Most traders review their results in dollars. It feels natural, and it is almost useless.
A $500 win means nothing on its own. Did you risk $100 to make it, or $2,000? Was it a clean setup that hit target, or a reckless size that happened to work? Dollars cannot tell you. R-multiple can.
What R actually is
R is your risk on a single trade — the distance between your entry and your stop loss, multiplied by your position size.
If you enter at 7516.75 with a stop at 7510.50 on 3 MES contracts:
- Distance = 6.25 points
- Point value = $5 per contract
- R = 6.25 × 3 × $5 = $93.75
That $93.75 is one unit of risk. Everything else is measured in multiples of it.
Calculating R-multiple
R-multiple = (Exit − Entry) ÷ (Entry − Stop) for a long trade.
Continuing the example above, if you exited at 7528.67:
- Reward = 7528.67 − 7516.75 = 11.92 points
- Risk = 7516.75 − 7510.50 = 6.25 points
- R-multiple = 11.92 ÷ 6.25 = +1.91R
Notice what disappeared: contract size, point value, and the dollar amount. The multiplier cancels out of both sides. That is the entire point.
Why this makes trades comparable
Consider two trades:
| Trade A | Trade B | |
|---|---|---|
| Market | MES futures | EUR/USD |
| Profit | $180 | $340 |
| Risk taken | $90 | $850 |
| R-multiple | +2.0R | +0.4R |
In dollars, Trade B looks nearly twice as good. In R, Trade A was five times better. Trade B risked a fortune to scrape a small return — the kind of trade that looks fine in a P&L column and destroys accounts over time.
R-multiple is the only unit that lets you compare a futures scalp to a swing trade in currencies, or this month to last month after you changed your position size.
What your average R tells you
Once you have a real sample, your average R per trade (expectancy in R) is arguably the single most useful number you own.
- Negative average R — the strategy loses money regardless of win rate
- 0 to +0.2R — barely breakeven after costs
- +0.3R to +0.5R — a solid, sustainable edge
- Above +1.0R — excellent, and worth checking whether your sample is large enough to trust
Here is the useful part: average R and win rate trade off against each other. A trend strategy might win 35% of the time with an average of +0.6R. A mean-reversion strategy might win 70% with +0.25R. Both work. Neither is judged by win rate alone.
The mistake that quietly ruins the number
Moving your stop loss.
R-multiple assumes your initial risk was real. The moment you widen a stop because price is going against you, the denominator you planned with no longer describes what you actually risked. Your calculated R becomes fiction — and always fiction in the flattering direction.
This is why the number only means something if you log your planned stop at entry and leave it there. A journal that lets you edit the stop after the fact is a journal that will lie to you.
Moving a stop to breakeven after taking partial profit is a different thing and it is fine — the initial risk was still genuinely taken. What corrupts the number is widening the stop to avoid being wrong.
Partial exits and R
If you scale out at multiple levels, R-multiple is calculated from the weighted average exit price, not from each exit separately.
Three contracts, one closed at TP1 and two at TP2:
- Weighted average exit = ((TP1 × 1) + (TP2 × 2)) ÷ 3
- Then apply the standard formula against your original entry and stop
Getting this wrong is common. Averaging the R-multiples of each exit instead of the prices produces a number that is close but consistently off — and the error grows with uneven position sizes.
How to actually use it
Log three things per trade and review them weekly:
- R-multiple — was the reward worth the risk
- Whether you followed your plan — was the result repeatable or accidental
- The setup name — so you can group and compare
Then look for the pattern. Most traders discover something uncomfortable and valuable: one or two setups carry the entire edge, and the rest are noise that costs money. You cannot see that in dollars. In R, it becomes obvious in a single sorted column.
The short version
Dollars measure outcome. R measures decision quality. Only one of them is under your control, and only one of them predicts what happens next.