is expectancy data collection.

Costs live inside it

Spread, slippage and swap subtract directly from every trade's R before expectancy is computed. A raw +0.15R edge with 0.1R average costs is a +0.05R business, and a fragile one. Cost-free expectancy claims are the first thing to distrust in any published result, including your own backtests.

FAQ

← All guides