What it means
Expectancy is the mathematical expected value of a single trade: (win_rate × average_win) - (loss_rate × average_loss). Expressed in R-multiples, it tells you how many R you should expect to make on average per trade. Positive expectancy = profitable system; zero or negative = unprofitable. Combined with trade frequency, expectancy determines annual returns: total_R_per_year = expectancy_per_trade × trades_per_year.
Why it matters
Most retail traders evaluate their strategies on win rate alone — a fatal mistake. A 70% win-rate system with average wins of +0.5R and losses of -1R has NEGATIVE expectancy (0.7×0.5 - 0.3×1 = +0.05R, marginal). A 40% win-rate system with +3R wins and -1R losses has STRONG expectancy (0.4×3 - 0.6×1 = +0.6R). Without expectancy, you can be very successful at being wrong.
How to use it
After 30+ trade samples, compute expectancy. If positive and consistent across sub-samples, the system has edge. If negative, stop trading it — variance can make a losing system look temporarily profitable. To improve expectancy: raise win rate (better entries), raise average win (better exits), or lower average loss (tighter stops). Each has tradeoffs.
100-trade sample: 45 wins averaging +1.8R, 55 losses averaging -1R. Expectancy = (0.45 × 1.8) - (0.55 × 1) = 0.81 - 0.55 = +0.26R per trade. At 200 trades/year, expected annual return = 52R. With 1% risk per trade, that's ~52% per year on capital — before any compounding or position growth.
The win-rate vs payoff tradeoff
Expectancy can be the same across very different win-rate/payoff combinations. Examples with +0.4R expectancy: (a) 70% win × +1R, 30% × -0.7R; (b) 50% × +1.4R, 50% × -0.6R; (c) 35% × +2.4R, 65% × -0.7R. Psychologically, traders find (a) easier (more wins) and (c) harder (more losses, but bigger wins). The math is the same; the experience is very different. Match the system to your psychology.
Sample size and expectancy stability
Single trades reveal nothing about expectancy. 30 trades: large noise, but indicative. 100 trades: expectancy starts stabilizing. 500+ trades: expectancy reflects true edge. Below 30 trades, do NOT make strategic decisions about whether your system 'works' — you're observing variance, not edge. Most retail traders make this error: 5 winning trades = 'I have a system'; 5 losing trades = 'system is broken.'
Frequently asked
What if my expectancy is negative?
Stop trading the system until you understand why. Common causes: (1) entry rules are noise (no statistical edge); (2) exit rules give back too much (small wins, big losses); (3) trading the wrong instrument (e.g., a momentum system on a mean-reverting asset). Backtest changes BEFORE applying to live capital.
Can a high-win-rate strategy have negative expectancy?
Yes, easily. 90% win rate with average wins of +0.1R and average losses of -1R: expectancy = 0.9×0.1 - 0.1×1 = -0.01R. Slightly negative. Win rate alone tells you nothing about profitability.
How do I improve expectancy?
Three levers: (1) better entries (raise win rate without changing exits); (2) better exits (let winners run longer, raise average win); (3) tighter stops (lower average loss). Most improvements come from #2 and #3, not #1 — exit discipline is where amateur traders lose the most ground.
Is expectancy stable over time?
It drifts as market regimes change. A trend-following system with +0.5R expectancy in 2017-2021 (steady trend) might have -0.2R expectancy in 2022-2023 (choppy). Re-evaluate expectancy on rolling samples (last 50-100 trades) to detect regime shifts before the drawdown becomes catastrophic.
Want a worked example or a deeper dive? Ask Rocky how this concept applies to your specific watchlist or trade idea.
Ask Rocky