Proposed complex EA. Is this system possible?

 
 
 

Freeman, what has been your winning pecentage with this approach manually, what pairs does it seem to work best with and is the trail stop your only exit? This would not be the easy to program for the uninformed - I may be able to help.

DeadEye96

 

DeadEye96,

I only trade this on the 3 main currency pairs (EUR/USD, USD/JPY, GBP/USD) and have been doing so for around 3 months now. Based on that win/loss percetage is around 65% and risk/reward ratio works out around 2.2. I can't do any thorough backtesting to get any better stats as unable to program the rules that don't lend themselves very well to backtesting hence the initial posting.

I do in fact use other criteria for exiting other than the trailing stop you queried (and also for entering as well). I place close atention to s/r, trendlines, pivot levels and also do no trade it approaching news. However, some of these other elements would be even more complex to programm a rules based EA around so I was just trying to get the nucleus of the system programmed as an EA first and then look to add other factors later.

Have you come across other EAs that automaticaly draw fib levles and trade off them? Be very interested in any help you could offer.

Eaglehawk,

Thanks for the link to the site. I'll have a look at it to see if it has any examples of doing anything so complex as what I am after.

 

Fib Sell Off

The only EA I am familiar with that sells off at prescribed Fib levels is the 4H-Vegas system that has been posted on lightpatch.com/forex. You may want to check it out. It uses a MA tunnel approach. I am sure that other Fib sell-off EA's exist though.

Regarding creating an EA, a suggestion. Define all of the indicators and trading rules to the point that you can build a flow chart/decision tree with verification through notes made on all of your manual trades. If you have complex decision rules, it is much better to define as many as possible up front rather than add them piecemeal. Some adjustments will probably be needed later anyway such as timefilters, etc.. This way, it will be much easier for someone to program your EA.

Reason: