Is a higher recovery factor always better?
A higher number means net profit covered the worst drawdown more times, which is usually preferable. A very high ratio on a tiny sample, or on a period that never saw a real losing streak, can still be fragile. Prefer a stable reading over a long journal to a spectacular reading after a handful of trades.
How is recovery factor different from Calmar ratio?
Both put drawdown in the denominator. Calmar ratio uses annualized return divided by maximum drawdown, so it is a yearly rate. Recovery factor uses total net profit over the sample, not an annualized return, so a longer profitable period can raise the number without the strategy getting safer. Use Calmar when you want a yearly comparison; use recovery factor when you want how completely the book recouped its worst hole.
Should I use dollars or percent for max drawdown?
Use the same unit as net profit. If net profit is in dollars, convert max drawdown to dollars (the peak-to-trough drop in equity). Mixing a dollar profit with a percentage drawdown produces a meaningless ratio. Mantis keeps the trade log in one currency so both sides of the formula stay aligned.
What if net profit is negative?
Then recovery factor is negative: the book lost money overall, so there is nothing to recover relative to the drawdown. A negative figure is a warning, not a ranking to optimize. Rebuild the sample after the strategy changes; do not compare a losing period's recovery factor with a later winning one as if they were the same test.
How many trades before recovery factor is meaningful?
You need enough history for a genuine peak-to-trough drawdown, not just a couple of losers. Tens of trades can sketch the formula; a few hundred, or several months of the style you actually trade, is a more honest reading. Recalculate after a new equity high or a new deepest trough, not after every single fill.