Recovery factor: formula, what counts as good, and its limits
The recovery factor is net profit divided by maximum drawdown: how many times a strategy earned back its worst loss over the whole test. It has no single good value, it grows with test length, and it only means something when compared at the same test length and checked against the margin you would hold.
By Soren Huang · Published · updated · Editorial policy
Contents
- What is the recovery factor and how do you calculate it?
- What recovery factor counts as good?
- Recovery factor vs profit factor: what is the difference?
- Recovery factor vs Calmar ratio: what is the difference?
- What does the recovery factor look like in a real backtest?
- Why does a short sample make the recovery factor unreliable?
- Frequently asked questions
- References
What is the recovery factor and how do you calculate it?
The recovery factor is net profit divided by maximum drawdown. It tells you how many times a strategy earned back its worst peak-to-trough loss over the whole test. A recovery factor of 3 means the test made three times its deepest loss; a value below 1 means the strategy never fully earned that loss back.
Maximum drawdown is the largest fall of the equity curve from a high point to the lowest point that follows, before a new high is made. The equity curve is the running total of closed-trade profit plotted over time. Both numbers come from that curve, not from the trade list: you cannot see a drawdown by looking at trades one at a time, because it is built from a run of trades in a particular order. If you have the equity series as a list of numbers, the free drawdown calculator finds the deepest fall and the time it took to recover.
The single formula in this article:
recovery factor = net profit ÷ maximum drawdown
Here net profit is the total result of all closed trades after commission and slippage, from the first bar to the last, and maximum drawdown is the deepest fall of the equity curve measured in the same money units. Both are read from a backtest that replays the rules on old prices with costs included, so the ratio is only as honest as the costs in the run.
- Run the backtest with commission and slippage set. A drawdown measured without costs is too shallow.
- Read the net profit of closed trades at the end of the equity curve.
- Find the deepest peak-to-trough fall anywhere on the curve. It is often not the last one and often not the longest one.
- Divide the first number by the second. Write down the test length in years and the trade count next to the result; the later sections explain why those two numbers belong with it.
Hypothetical example
A five-year futures backtest ends with a net profit of $60,000 after costs. The deepest fall of its equity curve, from a high in year two to a low eight months later, is $20,000. Recovery factor = $60,000 ÷ $20,000 = 3.0. The strategy earned its worst loss back three times over the five years.
What recovery factor counts as good?
There is no single good recovery factor, because the same number means something different for a two-year test and a ten-year test. Net profit keeps adding up as the test gets longer, while the worst drawdown is one event. A recovery factor of 3 over two years is a stronger reading than 3 over ten years, where the strategy had five times as long to earn its loss back. Judge the ratio next to the test length in years and the number of trades, never alone.
Two readings are safe to make. A recovery factor below 1 means the test did not earn back its own worst loss, which is a strategy you would have stopped trading at the bottom of the drawdown. A recovery factor comfortably above 1 over several years and a few hundred trades means the worst loss was small next to what the rules produced over the whole period. Between those readings, compare only tests of the same length on the same market, and prefer the strategy whose ratio was reached with the shallower drawdown, because a drawdown is the number you have to fund with margin.
One thing that does not improve the recovery factor is size. Trading two contracts instead of one doubles every trade result, so it doubles net profit and doubles the max drawdown, and the ratio does not move. The same is true of leverage: a smaller account with the same contract count has the same equity curve in dollars and the same recovery factor, but the drawdown is a much larger share of the account. How margin and account size turn a dollar drawdown into a percentage you can or cannot survive is covered in how futures leverage changes backtest drawdown.
Hypothetical example
One contract: net profit $100,000, max drawdown $40,000, recovery factor 2.5. Two contracts, same rules, same bars: net profit $200,000, max drawdown $80,000, recovery factor still 2.5. The ratio says the two runs are the same strategy, which is correct. Whether an account can hold an $80,000 drawdown is a separate question the ratio does not answer.
Recovery factor vs profit factor: what is the difference?
Profit factor ignores the order of trades and the recovery factor depends on it. Profit factor is the sum of all winning trades divided by the sum of all losing trades. Shuffle the same trades into any order and the profit factor stays the same, because both sums are unchanged. The max drawdown, and with it the recovery factor, can change a great deal, because drawdown is a path: it depends on which losses land next to each other.
Think of two hikers who climb the same total height over a day. One climbs in a steady rhythm of small ups and downs. The other walks downhill for the whole morning and climbs it all back in the afternoon. The total height gained is the same, like a profit factor. The deepest valley each one had to walk out of is very different, like a drawdown.
Hypothetical example
Ten trades: five wins of $2,000 and five losses of $1,000. Profit factor = $10,000 ÷ $5,000 = 2.0 in any order. Net profit is $5,000 in any order. If wins and losses alternate, the equity curve never falls more than $1,000 below a high, so the recovery factor is $5,000 ÷ $1,000 = 5.0. If the five losses come first, the curve falls $5,000 before the first win, so the recovery factor is $5,000 ÷ $5,000 = 1.0. Same trades, same profit factor, recovery factor five times apart.
This is why the two numbers belong together on a report. Profit factor tells you whether the trades, taken as a set, were worth taking after costs. The recovery factor tells you whether the order they arrived in was survivable. A strategy with a profit factor of 2.0 and a recovery factor of 1.0 earned good trades in a pattern that would have taken a large part of the account before paying back, and the streak of losses may land in a different place next time.
Recovery factor vs Calmar ratio: what is the difference?
The Calmar ratio is adjusted for time and the recovery factor is not. The Calmar ratio is the annualised return divided by the maximum drawdown. Because its numerator is a per-year figure, a three-year test and a six-year test of the same strategy land near the same Calmar ratio. The recovery factor uses total net profit, which roughly doubles when the test runs twice as long, while the worst drawdown stays one event. So the recovery factor of a longer test is higher for no reason other than length.
| Recovery factor | Calmar ratio | |
|---|---|---|
| Numerator | Total net profit over the whole test | Annualised return (profit per year, or compound yearly return on the account) |
| Denominator | Maximum drawdown | Maximum drawdown |
| Effect of a longer test | Grows with the number of years | Roughly unchanged, if the drawdown does not deepen |
| Best use | Comparing strategies tested over the same period | Comparing strategies tested over different periods |
| Quick conversion | Recovery factor ÷ years ≈ a Calmar-style figure in dollar terms | Calmar × years ≈ recovery factor |
Hypothetical example
A strategy makes about $20,000 a year with one contract and its worst drawdown is $30,000 in both tests. Over 3 years: net profit $60,000, recovery factor $60,000 ÷ $30,000 = 2.0, profit per year ÷ drawdown = $20,000 ÷ $30,000 ≈ 0.67. Over 6 years: net profit $120,000, recovery factor 4.0, profit per year ÷ drawdown still ≈ 0.67. The recovery factor doubled; the time-adjusted figure did not move. (The drawdown is held equal here to isolate the effect of length; in practice it tends to deepen, which the last section covers.)
In the strict form used for funds, the Calmar ratio uses the compound yearly return as a percentage of the account, which needs an account size. For a futures backtest run with a fixed contract count, the dollar version above, yearly profit divided by max drawdown, is the practical equivalent. Whichever you use, dividing the recovery factor by the number of test years is the one-step fix when you must compare a short test with a long one.
What does the recovery factor look like in a real backtest?
The run behind this article, made with the AlphaPrime (Valynox) engine, gives a recovery factor of about 1.5 when its net profit is divided by its max drawdown. The run tested NQ daily bars from 2019-01-02 to 2025-06-30, long and short with one contract, with $20 commission and $10 slippage charged on every trade. Its rule is a stand-in: a Bears Power rule (one half of the Elder Ray indicator) used as the only entry and exit, with no stop, target or filter, because the recovery factor is a report number and not a strategy, and any rule with a drawdown would do to show the reading.
Source: AlphaPrime product engine (CPU), run 2026-10-31. NQ daily bars, 2019-01-02 to 2025-06-30, long and short, 1 contract, $20 commission and $10 slippage per trade.
| Number | Result |
|---|---|
| Net profit (closed trades) | $130,510 |
| Max drawdown | $86,105 |
| Trades | 227 |
| Win rate | 52.86% |
| Profit factor | 1.229 |
| Average trade | $574.93 |
| Time in market | 55.42% |
Recovery factor = $130,510 ÷ $86,105 ≈ 1.52. Read with the two numbers that belong next to it: the test is about six and a half years long and holds 227 trades. The profit factor of 1.229 says the trades as a set were only modestly positive after costs, and the recovery factor says that the path they took included one fall of $86,105, which the remaining years earned back about one and a half times. Spread across six and a half years, that is a recovery factor of roughly 0.23 per year, which is the figure to put next to any shorter or longer test.
This candidate is the one the search kept out of 1,024 it tried, chosen for its return against its drawdown. Picking the best of many after the fact flatters every ratio built from profit and drawdown, because the kept candidate is partly the one whose losses happened to land in a convenient order on this history. Bailey, Borwein, López de Prado and Zhu (2014) show how quickly that selection effect grows with the number of candidates compared on the same data. The 1.52 is a worked reading of one past run, not what the Bears Power rule earns or what a reader can expect. In AlphaPrime, the Performance Report shows the equity curve, its drawdown and the trade list, which is where the two inputs to this ratio are read from.
Why does a short sample make the recovery factor unreliable?
Max drawdown is a single worst event, so in a short test the denominator of the recovery factor is mostly luck. A test with 40 trades or two years of history has only had a few chances to line up its losses in a bad run. Add more history and the worst run usually gets worse, not because the strategy changed but because more time means more chances for a bad streak. Magdon-Ismail and Atiya (2004) show for a random walk with a drift that the expected maximum drawdown keeps growing as the observation period gets longer. A recovery factor measured on a short sample therefore sits on a drawdown that is probably too small, and the ratio is inflated.
The same effect shows up without waiting for more history. Reshuffling the trades you already have, as in a Monte Carlo backtest of the trade sequence, produces many alternative equity curves from the same trade list, and the drawdown in a large share of them is deeper than the one your single backtest happened to show. That exercise does not change the profit factor at all, which is one more way to see that the recovery factor measures the path and the profit factor does not.
Three habits keep the ratio honest:
- Compare recovery factors only at the same test length, or divide each by its number of years first. A higher ratio from a longer test is not evidence of a better strategy.
- Treat the backtest drawdown as a floor, not a ceiling. Size the account so that a drawdown noticeably deeper than the measured one does not force you to stop. The futures risk indicator and margin ratio turns a dollar drawdown into the margin you would have to hold through it.
- Check the drawdown of everything you trade together, not of each strategy alone. When several strategies lose in the same weeks, the combined drawdown of the portfolio backtest is the one you live through, and a good recovery factor on each strategy separately says nothing about it.
A test you can run today: take your last backtest report, divide net profit by max drawdown, then divide that by the number of years in the test. Rerun the same rules on only the first half of the history and repeat the calculation. If the recovery factor per year from the half-length run is clearly higher than from the full run, you have just measured how much a shorter sample was flattering the number.
Frequently asked questions
What recovery factor is good?
There is no single good value. A recovery factor below 1 means the test never earned back its worst loss, and most systematic traders want a ratio clearly above 1 over several years and a few hundred trades. Beyond that, compare only tests of the same length, or divide the ratio by the number of years, because net profit keeps growing with time while the worst drawdown is one event.
How do you calculate the recovery factor?
Divide the net profit of all closed trades, after commission and slippage, by the maximum drawdown of the equity curve. Max drawdown is the deepest fall from a high point to the following low point before a new high is made, measured in the same money units. A free drawdown calculator finds the deepest fall from an equity series if your report does not print it.
What is the difference between recovery factor and profit factor?
Profit factor is the sum of winning trades divided by the sum of losing trades, and it does not change when the same trades are put in a different order. The recovery factor divides net profit by max drawdown, and drawdown depends on the order of trades. Two strategies with the same profit factor can have very different recovery factors if one of them groups its losses together.
How is the recovery factor different from the Calmar ratio?
The Calmar ratio divides the annualised return by the maximum drawdown, so it is adjusted for time and comparable across tests of different lengths. The recovery factor divides total net profit by the maximum drawdown, so a longer test of the same strategy shows a higher recovery factor for no reason other than length. Dividing the recovery factor by the number of test years gives a figure close to the dollar version of the Calmar ratio.
Does trading more contracts improve the recovery factor?
No. Doubling the contract count doubles every trade result, so net profit and max drawdown both double and the ratio stays the same. Size changes how large the drawdown is as a share of the account, which decides whether you can hold through it, but it does not change the recovery factor.
References
- Magdon-Ismail, M., & Atiya, A. F. (2004). Maximum Drawdown. Risk, 17(10), 99–102.
- Bailey, D. H., Borwein, J., López de Prado, M., & Zhu, Q. J. (2014). Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance. Notices of the American Mathematical Society, 61(5), 458–471.