engtraderfx

Re: Fapturbo EA - All Versions

Last version I have file is attached. I don't know why there are problems for some running this file sorry, I have it running on alpari for usdchf with no problems & making trades.

http://www.4shared.com/file/218579288/5f351212/ftu4s.html

chrs, D

Re: EA Generator

any chance of update to v1.07?

Re: Forex God Father Version 2.0 + Elite Version + Elite Bonus !

Just posted to the PM as well...I have had it trade on GoMarkets recently & Alpari for some time.

I found a few minor erorrs in original code but nothing that stopped it working.

Got that loss yesterday too unfortunately but not a problem if the risk is kept low!

Re: Forex God Father Version 2.0 + Elite Version + Elite Bonus !

well lets hope so, the only concern in my mind is that obviously you have to optimise the variables to get good performance which means there is a risk of curve fitting & failure in the future. The only thing you can do is optimise over as much data as possible & then test out of sample data.

I concur with Ansatsu_ that's it has very basic logic & the original settings are very optimised to recent data. If you are going to use it, as above I can only suggest optimising over a lot of data to get a more robust set & remember it's a long term trend type system which must be run at low risk (ie 1% risk) and you will get a number of losers in a row at time. At least if its wrong & the risk is low you don't lose too much!

ps....if anyone knows a better trending type Expert, let us know!

Re: Forex God Father Version 2.0 + Elite Version + Elite Bonus !

well lets hope so, the only concern in my mind is that obviously you have to optimise the variables to get good performance which means there is a risk of curve fitting & failure in the future. The only thing you can do is optimise over as much data as possible & then test out of sample data.

Re: FapTurbo v.49

not sure kilu, I have't used a cent account before. It does look at the broke lots size values so maybe it works. But if not just set manual lots as you you have.

chrs Dave

Re: Forex God Father Version 2.0 + Elite Version + Elite Bonus !

Godfather v2_ModA: I have been testing GF & modifying with a few extra features.

ModA includes following:

- VROC is now included in code, no separate indicator is needed.

- Breakeven Stop

- Trailing Stop

- "Semi-support/resist stop" (Optional, in testing, seems to help if check long term hi - low breached with a minimum take profit level.

The attached is my ea mod & optimised set file for interest. I have NOT started trading yet so do your own review & testing! Let me know if you find any errors.

http://[email protected]/167546

Hope it helps...Cheers D.

Re: FapTurbo v.49

Unplugged version attached, scalper only. f0rexsh@ring/382207.

Updated from v48 unplugged to include USDCHF default settings as per ft49.

Minor mods to code including:

Version 4s

- v49 update, all defaults updated

- added USDCHF pair with auto settings

- SCALPER ONLY VERSION

- add variable dpips to set deviation from mov avg

- added correct conversion of pips to dpips to allow testing of 2/3 digit codes (ie JPY)

- simplified Autosettings code, just make UseAutoSettings = True to use internal code, if using wrong pair will use default settings.

Have fun. Chrs, D

Re: FapTurbo v.49

don't forget the v48 unplugged is already available to run on any pair without dll for scalping. Just need to set the user variables to suit usdchf (tp 8, sl 82, & the trend roc parameter needs adjusting.

Re: Fapturbo v48

Invalid price errors mean that the number of digits is greater than the broker allowed digits (ie 4 or 5).

This does occur in backtesting due to rounding errors in historical data.

Many experts (& probably this one as well), don't do any rounding checks on the Bid / Ask price that is used for market entry orders so during backtestings you get these errors. Many (but not all) EA's do this properly for calculation TP & SL but many do not do it for the basic Bid/Ask price.

I have started making it a practice in all EA's to round or normalise even the Bid / Ask or final calculated price sent with all orders to minimise the number of errors.

Hope this helps..regard, David

Re: Fapturbo v48

hi...

dll must be activated in settings.

Also dll is not required for scalper testing but is required for long term system & you will need the dll saved in appropriate directory on metatrader. The dll is the ft48 version renamed (just look in the code for the name). The dll was included in the original download link.

Re: Fapturbo v48

Hi Ansatsu, to calculation the seconds required to a particular date, pull out excel & do the following calculation:

Excel dates converted to a number give number of days, so need to convert to seconds.

Cell1: 01/01/1970 00:00

Cell 2: NewDate

Cell 3: = (cell2 - cell1)*86400

That will give you the seconds required for MT4 from 1970.

The ammended code is below, it compiles so hope that helps!

Best Regards, Dave

int DSTSetting()
{
/*
Calculated DST adjustment.  If within DST time, adjustment = 1, else 0.  DST adj is subtracted from manual GST setting.
ALPARI / EUROPEAN Daylight savings adjustments
// time is number of seconds since 00:00 Jan 1 1970
*/
double gi_timeCurrent = TimeCurrent();
if(gi_timeCurrent < 941068800) return(0); //1999-10-28
else if(gi_timeCurrent <   972345600)  { if(gi_timeCurrent >= 954460800 )  return(0);    
else if(gi_timeCurrent <   1003708800) { if(gi_timeCurrent >= 985996800 )  return(0); }
else if(gi_timeCurrent <   1035763200) { if(gi_timeCurrent >= 1017532800)  return(0); }
else if(gi_timeCurrent <   1067212800) { if(gi_timeCurrent >= 1049068800)  return(0); }
else if(gi_timeCurrent <   1098921600) { if(gi_timeCurrent >= 1080691200)  return(0); }
else if(gi_timeCurrent <   1130284800) { if(gi_timeCurrent >= 1112230800)  return(0); }
else if(gi_timeCurrent <   1161648000) { if(gi_timeCurrent >= 1143774000)  return(0); }
else if(gi_timeCurrent <   1193011200) { if(gi_timeCurrent >= 1175310000)  return(0); }
else if(gi_timeCurrent <   1225065600) { if(gi_timeCurrent >= 1206928800)  return(0); }
else if(gi_timeCurrent <   1256954400) { if(gi_timeCurrent >= 1238464800)  return(0); } // 31/3 and 31/10/2009 2 am
else if(gi_timeCurrent >=  1270000800) return(0);} //2010-03-31 02:00	
return(1);
}

Re: Fapturbo v48

Fair enough, I thought you wanted to review the past GMT changes. You will need to add new dates as time goes on. That calc isn't hard, is just the number of seconds after 1970 (just check MT4 editor help for time functions to confirm.

When I looked at the GMT/DST website I found its not the same day every year so unfortunately you can't just do DST say "every 3rd Friday in March & October" (although it would be a close approximation!). So you have to update the DST code as time progresses.

I won't make any more updates, I did that as an exercise to see if it made much difference. From my point of view it was interesting but didn't make a massive difference. As we all know, forward testing is the real acid test!

Chrs, Dave

Re: Fapturbo v48

absolutely, should do your own testing to find acceptable sl / tp values to you and keep risk down. I only use fapt for EURCHF currently at risk level 1 to 2 %. I use my own ea's for usdcad & others.

Re: Fapturbo v48

Hullo,

I finally got to update faptu to v3 with a couple of mods as requested.

V3 changes include:

1. Include DST Setting adjustment for Alpari data. Works by reducing manual GMT by 1 hour during certain dates.

2. Logic for "anypair" altered so that 'default' values are used if not a normal fapturbo pair (EurGBP, EURCHF, USDCAD, GBPCHF).

Please note the DST setting changes are for Euro session (Alpari) only and the dates are from another posting & are assumed correct. I did a brief check on a GMT website but didn't find any major errors. This was done as an exercise to see if it effects backtesting. A check of EURCHF does show a minor improvement in backtesting but does not removal all stoplosses. If anyone finds an error just let me know & will update.

load files from http://www.f0rexsh@ring_com/358627

Please remember, use at own risk & do own backtesting to confirm settings. I can only suggest keep hours short & TP about 4 or 5. My own settings are included as a start & I use at low risk (risk of 1 or 2).

Re: Fapturbo v48

why not u allow any pair set it into true.... if it false, it cant trade other than eurchf, eurgbp, usdcad and gbpchf..

that is correct...set "any pair" to true = run on any pair. If false, only allows original fapturbo pairs.

Also, you can override the fapturbo scalper settings (ie tp, sl etc) by setting "UseDefaultSettings" = True. Set to False to use inbuilt fapturbo settings.

I always leave the original code & options available in case someone wants to use them.

Just flick the switch to go between them.

Hm.....I just realised when setting "UseAnyPair=True" you have to use the default settings (set to true) or it will look for the standard fapturbo pairs. That may be why it hasn't worked for other pairs for some. I'll have to adjust that later to force it to use default settings.

Re: Fapturbo v48

hi apple...sorry but should work, have it running live now on eurchf.

Re: Fapturbo v48

Thanks, I have a number of those versions too, unfortunately fun1() is always used which is a dll call. That's the one I can't figure out. I'll just have to try & nut it out!

Re: Fapturbo v48

I have an old version of FAP & tried the rules but didn't match fapturbo, I got something wrong. I'll havea look if you have a good version you think will help.

Re: Fapturbo v48

saendee...what I need is the actual code logic for the long term EURUSD system. I think its similar to the original FAP expert which I have, but its not quite the same so can't reproduce the results.

Re: Fapturbo v48

thanks engtraderfx

i have tweak the barheightmin and trendroc ,but it seems they trade at the same position.when trendroc became smaller it trades less.

That sounds right..

Barheight is just a filter, it doesn't allow a trade if the previoius bar range is too high. It doesn't set the trade price (that would be a new rule).

TrendROC - that is correct, if you make the limit smaller, you are asking the long term trend to be closer to zero which gives less opportunities. Increasing the roc does allow more trades but at some point will start letting in losers.

Chrs, Dave

PS - found the daylight savings code from another post, will look to include soon.

PPS - anyone know the rules for the long term? I would still like to unplug the whole expert.

Hi everyone,

I have a general question for anyone interested in chipping in about overall trade management. Its not something discussed much here. As background, I have a concept in mind about developing a risk management module in form of an expert advisor whose function is to calculate various account performance & risk criteria that can then be referenced by individual experts (via global variables). I have heard of people using a 'master' expert which I assume is this kind of thing.

I think this is important now with the number of expert we are running (ie at least 10 or more), often many with similar strategies (mainly thinking scalping here) & sometimes doubling up on currencies eg dragon, fapturbo, megadroid etc all do pretty much the same thing.

At a basic level we could just program some things into individual EA's (like max no posn's, max same pairs open etc) but seems a large duplication of processing time. I know the most basic method is just setting every ea risk very low, but there could be benefit in smarter account management. So generally I was wondering does anybody bother looking at their total risk in terms of total max loss exposure, margin usage, no positions open etc, equity gain / loss per day / week etc.

Some metrics I am thinking of calculating in either real time (tick) or other (eg X min) are:

- total exposure (max loss of all open positions at stop loss)

- total margin utilized vs max limit

- no of total positions

- no of same pairs open

- no of correlated pairs open

- equity profit / loss per day / week / month

Then the ultimate idea would be to create global variables of these metrics with limits that each EA much check to allow a trade. Any other ideas welcome...

Chrs, Dave

Re: Fapturbo v48

Unplugged [-O<

I have been trying but couldn't figure out rules for the long term code. All other functions are unplugged tho. You can actually run the scalper part without the dll so it is effectively unplugged for scalping.

Chrs, Dave

Re: Fapturbo v48

hi sunyc... the trend filter is as per the response above, the trend filter is checking if the velocity or the rate-of-change of the mov average is close to zero (ie not moving much). The rate-of-change is changed to an absolute value so it looks for a value within +/- 0.3 (for example).

Increasing the value from 0.15 to 0.35 allows for more movement in the mov average & hence allows more trades. Backtesting is needed to see if it still works ok without letting in too many stop losses.

As always tho its not perfect as the moving average can be moving very slowly just before a major breakout (& hence the occassional stopout).

Hope that helps..chrs, Dave

Re: Fapturbo v48

hi guys...

gi_605 = 25 is the hi lo difference setting...fapturbo looks at the last 2 bars to see if the high - low range is greater than 25 pips or won't trade. If you want to test other values, try the default settings (set UseDefaultSettings = True) & set default_barheightmax = 35;

gd_648 is the trend filter limit. The trend filter is the absolute value of the change in the moving average over the last 4 days. Because its an absolute value it will always be +ve & means that the moving average trend filter will be between -0.2 & 0.2 when true. The concept is that fapt is looking for a trend with a small rate of change around zero.

Will get to the gmt stuff later guys, maybe the weekend. I started back at work this week & travelling for clients so out of action this week.

chrs, Dave