A planned reward-to-risk ratio compares potential reward with planned loss, while win rate measures how often a defined process wins. Neither number proves profitability by itself. The relationship must be tested with realised outcomes and costs.
Calculate the planned ratio
For a long setup, price risk is entry minus stop and potential reward is target minus entry. Divide reward by risk, then confirm that spread, slippage and charges do not materially change the relationship.
Estimate the break-even win rate
Ignoring costs, a 1:1 reward-to-risk relationship needs a win rate above 50% to create a positive average outcome, while a 1:2 relationship has a lower mathematical break-even rate. Realised losses and gains rarely match the plan perfectly.
- Average realised gain
- Average realised loss
- Win rate
- Trading costs
Watch for unrealistic targets
A distant target may create an attractive ratio on paper but have a low probability of being reached within the chosen horizon. Reward should come from market structure, not from selecting a number to make the ratio look better.
Measure expectancy across a useful sample
Track the average result per decision over a consistent set of rules. A small sample or selective record can make win rate and payoff appear more stable than they are.
Calculate expectancy from realised outcomes
For a consistent sample, multiply win rate by average realised gain and subtract loss rate multiplied by average realised loss. Include charges and classify scratch or partial outcomes consistently.
Compare planned and realised ratios. Early exits, slippage, gaps and missed targets can make the strategy's actual payoff very different from the ratio displayed before entry.
- Win rate
- Average realised gain
- Average realised loss
- Net cost per decision
Test sensitivity instead of relying on one average
Recalculate expectancy with a slightly lower win rate, smaller average gain and larger average loss. If a small change removes the edge, the process may be fragile or the sample may be too small.
Use the result to define review questions, not to guarantee future profitability. Market conditions, execution and behaviour can change after the historical sample.
- Base case
- Lower win-rate case
- Higher-loss case
- Minimum acceptable sample