Have you thought of maybe attaching an ATR Indicator to your EA. Maybe that will help in detecting an oncoming trend. I would be very interested in testing this EA. Would you like to share the EA please?
Hi Jim, the EA that I am testing at the moment is called Killa Gorilla EA (just do a Google search).
I did think about using an ATR indicator on a separate chart, but this would involve some manual intervention and the idea behind my trading is to be completely automated. The EA does say that you can trade 100% of the time in automated mode, so I guess I can use some of the safety mechanisms it has like the trend and spike protection. It's definitely one of the better EAs I have come across in the last couple of years.
Thanks for your advice.
I was thinking how to improve this system even more and I think I have succeeded in doing that. So the main problem occurs during the ranging market conditions. Then the EA will keep opening buy and sell positions every time the price will cross the recovery line levels. In worst case scenario it could open even 10 positions (or even more). This will totally destroy the FreeMargin level and introduce a risk of "unfinished recovery" cycle. So my idea
was to introduce a "moving recovery target" mechanism. The idea is to shift the recovery levels closer to the market, each time a new recovery trade is opened. Doing so, the EA would be targeting lower and lower recovery levels and this would increase the chance of successful recovery.
To test this improved recovery system the EA (version 01.1) is programmed to enter the market almost randomly, where every new market entry decision is based only on the information from the previous candle and a new position is automatically opened when the previous trade (or trade group) is closed. The entry logic is:
This is of course the stupidest entry algorithm ever, so normally your account would be wiped out in several days. This kind of trading would end in a random walk towards the $0 line. So let's run a backtest from 2007 until now and see if this system can survive this worst case random test.
- initial deposit $10000
- EA settings: see tester report
- EURUSD @ H4
- dukascopy data 2007-2014 is used
- fixed spread of 3pips (to simulate my broker's spread)
- backtest performed using tickdata with 99% of modelling quality, this is extremely important since the results are very
sensitive to small price changes
- back tester set to "every tick" mode
Strategy Tester Report:
Link to the full statement:
Strategy Tester: CoensioRecoveryEaV01
So as you can see, apparently this strategy can withstand even 7 years of random trading! Personally I think it is, a very GOOD result.
p.s.: I'll post the updated EA version 01.1 as soon as possible.
it is very interesting
is it something like this?
i need help cuz i just looked it up and found it interesting
Plenty of them in the market. Blackstar and cap zone use similar strategy.
This kind of system made of pure hedging is very difficult to elaborate, there's a lot of constraint.
I never saw one stable with no stopout in long term.