The Journey to RB51362

[rsmereka]

Please join me on the journey to building an EA from an existing manual system.

The original manual system is Rob Booker's 5/13/62 which is a manual system based (in part on 3 EMA's, 5 (fast), 13 (medium) and 62 (slow). Here are the rules in brief:

- 5 crosses 13
- 13 crosses 62
- 13 at least 30-40 pips from 62
- wait for price to retrace back to 62
- open trade with 25 pips SL
- TP at previous support/resistance
- TS of 20

The complete rules (and comic relief) is here.

When I started to plan building the EA, I realized that I don't know how to program a triple MA cross in MQL4 so I went hunting for examples. I came across the Three Moving Averages EA at tradingsystemforex.com. This is a completely free EA (you have to register however). This EA is from 2008 and it is not aware of 5-decimal brokers (you have to multiply the SL, TP, TS by 10. It is also not ECN aware.

My plan is to implement the rules of Rob Booker's 5/13/62 one at a time and backtesting to make sure there are no bugs. I will use the triple MA cross from the Three Moving Averages EA.

I call this new EA RB51362 (creative eh?).

My first step is to clean up the original source code. It's a mess with no proper indentation and everything is jam-packed together and run some backtests just testing the straight 5/13/62 EMA cross with a 25 pip SL. I am running an optimization right now to determine the best TP.

More later,

Update 20110418 - Attached version .01-2011.04.18, License is GPL V2
Rick

[rsmereka]

The EA with basic logic is done. I have attached the EA to post 1 of this thread. This version has MM or fixed lots (MM is turned on by specifying a risk greater than zero). EA is aware of five decimal brokers but not capable of running on an ECN broker.

The SL and TP are fixed at the moment. Attached is a backtest from Jan 3, 2011 to Apr 18, 2011 at risk 2% (0.02) with a SL of 25 and a TP of 98 (that's a risk:reward just shy of 1:4). The winning rate is not even 30% but with that risk:reward, the roi is 242.74% in less than four months.

This trading strategy can work on any timeframe. I just picked M30.

The next step is to make the SL and TP dynamic based on levels of support/resistance.

I also ran a backtest from Jan 3, 2000 to April 18, 2011 and the EA blew the account. It seems clear that the 5/13/62 logic needs a dynamic TP and SL.

Rick

[Stormy]

I have been a fan of the 5-13-62 for some time - do you have any updates or additional results on this EA?

[rsmereka]

Originally Posted by Stormy

I have been a fan of the 5-13-62 for some time - do you have any updates or additional results on this EA?

Hi and Welcome,

No further updates. I lost interest in this system and EA.

Rick

[Stormy]

rsmereka -

Thanks for the reply and the welcome.

I am studying your EA because when I look at this method and the charts it looks like the results could be inproved if you entered when the fast and medium both cross the slow MA. But close the position as soon as the fast crosses the medium and don't wait for the slow.

Do you know how this EA could be modified for this condition?

[rsmereka]

Originally Posted by Stormy

rsmereka -

Thanks for the reply and the welcome.

I am studying your EA because when I look at this method and the charts it looks like the results could be inproved if you entered when the fast and medium both cross the slow MA. But close the position as soon as the fast crosses the medium and don't wait for the slow.

Do you know how this EA could be modified for this condition?

Yes I do,

It's too technical to explain here. Unless you know MQL4, you would not understand.

My original EA has a GPL license which means that it's ok to modify it and produce offspring as long as the offspring are also GPL.

If you are interested in a strategy other than the original, go ahead.

Rick

[Stormy]

Rick -

I'm okay with MQL4 - I've been studying it for a few months. I really liked you code because it is clean and easy to follow I'm just not sure exactly how to implement this closing function for just the cross of the MAF and the MAM but not the MAS.

Thanks for any guidance you can provide.

Stormy

[rsmereka]

Originally Posted by Stormy

Rick -

I'm okay with MQL4 - I've been studying it for a few months. I really liked you code because it is clean and easy to follow I'm just not sure exactly how to implement this closing function for just the cross of the MAF and the MAM but not the MAS.

Thanks for any guidance you can provide.

Stormy

I got tons of other projects that need my attention,

I wrote this code in Apr of 2011, almost a year ago. It would take me too long to acclimate.

Sorry, I don't have time or the inclination (as previously mentioned).

If you know MQL4, you do it and post it here.

Rick

[Stormy]

Thanks Rick - I understand - I'll post it and the results when I get it done.