If we could tie the martingale coefficient or step size to volatility we could get better results and reduce risk.
As volatility picks up, we increase the coefficient for the next martingale. That way we will have a higher chance to "catch" the big swings. Meanwhile when volatility is low, our martingale spread would be also smaller, thus optimizing profitability.
So an market adaptive system should be more efficient.
Roman, have you tried Martingale type "ATR * coefficient"? Is it what you want to implement?
Andrey, I have used ATR coefficient type with some success.My assumption is that it always uses ATR and ignore Step Size. This can be positive but also negative. I have following thoughts:
1. Ideally we can track two different volatilities. The long term volatility and short term volatility.
2. Long term volatility (weeks or days) would influence step size for the first 3 martingale steps, thus optimize profit especially in low volatility phases.
3. Short term volatility (minutes or hours) would influence step size for the latter martingale steps and serve mainly as a recovery mechanism. It would expand the steps so wide that we "catch" the occasional big swing that often happens with news.
4. To realize this we would need a martingale "phase" separator (# from when to use second config), two configurable ATRs, two atr * coefficients, two lot *coefficients.
Andrey, I have used ATR coefficient type with some success.My assumption is that it always uses ATR and ignore Step Size. This can be positive but also negative. I have following thoughts:
1. Ideally we can track two different volatilities. The long term volatility and short term volatility.
2. Long term volatility (weeks or days) would influence step size for the first 3 martingale steps, thus optimize profit especially in low volatility phases.
3. Short term volatility (minutes or hours) would influence step size for the latter martingale steps and serve mainly as a recovery mechanism. It would expand the steps so wide that we "catch" the occasional big swing that often happens with news.
4. To realize this we would need a martingale "phase" separator (# from when to use second config), two configurable ATRs, two atr * coefficients, two lot *coefficients.
The drawback I could see is, by using long term volatility for first 3 martingale steps, we may be losing our profits slightly, but will improve DD. Gave my vote, Thanks.
Is this even possible to automate!!! If it is, then I definitely give it 5 stars