Anyone do forex coding?!

[Drazen]

Steve

What about a degree of randomness in Forex?

If it is same like on roulette some people here might beat it 

Cheers

[Redherring]

Please explain what you mean by "locked"? If it is copyrighted I will NOT touch it.

If you cannot attach the file, rename it from mq4/mq5 extention to a .txt file, this forum does allow people to attach text files.

In my opinion, the best chance to fulfil your request is to post it on a forex forum, not a roulette forum.

Copywrite indeed. Is a good indicator anyway

Roulette forum isn't a bad place specially in an off topic section, many here have done it and some here are coders so surely it's more relevant than another thread bitchin about forum members 

[Steve]

Members and balanced moderation make the forum what it is.

Forex is still just cause and effect.

[Still]

Copywrite indeed. Is a good indicator anyway

Roulette forum isn't a bad place specially in an off topic section, many here have done it and some here are coders so surely it's more relevant than another thread bitchin about forum members 

If it's copyrighted, you should be able to name it, and we can find it out on the web, and a description, either for free, or for fee. 

If it looks promising, someone here might undertake it. 

I've dabbled with MT4 and Tradestation/EZ language.  But time is short unless really promising. 

What i recommend is looking for a way to do everything out of Excel. 

You can coordinate Excel with MT4, but also, you can also test the indicator out on random roulette numbers.

Here is a link to an expert in putting indicators into Excel, and testing them. 

link:s://:.tradinformed.com/

He has how to e-books, spreadsheets filled with indicators, systems, and testing methods.   

Lot's of free vids, giving away some valuable info, if you consider a profit factor above 1.4 to be valuable (it is). 

[Steve]

No prediction method is better than understanding the cause of fluctuations. Otherwise you'd be trying to use the effect to predict the effect. It can work but nowhere near as well  Proper prediction requires understanding cause. For example, seeing news before investors creates opportunity.

[Still]

In the following video, the author mentioned in post #18 tests random entry against a couple of indicators, and was able to come up with profit factors of 1.04 to 1.09 (anything above 1.00 is better than break even) using average true range as a trailing stop, besides a couple of other very simple approaches.  More interesting than that were the 60%-73% win rates for various tests. 

Here's another of his vids on the same theme:

[Still]

Here he is discussing an Ichimoku (trend following) based strategy with profitable results.

Here is an article that goes with it, laying out the statistics:

link:s://:.tradinformed.com/3-profitable-ichimoku-trading-strategies/

[cht]

No prediction method is better than understanding the cause of fluctuations. Otherwise you'd be trying to use the effect to predict the effect. It can work but nowhere near as well  Proper prediction requires understanding cause. For example, seeing news before investors creates opportunity.

Disagree.  It's difficult to impossible to align a cause on Tick chart up to 1hr tf environment.  Yet it presents great trading opportunity.

Causality lends credence to efficacy but not a must have.

One only needs to know the what,  when and how,  why doesn't change a thing.

If the why is made a neccesity then it explains for why so much has been missed because you can't see the must have why.  Make the dough first,  leave the why to the academics which is mostly hindsight.

[Still]

Here again, someone has tested a random entry against some clever trade management.  This guy tested a trade management plan spelled out in a book by Van Tharp called Trading Your Way To Financial Freedom.  This one is interesting to me as i happen to have the book on my nightstand, and i do recall how Van Tharp had glowing reviews for this particular method from original author Tom Basso. 

The results were positive for random entry followed by a trailing stop based on average true range, using a 1% position sizing model. 

Assuming financial markets are random, this is promising. 

Here is another guy testing this theory against a live forex trading account, albeit a very small account, and very small leverage. 

He is picking up an average 3 pips a day, with a 67% win percentage using random entry, and an undisclosed, proprietary trade management that he will sell, along with a spreadsheet. 

Would it be possible to improve performance with more logical entries?

[Redherring]

If it's copyrighted, you should be able to name it, and we can find it out on the web, and a description, either for free, or for fee. 

If it looks promising, someone here might undertake it. 

I've dabbled with MT4 and Tradestation/EZ language.  But time is short unless really promising. 

What i recommend is looking for a way to do everything out of Excel. 

You can coordinate Excel with MT4, but also, you can also test the indicator out on random roulette numbers.

Here is a link to an expert in putting indicators into Excel, and testing them. 

link:s://:.tradinformed.com/

He has how to e-books, spreadsheets filled with indicators, systems, and testing methods.   

Lot's of free vids, giving away some valuable info, if you consider a profit factor above 1.4 to be valuable (it is).

have been playing with harmonics for a while and doing a search took me to the pz website (pointzero) where I saw this for something like $299. Google then found me a free download. The indicator has very recently been upgraded on the website but it has icustom code ready for developers to get stuck in. However, it looks like it wants license codes which I don't have as it's an older free downloaded copy.

As far as harmonics go, I've found this to be a really good one. I use M5, set alerts for audio and visual. When the alert sounds from a complete pattern I try scalp 10-20 pips. The alert has gone off up to 60 times a day (I usually have 9 windows open, none in particular each day) but when I place a trade I'm too slow, from the alert to me trading there is a good few pip difference. If auto trade on alert, it would be very good.

One day from 0800-1300 I completed 25 trades, 21 were successful but the 4 bad ones were stoploss errors - wrong amount or didn't set. Think there was 42 trades in total I could've done. 

I've attached it in txt format just rename to ex4 and it should be fine. I understand it will need decompiling in order to then make into an EA which would simply trade to a set number of pips profit or stoploss but obviously see for yourself

Cheers

[Still]

It looks like the Point Zero developer, Arturo, made this into an EA back in March 2014. 

The stats from a ten day test period appear at the end of the video. 

Looks like it averages 25 trades a day, for a profit factor of 2.5 which is very good. 

I don't see the EA for sale on his website, but for $299, you'd think it would be included. 

I recommend contacting him and see what he'd be willing to sell it for.

This may be too much for Excel. However, it may be possible to substitute random data in place of market data, that can then be consumed by MT4. 

Btw, the file you attached is empty. 

[Redherring]

That's the one. I have emailed him and he's keen to reply but not interested in making it into an EA

Sorry I should've checked the file! I'll try again later when at a computer

[Still]

That's the one. I have emailed him and he's keen to reply but not interested in making it into an EA

Sorry I should've checked the file! I'll try again later when at a computer

Maybe an EA is too much liability, too much upkeep, or maybe he doesn't want to dilute it with a lot of people acting on the same signals. 

But if the indi code is available with the purchase, any developer could put it together. 

My understanding is MT4 is able to pull data from files, for testing. 

It would be possible to format random data and disguise it as a consumable MT4 file. 

Now THAT would be interesting. 

[Redherring]

But if the indi code is available with the purchase, any developer could...

Agree. But I don't have spare funds to purchase, and don't even know if anyone would purchase, develop and then give out

[cht]

Check if the indi has buffers.  If there is then the EA can be coded,  no need to purchase. Else if on object,  you need the source code which will not be given out even with the purchase.