Average MFE in R is each trade's MFE divided by the risk it carried, then averaged. It answers a question MFE in currency cannot ask: was the best moment large relative to what you were risking?
What the ratio compares
Contents
MFE in currency has a flaw R fixes: it depends entirely on position size. An MFE of $312 on a trade risking $100 and one of $312 on a trade risking $900 do not tell the same story, and the average of the two describes nothing.
MFE in R brings each best moment back to the risk unit of the trade that produced it. An MFE of 2.1R means the market went twice as far in your favour as the distance to your stop, whatever the amount committed. That compares across trades, across markets and across accounts.
Its reading only makes sense against average R banked. The first says what the market offered in risk units, the second what you kept of it. The gap between them is the information, and neither contains it alone.
MFE in R = trade MFE ÷ |entry − stop| — Average MFE in R = mean over trades with a stop
The risk is the one at opening, as for any R calculation: moving the stop along the way does not change the denominator.
A trade with no stop has no MFE in R. It is excluded, and you need to know how many trades that represents before reading the average.
MFE in R and the RRR you aimed for are not the same thing. The first is what the market offered, the second what you had planned to target.
| Average MFE in R | What it describes |
|---|---|
| under 1R | The market almost never exceeds your risk. The problem is the entry or the market chosen. |
| 1 to 2R | Ordinary. What you keep of it decides everything. |
| 2 to 4R | The move is there. If your banked R stays under 1, the exit is what costs. |
| above 4R | Very tight stops, or very mobile markets. Check first that the logged stops are the real ones. |
The last row meets the R-multiple trap, for the same reason: a denominator that is too small inflates everything. An average MFE of 6R almost always comes from a logged stop tighter than the one held, not from an exceptional market. The check is the same, compare average loss with declared risk.
The rest of the table always reads in pairs. An MFE of 3R with a banked R of 1.2 is a good system leaving money behind; an MFE of 3R with a banked R of 0.1 describes someone cutting almost immediately. The first figure is identical and the conclusion opposite.
The market offered 2.1 times the risk and you keep 0.8. Expressed in R, that finding compares directly with another trader's, another market's or another year's, which the currency equivalent never allows.
This is what R allows and currency forbids. A scalper and a swing trader do not deal in the same amounts, but both can say how many times their risk the market offered them, and how much they kept.
That comparability mostly serves you against yourself over time. An MFE in R staying stable while capital changes says the market still offers as much; the same measure in currency would have moved for nothing.
A target is set in risk multiples. If your average MFE is 2.1R, a 3R target is only rarely reached, and the figure says so before the test rather than after three months.
It does not replace the test, because moving the target changes how many trades reach it. But it eliminates targets the market never offers, and that is already a lot.
The first flaw is inherited from R: it silently excludes trades with no stop. An average MFE in R computed over 55% of your activity is not wrong, it describes 55% of your activity, and almost no tool writes that percentage alongside.
The second is confusion with the RRR you aimed for. Many read 'MFE 2.1R' as 'my trades go to 2.1 times risk', so as a promise of a target. It is a maximum reached at an instant, not a level where you could have calmly exited, and the nuance is worth several months of pointless tinkering.
The third is looking at it without banked R. Alone it says nothing: the gap between the two is the information, and a high MFE can describe an excellent system just as well as a trader cutting everything after three minutes.
The classic mistake. Taking average MFE in R for a reachable target. It is a maximum touched at an instant, often a spike, not a level where you could have calmly exited. A target set at average MFE is only reached by the half of trades exceeding that average, that is by very few.
| Indicator | Formula | What it measures | When to use it |
|---|---|---|---|
| Average MFE in R | MFE ÷ risk taken | What the market offered, in risk units | Comparing an exit across styles |
| MFE | best point − entry | The same thing in currency | Quantifying a gap on a given account |
| Average R | result ÷ risk taken | What you kept, in risk units | Sitting opposite MFE in R |
| RRR | average win ÷ average loss | What you aimed for, not what was offered | Setting targets |
Both figures in the same unit, side by side. Their gap is the only information, and each taken alone lends itself to two opposite readings. It is the most profitable display setting in this group of terms.
An MFE in R computed over half your trades describes half your activity. Note the percentage of trades with a stop next to the figure, once, and you will know how much the average speaks about you.
A target beyond average MFE is only reached by the minority of trades exceeding the average. That is not forbidden, but you need to know it beforehand, not after watching your win rate collapse.
It is each trade's best moment divided by the risk that trade carried, then averaged. An MFE of 2.1R means the market went in your favour twice as far as the distance to your stop.
Because MFE in currency depends on position size. The same move gives $312 on one trade and $31 on another, and the average of the two describes nothing. In R, the figure compares across trades, across markets and from year to year.
With average R actually banked, in the same unit. 2.1R offered for 0.8R taken means 62% of the available move is given back. It is the gap that informs, neither figure alone.
It is risky. An average is exceeded by only half the trades, and MFE is often a spike touched for an instant rather than a level held. A target set there will be missed far more often than imagined, and it is better tested over thirty trades before believing in it.
They are excluded, as for any measure in R: with no stop there is no defined risk for the denominator. What matters is knowing what share of your activity that represents, because a figure computed over half your trades describes only that half.
Tradoshi computes average MFE in R as soon as a trade has a known stop and its historical candles could be downloaded. It sits in the Reports screen, in the same metric group as MFE in currency and MAE in R.
The setting that makes this term useful takes one action: display in the same table average MFE in R and realised average R. Both columns are in the same unit, the gap reads without calculation, and it is exactly the pair described above. The Tag dimension then shows on which setups the gap widens.
To go deeper on this: read the full guide. Every term is gathered in the glossary.