Expert Advisor Builder | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Expert Advisor Builder 7 replies

MetaTrader Expert Advisor BUILDER for backtesting and auto trading 4 replies

How to get many trades with Expert Advisor Builder? 1 reply

How to mod Expert Advisor Builder code 0 replies

Expert Advisor Builder Help 2 replies

Attachments: Expert Advisor Builder
Exit Attachments

Expert Advisor Builder

  • #61
  • Aug 4, 2013 10:22am Aug 4, 2013 10:22am
Hey, RaptorUK. Good to see you again. I remember having a big discussion about a builder with you, WHRoeder, and Onewithzachy over at mql4.com.

mqlautocoder.com, to some extent, grew out of that discussion. I'm not here, or there, to promote any individual EA builder. The goal is to allow users of the builders to decide which software is preferred. Note that, under the site's rules, software publishers must disclose identity up front. Coders for hire would be well advised to do the same.
An ounce of profitable logic is worth a pound of efficient code.
  • #62
  • Aug 4, 2013 10:56am Aug 4, 2013 10:56am
Quoting RJo
Hey, RaptorUK. Good to see you again. I remember having a big discussion about a builder with you, WHRoeder, and Onewithzachy over at mql4.com. mqlautocoder.com, to some extent, grew out of that discussion. I'm not here, or there, to promote any individual EA builder. The goal is to allow users of the builders to decide which software is preferred. Note that, under the site's rules, software publishers must disclose identity up front. Coders for hire would be well advised to do the same.
Most importantly, new Forex traders need to TAKE SERIOUS LESSONS from the FACT that NO EA can make trades for you to profit consistently.

Markets have far too many variables for any one EA to deal with. Trending markets are not ranging markets etc.

EAs DEFINITELY have a role in "trade management" functions not provided by MT4/5.

For successful trade entries they are USELESS most of the time.
  • #63
  • Aug 4, 2013 11:16am Aug 4, 2013 11:16am
Indeed, attempting to use one EA to handle every market environment is probably not advisable. Seasoned autotraders generally use a "portfolio" of EA's for different pairs, session times, trends, ranges, etc.

Having said that, some of the EA builders can create pure exit/pending order cancellation EA's for semi-automatic trading strategies. Just delete the magic number code or use a magic number of 0.
An ounce of profitable logic is worth a pound of efficient code.
  • #64
  • Aug 4, 2013 11:22am Aug 4, 2013 11:22am
Quoting RJo
Indeed, attempting to use one EA to handle every market environment is probably not advisable. Seasoned autotraders generally use a "portfolio" of EA's for different pairs, session times, trends, ranges, etc. Having said that, some of the EA builders can create pure exit/pending order cancellation EA's for semi-automatic trading strategies. Just delete the magic number code or use a magic number of 0.
AGREED.

The most important function of trade management EAs that has been consistently ignored is the hedging function, and its derivatives.

The Forex world of retail traders seem forever lost to the use of stop losses, as if they were ever useful.

Stop losses are for losers; losers confirming their losses such that they cannot re-enter a trade with the same lot size. Hedge management in the same currency pair, (not to be confused with arbitraging against other currencies) is an important school of study NOT found on Forex forums.

And so profitable trading by the masses suffers.
  • #65
  • Aug 4, 2013 1:06pm Aug 4, 2013 1:06pm
Hear hear, dancingphil! Hedging is a huge feature to look for in an EA builder. For anyone in the States, you'll have to escape NFAland to use it. This is difficult, but not impossible.
An ounce of profitable logic is worth a pound of efficient code.
  • #66
  • Aug 4, 2013 2:35pm Aug 4, 2013 2:35pm
Quoting RJo
Hey, RaptorUK. Good to see you again. I remember having a big discussion about a builder with you, WHRoeder, and Onewithzachy over at mql4.com. mqlautocoder.com, to some extent, grew out of that discussion. I'm not here, or there, to promote any individual EA builder. The goal is to allow users of the builders to decide which software is preferred. Note that, under the site's rules, software publishers must disclose identity up front. Coders for hire would be well advised to do the same.
Ah yes, I didn't recognize your ID, Molanis and it's poor mql4 code was the root of the discussion as I recall. If your "coders for hire" comment was aimed at me it is misdirected, I've not coded for any amount of cash, for free for some people, yes, but not for hire.

My question assumed that you had experience of many of the EA builder type apps and might be able to offer an opinion of which produces the best code, just wondered if you had an opinion ? I not used any and have only seen the end results from a few, they have all been woeful, maybe there is a good one that I haven't seen ?
20 pips a day isn't too much to ask . . .
  • #67
  • Aug 4, 2013 3:24pm Aug 4, 2013 3:24pm
The phrases, "software publishers" and "coders for hire," are both aimed at those who might want to seed or flame an otherwise neutralt forum that I fund out-of-pocket and moderate.

Offering up my own opinion on any EA builder(s) here would be improper and commercial... although I'm flattered by your request.

Yes, mql4.com... And then there was the time we crossed paths at that software publisher's support forum...

Like you, I have only used a few EA builders--another good reason not to make generalized statements regarding all 17.

I suspect that an expert programmer will never have a need for an EA builder, and autocode will never be acceptable to that programmer. Unfortunately, the nonprogrammer is faced with begging, borrowing, or buying (or stealing). What would it mean to the programmer if autocode were acceptable?
An ounce of profitable logic is worth a pound of efficient code.
  • #68
  • Aug 4, 2013 3:35pm Aug 4, 2013 3:35pm
Quoting RJo
The phrases, "software publishers" and "coders for hire," are both aimed at those who might want to seed or flame an otherwise neutralt forum that I fund out-of-pocket and moderate. Offering up my own opinion on any EA builder(s) here would be improper and commercial...

I wouldn't be surprised if your initial post may were to end up being regarded as commercial anyway . . .

I don't think anyone would say that you aren't entitled to voice your opinion.


Quoting RJo
Like you, I have only used a few EA builders--
Nope, I have never used any other than my installation of Molanis and subsequent discovery that I couldn't get access to the mql4 it produced.


Quoting RJo
What would it mean to the programmer if autocode were acceptable?
Relief, it would mean fewer requests over at the mql4 forum from people asking for their bad code to be fixed. For other less experienced programmers it could be a source of base code and a point from which to learn.
20 pips a day isn't too much to ask . . .
  • #69
  • Aug 4, 2013 6:45pm Aug 4, 2013 6:45pm
Quoting RaptorUK
I wouldn't be surprised if your initial post may were to end up being regarded as commercial anyway . . .
How ridiculous. In a thread titled, Expert Advisor Builder, riddled with commercial videos and direct sales links; I post a link to a comparative review forum with only one affiliate link--my host to discount my hosting expense. You'd better explain this to FF members like seanmcgu, who was first to reply to my initial post.
Quoting seanmcgu
Thanks, RJo. And I've also found these, so now, I should be able to put something together ...
Quoting RaptorUK
I don't think anyone would say that you aren't entitled to voice your opinion.
This would be a bit like Angie endorsing a contractor who appears on Angie's List.

Quoting RaptorUK
I couldn't get access to the mql4 it produced.
For that particular EA builder, you need the commercial license to view/share the code. It's not available during the trial period. This is explained at mqlautocoder.com.

Quoting RaptorUK
Relief, it would mean fewer requests over at the mql4 forum from people asking for their bad code to be fixed.
Easy there... I've seen others fix your code too. This is not mql4.com.

Quoting RaptorUK
For other less experienced programmers it could be a source of base code and a point from which to learn.
Less experienced programmers probably constitute the gross majority of MT4 users, including me. I think we've reached a point of agreement.
An ounce of profitable logic is worth a pound of efficient code.
  • #70
  • Edited 2:27am Aug 5, 2013 12:07am | Edited 2:27am
Quoting dancingphil
{quote} Most importantly, new Forex traders need to TAKE SERIOUS LESSONS from the FACT that NO EA can make trades for you to profit consistently. Markets have far too many variables for any one EA to deal with. Trending markets are not ranging markets etc. EAs DEFINITELY have a role in "trade management" functions not provided by MT4/5. For successful trade entries they are USELESS most of the time.
Phil, good to see you posting again, but I must beg to differ with you somewhat.

My trading buddy and I have thus far developed 4 different trend-following EAs that are long term profitable. None of them hedge; they all use stoplosses productively. Granted, they generate a relatively modest income, but all win more than they lose, over time. There's no curve fitting involved, nor user intervention required; all run continuously, and simply use key levels and patterns for their entries, and then let profits accumulate on all surviving entries, whenever long timeframe trends persist.

The EAs operate using fixed MM; there is no attempt to recover from losses, average down, or use progressive lot sizes (like martingale). They simply accept losses and wait for the next setup, working on the premise that there is overall positive expectancy from the entries and exits alone.

A lot of research has gone into these EAs; I agree that it's by no means easy to find concepts that are both totally mechanical, and capable of surviving all types of market conditions. I also agree that analysis performed by an experienced human trader will always be superior to that of a 'dumb' EA.

The EAs have been manually coded; I doubt that they could have been created using a code builder.

@everybody: In case anybody reading this is wondering, we won't be running a Trade Explorer to prove this claim, nor sharing the concepts involved, nor selling the EAs at any price. People can believe whatever they want to.
  • #71
  • Aug 5, 2013 4:30am Aug 5, 2013 4:30am
Going slightly (?) off topic here but I was drawn to this tread through a search. I am just exploring the feasibility of backtesting a strategy on MT4 and the UTube video provided by my broker stated that I had to do the backtesting on an EA. I wanted to backtest on a indicator (a simple moving average cross) but don't know if this would need to be converted into an EA and whether it would need to action entry into trades. In reality I would still carry out due diligence in live trading and enter and exit trades manually - anyone any advice on this or guidance on the way forward/
  • #72
  • Aug 5, 2013 9:50am Aug 5, 2013 9:50am
Quoting TraderJak
Going slightly (?) off topic here but I was drawn to this tread through a search. I am just exploring the feasibility of backtesting a strategy on MT4 and the UTube video provided by my broker stated that I had to do the backtesting on an EA. I wanted to backtest on a indicator (a simple moving average cross) but don't know if this would need to be converted into an EA and whether it would need to action entry into trades. In reality I would still carry out due diligence in live trading and enter and exit trades manually - anyone any advice on this...
This is an EA that even the most basic EA builder can code. There are plenty off free ones capable of making a simple MA EA
An ounce of profitable logic is worth a pound of efficient code.
  • #73
  • Aug 5, 2013 10:15am Aug 5, 2013 10:15am
Quoting TraderJak
Going slightly (?) off topic here but I was drawn to this tread through a search. I am just exploring the feasibility of backtesting a strategy on MT4 and the UTube video provided by my broker stated that I had to do the backtesting on an EA. I wanted to backtest on a indicator (a simple moving average cross) but don't know if this would need to be converted into an EA and whether it would need to action entry into trades. In reality I would still carry out due diligence in live trading and enter and exit trades manually - anyone any advice on this...
If you want to test a manual strategy in the Strategy Tester look up vhands, it allows you to place trades manually.
20 pips a day isn't too much to ask . . .
  • #74
  • Aug 6, 2013 12:07pm Aug 6, 2013 12:07pm
Quoting hanover
Good to see [Phil] posting again...
Judging by your signature, it's good to see you posting.

Quoting hanover
[My] EAs have been manually coded; I doubt that they could have been created using a code builder.
Not unless: (a) the EA's were relying on pattern indicators, etc. with well defined indicator buffers for use with iCustom, or (b) you used a builder having a custom code generation interface--building each condition bar by bar (yes, this actually exists), or (c) you were willing to type each entry condition manually into the builder's interface (a.k.a., partial manual coding).

Quoting hanover
We won't be running a Trade Explorer to prove this claim, nor sharing the concepts involved, nor selling the EAs at any price.
I suspect every profitable trader has a secret... or 4. My sincere congrats on 4, BTW. Even I have one EA that I refuse to share. Yes, the initial version was autocoded in an EA builder... and yes, I added code manually.
An ounce of profitable logic is worth a pound of efficient code.
  • #75
  • Aug 6, 2013 4:24pm Aug 6, 2013 4:24pm
Quoting RJo
{quote} Judging by your signature......
I agree that code generators can be useful, but I prefer to use my own code libraries. Software development has been my career since 1979, and over the last few years I've written more than 400 MT4 EAs and indicators. I find it easier to simply copy/paste modules whose logic I'm familiar with, use them as building blocks, and stitch them together myself. I've created my own templates for frequently used operations, and have also extended the MQL4 language by adding around 100 new functions, which are available here, free of charge, to anybody who wants them.

For traders who don't have programming experience, code generators are a good alternative IMO, provided that the trading rules and requirements aren't too complex. With more complex logic, it's almost certainly going to be necessary to add to, or modify, the generated code. And with many generators, the code may not necessarily be efficient or easily readable, especially to a non-programmer, which can compound the difficulties involved.
  • #76
  • Aug 6, 2013 4:53pm Aug 6, 2013 4:53pm
@hanover, thanks. I welcome any script/EA/indicator template you might be willing to share publicly on the review site.
An ounce of profitable logic is worth a pound of efficient code.
  • #77
  • Aug 7, 2013 5:55pm Aug 7, 2013 5:55pm
Quoting hanover
{quote} Phil, good to see you posting again, but I must beg to differ with you somewhat. My trading buddy and I have thus far developed 4 different trend-following EAs that are long term profitable. None of them hedge; they all use stoplosses productively. Granted, they generate a relatively modest income, but all win more than they lose, over time. There's no curve fitting involved, nor user intervention required; all run continuously, and simply use key levels and patterns for their entries, and then let profits accumulate on all surviving entries,...
Thanks Hanover; you are one of the few contributors I respect here on FF - both for your IQ and your EQ, so your welcome back means something and is appreciated.

I think it was RonaldRaygun who said (maybe even years ago) that he has 34 (I think it was) EAs in his arsenal; each suitable dependent on the market condition, and I have no idea if any were related to particular currency pairs given how broadly the pairs differ in volatility.

I will agree with you that for any particular market condition, it is conceivable to build an EA. WHEN to engage an EA correctly for that market condition would be the discretionary component necessary to succeed.

To have EAs that trade non-stop using stop losses and essentially therefore playing the probabilities of "wins being better than losses" on a daily basis, it is my guess that your EAs are basket traders, or alternatively multi-currency pair traders. If they are sufficiently but ALWAYS net profitable each day, then it does make sense to stick a million $ in an account for each of them, then let those puppies loose and check in on them from time to time.

I salute you for your success with them.

Phil
  • #78
  • Aug 8, 2013 10:00am Aug 8, 2013 10:00am
Quoting dancingphil
To have EAs that trade non-stop using stop losses and essentially therefore playing the probabilities of "wins being better than losses" on a daily basis, it is my guess that your EAs are basket traders, or alternatively multi-currency pair traders.
a.k.a. indirect hedging?
An ounce of profitable logic is worth a pound of efficient code.
  • #79
  • Aug 8, 2013 12:21pm Aug 8, 2013 12:21pm
Quoting RJo
{quote} a.k.a. indirect hedging?
I have no idea about Hanover's EAs and whether or not they have some aspect of arbitrage built into their designs.

My point was that multi-pair trading EAS cannot use significant lot sizes on any one currency pair. All of my work in hedging is always within the same currency pair; not arbitraging.

I trade 2.4 lots per $1,000 in my account, and I trade the Eur:USD. I commit the full weight of my account to the trade, and I am very focused on PA during the trade. Every pip means 2.5% equity gain to me.

Assume 24 pairs are traded with an EA, no more than 0.1 lots can be committed per pair to trade as heavy as I do. Probably half that lot size actually with stop losses being used, and ultimate success being an average of wins over losses being higher to some degree over time.

I don't know if I've made that easier to understand or complicated it.
  • #80
  • Edited 2:14pm Aug 8, 2013 1:08pm | Edited 2:14pm
@dancingphil, I'm PM'ing you.

@everyone, apologies for going off topic.
An ounce of profitable logic is worth a pound of efficient code.
0 traders viewing now