Recovery factor is your net gain divided by your maximum drawdown. It answers the question the gain alone never asks: that result, at the cost of what decline did you get it?
From peak to trough
Contents
An annual gain cannot be judged alone, and recovery factor is the simplest way of saying so in one number. It divides what you earned by the worst decline you went through to get there. Making 20% with a 5% decline and making 20% with a 40% decline are not the same performance, and this ratio separates them immediately.
Its interest is using observed risk rather than declared risk. Percent of capital risked describes an intention; maximum drawdown describes what actually happened. So recovery factor is built on the only risk measure a trader cannot flatter.
It is also a unitless figure, so comparable across accounts and methods, which makes it one of the few indicators a funder looks at first. A system with a high recovery factor is fundable: it returns a lot for little decline.
Recovery factor = net P&L ÷ maximum drawdown
Both terms are taken over the same period. An annual gain divided by a monthly drawdown produces a flattering figure that means nothing.
The drawdown in the denominator is the one measured from a peak, not from starting capital. That point decides the result more than the numerator does.
Over a losing period, the ratio is negative and no longer means anything. Recovery factor only reads on a period whose P&L is positive.
| Recovery factor | What it describes |
|---|---|
| under 1 | The worst decline exceeds the period's gain. Profitable, but very hard to hold. |
| 1 to 2 | Common. Gain and decline are of the same order of magnitude. |
| 2 to 3 | Good. The system returns clearly more than it makes you go through. |
| 3 to 5 | Very good, and that is the order of magnitude a funder looks at. |
| above 5 | Rare on a long track record. Check the period is not too short to contain a real decline. |
The last row is the check to run first. A very high recovery factor almost always comes from a drawdown that is too small because the period is too short: over three calm months, the denominator had no time to be tested. The figure is correct and the conclusion false, and it is the same flaw as on profit factor.
The second point is the blind spot of time. This ratio does not distinguish a $1,800 decline crossed in three weeks from an identical one spread over eight months. The first is an accident, the second is a whole year of doubt, and it is the second that makes people abandon a perfectly profitable method.
A recovery factor of 1.17. The account earned barely more than it lost in its worst sequence, which describes a profitable and gruelling year. The same gain with a $600 decline would give 3.5, and that would be an entirely different year lived for exactly the same final result.
This is the use this ratio exists for. Two methods with the same annual return do not compare until you know what they made you go through. Recovery factor puts both on the same scale in one division.
It is also the comparison a funder makes first, before even looking at the return. A system returning 15% with a 3% decline is more interesting than one returning 40% with a 35% decline.
Reducing risk per trade reduces gain and drawdown in roughly the same proportion, so recovery factor barely moves. That is precisely what makes it useful: it isolates system quality from sizing.
If it moves clearly after a size change, something else changed at the same time, and it is worth finding what.
The first flaw is the period being too short. A maximum drawdown is only reliable over a long history; over three months it describes what happened, not what can happen. A recovery factor of 8 over a calm quarter will become 2 at the first real bad run, with the system unchanged.
The second is the absence of time. The ratio treats a short decline and a long one exactly alike, while duration is what decides abandonment. Two systems with the same recovery factor can be one workable and one unliveable.
The third is the badly computed denominator. If drawdown is measured from starting capital instead of from the peak, it is heavily understated, and recovery factor inflates by as much. It is the same error as on drawdown itself, and it propagates here silently.
The classic mistake. Celebrating a high recovery factor over a short period. The denominator is a maximum drawdown, and a maximum drawdown only means something over a long history: over a calm quarter it simply has not been tested yet. The figure is correct, and it describes a period, not a system.
| Indicator | Formula | What it measures | When to use it |
|---|---|---|---|
| Recovery factor | net P&L ÷ maximum drawdown | Gain against the worst decline endured | Comparing two systems at unequal risk |
| Maximum drawdown | (peak − trough) ÷ peak | The decline alone, without the gain | Sizing positions |
| Sharpe ratio | mean ÷ standard deviation | Steadiness, not depth | Judging the stability of a performance |
| Profit factor | Σ gains ÷ |Σ losses| | The ratio between two totals | Judging a system trade by trade |
A high recovery factor over three months describes a calm quarter, not a good system. The denominator has not yet been tested, and it will be. That check takes a second and avoids concluding on a sample that went through nothing.
A recovery factor below 1 means your worst decline exceeds your annual gain. Reducing risk per trade reduces both, so it barely changes the ratio, but it makes the period liveable, which is the condition for it repeating.
That is the information this figure does not contain and nothing else gives. A six-month decline and a three-week one give the same recovery factor and two unrelated experiences. One line per quarter is enough.
It is your net P&L divided by your maximum drawdown over the same period. It says how much you earn per unit of worst decline endured, and it immediately separates two traders showing the same annual gain with very different trajectories.
Between 2 and 3 is good, between 3 and 5 is very good and matches what a funder looks at. Below 1, your worst decline exceeds your gain for the period: that is profitable and very hard to hold. Above 5, first check the period is long enough to contain a real decline.
Both divide a gain by a risk, but not the same one. Recovery factor takes the worst decline, Sharpe takes the spread of results. A very steady system suffering one brutal fall will have a good Sharpe and a bad recovery factor, and both figures will be right.
Barely, and that is precisely its interest. Reducing risk per trade reduces gain and drawdown in roughly the same proportion, so their ratio stays stable. That is what allows using it to judge a system's quality independently of its sizing.
Time. A $1,800 decline crossed in three weeks and the same one spread over eight months give exactly the same recovery factor, while the first is an accident and the second a whole year of doubt. Duration is what makes people abandon a profitable method, and it is in none of these ratios.
Recovery factor is a metric in the Reports screen, computed from net P&L and from maximum drawdown measured from each peak of your equity curve. That last point matters: measured from the starting balance, the denominator would be heavily understated and the ratio would inflate silently.
It reads next to max drawdown and the Sharpe ratio in the same table, and it is that trio which describes a trajectory: the depth, the steadiness, and the gain against the depth. The duration of the decline is in none of the three, and it is the one thing to log by hand.
To go deeper on this: read the full guide. Every term is gathered in the glossary.