~What is your EA recovery factor?~(Double-R FX Miscellaneous Talks)
★As a criterion for EA selection, it is most important to compute the recovery factor from backtesting.
…The recovery factor is
calculated as (net profit / maximum drawdown) during backtesting.
A desirable value is 10 or more over 10 years.
※But, per one position, it is
※Moreover, it must be backtested with a sufficiently large spread.
▼Everyone, how is your current EA?
It may surprisingly not clear this criterion.
▲This is an important standard for whether an EA will continue to be profitable in the long term.
▼Even if forward performance is good in the short term, if the recovery factor isn’t high, you may incur large losses in the future.
▲If the recovery factor over a 10-year backtest is 10, then, by simple calculation, it is 1 per year.
▼In other words, if you consider the maximum drawdown, it will take about a year to recover.
▼You might think that’s surprisingly long, but if the recovery factor falls below 10, it becomes even longer.
▲Not only for the EA you operate but also for commercially available EAs, it’s good to run the calculation.
★In addition, if the backtest equity curve is rising to the right at all times, (there are software to analyze this), you can say you’ve gained a step on choosing the EA.
★This is, of course, only when you operate in the long term.
For example, you might start with a small amount with an EA that could profit explosively in the short term, and when you make profits, you could run the stable EA described above and leave it running for the long term; whether this idea is OK or not is a matter of opinion, I think it’s OK.
♪For reference, I’ll list as much as I can think of now the checklist when looking at backtests.
① Is the recovery factor high?
② Is the backtest conducted with a sufficient spread?
③ Is the expected gain not overwhelmed by the spread?
④ Is it rising to the right at all times?
⑤ Is it based on the opening price movement?
⑥ Is the backtest over 10 years?
⑦ Is it on 5-minute bars or higher?
⑧ Is the number of trades sufficient?
× ![]()