A system making $48,000 of net profit with a worst drawdown of $12,000 has a recovery factor of 4. It is a quick, unitless read on whether the returns justified the worst stretch.
Rough interpretation for a multi-year record: under 2 is thin, 3 to 5 is solid, above 10 usually means the sample is short or the worst drawdown has not happened yet. Unlike calmar-ratio it is not annualised, so it rises mechanically with record length - which makes it useful within one backtest and misleading across records of different lengths.
Its best use is comparative, across parameter settings or strategy variants tested over the identical period. There it answers the practical question directly: for the same pain, which version paid more?
Related: calmar-ratio, max-drawdown, profit-factor, drawdown-recovery-maths