A trade's R-multiple is its result divided by the risk it carried, that is the distance between entry and stop. A trade returning three times what it risked is worth +3R; a stopped-out trade is −1R. Average R is the average of those values across every trade with a known stop.
The levels of a trade
Contents
The R-multiple answers a question the result in currency never asks: did this trade pay well relative to what it put at stake? A $300 gain is excellent if it risked $100, and mediocre if it risked $900. R makes that comparison immediate by taking risk as the unit.
The letter R stands for risk, and a trade's R reads as a multiple of that risk. The important point is that −1R is fixed: it is what a trade going to its stop costs, by definition. Everything else is measured against that marker, which gives a stable scale where currency depends on capital, market and leverage.
Average R is the mean of those multiples across every trade with a known stop, winners positive and losers negative. It is expectancy expressed in units of risk rather than in currency, and that is what makes it the most portable statistic from one journal to another.
Trade R = net result ÷ risk taken — Risk taken = |entry − stop| × position size
The risk taken is the one that existed at opening. Moving your stop during the trade does not change R, otherwise the measure would be circular.
The result is taken net, fees included, for the same reason as everywhere else: an R computed on gross is systematically flattered.
A trade with no stop has no R. It does not have a bad one, it has none, and forcing it in by assuming a stop manufactures a statistic that does not exist.
| Average R per trade | What it means |
|---|---|
| negative | The system loses in risk units. Trading bigger only speeds the loss up. |
| 0 to 0.1 | Break-even. Fees and slippage alone can tip the period over. |
| 0.1 to 0.3 | Decent, and common among profitable traders. Volume then makes the income. |
| 0.3 to 0.5 | Very good, provided the sample exceeds a hundred trades. |
| above 0.5 | Rare on real history. First check that the recorded stops are the real ones. |
The last row deserves to be taken seriously, because it is the most frequent way of fooling yourself. An abnormally high average R almost always comes from a recorded stop tighter than the one actually used: the risk in the denominator is smaller than the risk carried, and every multiple inflates. The check is one comparison: the average loss on your losing trades should be close to the planned risk.
The second point concerns trades with no stop. They are excluded from the calculation, which is correct, but it means a flattering average R can coexist with a handful of catastrophic trades that do not appear in it. So average R is always read next to the share of trades with a stop: at 60%, it describes a little over half your activity.
This trade is worth +3R, and it would be worth +3R on a $500 account as on a $500,000 one. That is exactly what the measure brings: the figure describes the decision, not the capital that executed it.
This is the main contribution. A gold trade and a currency pair trade, a 2024 trade on a small account and today's trade on an account ten times bigger: in currency those four figures do not compare. In R, they do.
It is also what makes the measure useful for judging progress. An average R rising while capital changes says something about the decision; a P&L rising while capital changes says nothing at all.
Average R multiplies. At 0.2R per trade, 20 trades a month and 1% of capital risked per trade, monthly expectancy is 0.2 × 20 × 1% = 4% of capital. It is a rough calculation, and it is the only one linking a trading statistic to an income.
It is mostly useful for seeing what is missing. The same formula shows that a decent average R with three trades a month produces nothing, and that it is frequency, not quality, that needs work.
The most widespread flaw is confusion with RRR, encouraged by the fact that both are often written with the same letter. RRR compares average win to average loss, in currency, without ever looking at your stop; the R-multiple relates each trade to its own stop. A large gain with a wide stop gives a flattering RRR and a mediocre R, and that is the difference between a result and money management.
The second is the declared stop being tighter than the one actually held. Nobody does it on purpose, and yet it is frequent: you log the plan's stop, you let it run a little when the moment comes, and the denominator becomes a fiction. Average R rises without any decision having improved.
The third is the silent exclusion of trades with no stop. Most tools compute an average R over eligible trades without saying how many they dropped. A statistic built on 60% of your activity is not wrong, but it does not describe what you think it describes.
The classic mistake. Logging the plan's stop rather than the stop you held. The risk in the denominator then becomes smaller than the risk actually carried, every multiple inflates at once, and average R improves without any decision having changed. It is the quietest way of lying to yourself with this statistic.
| Indicator | Formula | What it measures | When to use it |
|---|---|---|---|
| Average R (SL) | mean of (result ÷ risk taken) | What a trade returns, in risk units | Comparing different trades and accounts |
| RRR | average win ÷ average loss | How winners weigh against losers | Setting targets, without looking at stops |
| Expectancy | net P&L ÷ trade count | What an average trade returns, in currency | Estimating income at constant capital |
| Profit factor | Σ gains ÷ |Σ losses| | What you earn per $1 lost | Judging a whole system, or one setup |
Compare the average loss on your losing trades with the risk you claim to take. If the former exceeds the latter, a handful of trades where the stop got pushed explain almost the whole gap, and your average R is inflated by as much. It is a one-minute check and it is worth more than any optimisation.
Average R only speaks for trades with a stop. Note next to it what share of your activity that is. An average R of 0.3 over 95% of trades and the same over 50% do not describe the same person, and only the second is hiding something.
An average R of 0.2 over thirty trades a month returns more than an average R of 0.5 over five. Many traders try to improve the quality of their entries when their problem is the number of opportunities, and the R × frequency × risk formula shows it in one line.
It is a trade's result expressed in multiples of the risk it carried. If you risked 30 pips and gained 90, the trade is worth +3R. A trade that goes to its stop is −1R by definition, and it is that fixed marker that makes the scale usable.
Divide the trade's net result by the risk taken at opening, that is the entry-to-stop distance times the position size. The risk is the one at the start: moving the stop along the way does not change it, otherwise the measure would go in circles.
Between 0.1 and 0.3 per trade is decent and common among profitable traders; above 0.5 on real history, first check that the recorded stops are the ones actually held. The figure matters less than its multiplication by your frequency: 0.2 over thirty trades beats 0.5 over five.
RRR compares the average size of your winners to your losers, in currency, without ever looking at your stops. Average R relates each trade to its own stop, so it measures your money management. A large gain obtained with a wide stop gives a good RRR and a mediocre R: both describe the same series and not the same thing.
It is excluded, and that is correct: with no stop, there is no defined risk to put in the denominator. The problem is that this exclusion is silent in most tools. An average R computed over 60% of your trades is not wrong, it simply describes 60% of your activity, and that is information to display alongside.
Tradoshi computes a trade's R as soon as a stop is known, and it takes that stop directly from what your broker returns, not from a manual entry. Average R appears in the Reports screen, next to win rate, expectancy and maximum drawdown, and it recomputes on every sync.
The check described above is directly readable: the same screen gives the average loss and the share of trades with a stop. If the average loss exceeds your planned risk, or if coverage is low, the average R displayed describes less than it appears to, and those two figures sit two lines apart.
To go deeper on this: read the full guide. Every term is gathered in the glossary.