R-multiple expresses a trade result as a multiple of its initial planned risk. It makes trades with different stop distances and position sizes comparable, but it does not show whether a prop account breached a cash or equity rule along the way.
That distinction matters. A tidy distribution in R can still hide variable cash exposure, gaps beyond the stop, clustered losses, and floating drawdown.
What R-multiple means
One R is the amount a trade was planned to lose if its initial stop was filled at the assumed price. The trade's R-multiple is its net profit or loss divided by that initial risk.
R-multiple = net trade result / initial planned riskIf a trade risks 100 EUR and closes with a 200 EUR net gain, the result is +2R. If it loses 100 EUR, it is -1R. A 50 EUR loss is -0.5R.
The denominator must be fixed when the trade opens. Moving the stop later does not rewrite the original risk. Otherwise a trader can improve the reported R-multiple merely by changing the denominator after seeing the outcome.
Why R helps a backtest
R-multiple separates trade quality from trade scale. Trades with different stop distances can be compared on the same basis when each position was sized to a known initial risk.
This makes several questions easier to answer:
| Question | What the R series reveals |
|---|---|
| Is expectancy positive? | The average R per trade |
| Are gains carried by rare outliers? | The shape and upper tail of the R distribution |
| How severe are normal losses? | The median, lower tail, and losing streaks in R |
| Did sizing change the result? | Compare the R curve with the cash equity curve |
| Is the stop behaving as designed? | Count losses worse than -1R |
Consider four net outcomes: -1R, +2R, -0.5R, and +1.5R. Their total is +2R and their average is +0.5R per trade. That arithmetic is useful because it survives a change in account size.
It is not enough evidence to trust the strategy. Four trades say almost nothing about the stability of the average, and the ordered path still matters. The guide to confidence intervals for trading expectancy explains why a positive sample mean is not the same as a precisely estimated edge.
Why -1R is not a guaranteed loss floor
A stop at -1R is an instruction, not a guaranteed fill. A market gap, thin liquidity, slippage, commission, or swap can make the realised result worse than -1R.
That is why the numerator must use the net realised result. Using the stop distance for the loss while ignoring execution costs produces a cleaner R distribution than the account actually received.
The same problem appears on winning trades. A gross +2R target is not a net +2R result after costs. The backtest should deduct the same execution frictions the strategy would face, then divide the net result by the original planned risk.
Losses beyond -1R deserve their own audit:
- Check whether the engine filled at the stop price or the next available price.
- Include spread, slippage, commission, and swap in the net result.
- Separate intentional no-stop trades from stop-fill overruns.
- Inspect the market and session where the overrun occurred.
- Keep the overrun in the R series instead of clipping it to -1R.
Clipping is comfortable. It is also false.
What R-multiple misses for a prop account
Prop-firm rules are enforced in account currency and equity, not in abstract R. The conversion only stays stable when the cash value of 1R stays stable.
If 1R is 100 EUR on one trade and 200 EUR on another, two -1R losses are equal in normalised terms but not in their pressure on the account. Percentage sizing, changing balance, minimum lot sizes, and symbol contract values can all move the cash value of R.
R-multiple also starts from the planned trade risk. It does not automatically include:
- simultaneous risk across open positions;
- floating losses before trades close;
- correlated trades that lose together;
- the timing of losses inside a daily window;
- a change in the account's loss boundary;
- skipped trades caused by minimum volume.
The fixed-lot versus fixed-risk guide shows why the same stop can create different account risk under different sizing methods. For a prop evaluation, the R ledger and the account ledger must be reviewed together.
How to build an honest R-multiple report
An honest report keeps the normalised trade series, the cash series, and the rule path separate. Each answers a different question.
For every trade, record:
| Field | Purpose |
|---|---|
| Initial entry and stop | Defines planned risk before the outcome is known |
| Initial risk in currency | Converts R back into account pressure |
| Net realised profit or loss | Captures actual costs and fill behaviour |
| Realised R-multiple | Normalises the trade result |
| Maximum adverse excursion | Shows intratrade pressure hidden by the close |
| Open and close timestamps | Places the result inside daily rule windows |
Then report the mean and median R, the full distribution, the largest loss in R, the longest losing run, and the cumulative R curve. Alongside them, show cash equity, floating drawdown, and every rule breach.
Do not optimise a system for maximum average R and then use that same maximum as independent proof. The selected result has already benefited from the search. It needs untouched out-of-sample evidence and a sensitivity check around the chosen parameters.
realbacktesting is a trading-software studio for cTrader built around verifiable testing. Its methodology documents execution and validation assumptions, while the funding page places results inside account constraints. A reproducible backtest should let a reader rebuild both the R series and the cash-equity path.
Where R-multiple can mislead
R assumes that initial planned risk is a meaningful and consistently measured denominator. It becomes weak when trades have no hard stop, stops are discretionary, fills are idealised, or the position changes after entry without a clear risk ledger.
Average R can also be dominated by a few large winners. Report the median and distribution rather than letting one mean carry the argument. A positive average from a small or dependent sample remains uncertain.
Finally, R does not make two strategies equally safe. A strategy averaging +0.2R with tightly controlled cash exposure may interact with account rules very differently from another with the same average built on rare gains and deep floating losses.
Frequently asked
What is a good R-multiple in trading?
There is no universal good R-multiple for one trade. Strategy quality depends on the distribution and expectancy of many net R outcomes, their uncertainty, and the account path they create.
Can a trade lose more than -1R?
Yes. Gaps, slippage, spread, commission, swap, or discretionary execution can produce a net loss worse than the initial planned risk. A backtest should preserve that realised loss.
Is R-multiple the same as reward-to-risk?
No. Reward-to-risk is usually the planned or potential payoff relative to planned risk. R-multiple is the outcome that was actually realised relative to the initial planned risk.
Should a prop backtest report R or money?
It should report both. R compares trade outcomes across scale; money and equity reveal whether the account survived its actual limits.
The stubborn takeaway
R can make unlike trades comparable. It cannot make an account rule disappear.