Risk/Reward Ratio (R:R)
Ratio of potential loss (1R = stop distance) to potential gain. Together with the win rate it determines profitability.
Definition
R (risk) is the loss on a stop hit. All gains are expressed in multiples of R.
- Stop at 100 USD, entry at 102 USD → 1R = 2 USD.
- Take-profit at 96 USD → +2R (twice the gain relative to risk).
The R:R ratio of a trade is target distance / stop distance.
The relationship with win rate
The break-even win rate depends on the R:R:
| R:R | Break-even win rate | Comment |
|---|---|---|
| 1:1 | 50% | coin flip, any improvement is edge |
| 1:2 | 33.3% | classic trend-following profile |
| 1:3 | 25% | many losses, but the wins compensate |
| 2:1 | 66.7% | high-win-rate strategies (mean reversion) |
Expectancy formula:
E = (win_rate * avg_win) - ((1 - win_rate) * avg_loss)
As long as E > 0, the strategy is profitable in the long run — regardless of the win rate.
R:R vs. Calmar
Both measure risk efficiency, but on different levels:
- R:R — per trade. A tactical quantity.
- Calmar — per strategy, across many trades. A portfolio quantity.
A strategy with a high per-trade R:R can still have a poor Calmar if the win rate is too low for that R:R.
How Botty uses it
- Initial stop via
ATR_STOP_MULTIPLIER(e.g. 2 × ATR) defines 1R. - Trailing stop leaves gains open-ended, no fixed R target.
- Reporting in
reporting/computes average win/loss in R units for every experiment.
Beginner trap
"At least 3:1 R:R on every trade" is not a universal law — it fits trend following (Raschke, Weinstein), but not mean-reversion systems with a 70% win rate and R:R = 0.7. The combination of win rate × R:R is what counts, not the single number.