Download MetaTrader 5

Looking for a partner to help code a trading sytem

To add comments, please log in or register
Mongoose
114
Mongoose  

Hello all;

I am Looking for a partner to help me (eventually) create an EA. I have created a unique profitable trading system, and have come to the end of how much better I can make it.

Unfortunately the 4 indicators that the system uses are not optimally coded to do exactly what I need them to do. To be honest I don't even know if some of the things I need done are even possible in MT4. One such problem is taking the angle of a Linear Regression channel, turning that into a variable,(stated in degrees) and being able to use that as a basis for SL's and TP's.(plus or minus a few pips) If that cannot be done, I hope that it would be possible to at least take the last value of the top/bottom of the channel and divide that by the first value of the top/bottom of the channel, thereby giving a value which would give an indication of how much the channel is actually tilting, up, down, or not at all. I do have a work around for this if it is not possible to do either.

To make things even more difficult, none of this is done on a time based chart,(I use median renko bars) however I would need to pull variables from a time based feeder chart and display and manipulate them on the offline chart. When the EA is actually created, all the trades would be based off the indicators on the offline chart, using a few variables from the live chart. I am sure you know that while median renko bars are not time based, the values of the regression channels are constantly changing on a per tick basis, as data is fed from the feeder chart.

I designed it to trade whether in a trend or in a range, although when the pair is ranging it does not trade as much, because there is not enough pips to be had between the outside lines of the channel. Risk/Reward is actually controlled within the system, if there is not enough of a potential reward the system wont open a position. The bare minimum is 1:1 and that only happens off a bounce with both K% and D% of the Stoch in agreement, as well as the SSRC.(normal Risk/Reward is 1:2.85, I have read that pro traders like 1:3, however 1:2.85 is the best I could do, realistically) Of course Risk/Reward is honestly based off the pair you are trading, right now the 50 day average range of USD/CAD is 47 pips(bread and water) EUR/JPY is 187 pips.(steak and potatoes)

To give you a rough idea of what needs to be done:

Two indicators need to be optimized/changed, should be simple as it is a daily H/L/O/C and a default linear regression channel.

A custom indicator will then need to be created to display merged data from those two indicators, on the offline chart.(Along with the variables from the feeder chart)

The custom Stoch indicator that I have needs to be overhauled.

 

The custom SSRC indicator that I have needs to be adjusted.

 A custom indicator would need to be created to incorporate the system into symbols that would show the state of the indicators, along with a simple Buy/Sell/No Trade, signal.

I would need some time to trade the system manually with real money.(my money) I do not back test, back testing is deceptive in nature: no spreads, no news, no “bank burst” buying or selling, that is why I do not back test. I only trade with micro positions when testing, 10 cents a pip, so live testing is really not that big of a deal.

If everything is good at this point, the work can begin on the EA. I think this will be the most difficult part. Knowing when to open or close a position is simple compared to the task of actually optimizing the automation of managing the trade. If you have used renko or median renko then you know that during news the chart can print 10-20 bars within a few seconds,(and that is using 6-10 pip bricks) having the software deal with this in real time will be difficult. The news is not that hard, you can always turn off the EA on big news releases, but you cannot know the future time of world events,(NK launching a missile, or a terrorist action, as examples) and you cannot know the time of a bank taking a position with a few billion dollars. We can both use a demo account for the EA, when it is created, to compare settings and see which are best for automated, live trading.

So if anyone, that is extremely knowledgeable with coding indicators and EA's for mt4, is interested just leave a reply. This is not a project that will be completed overnight, or in a few days, however I will promise not to waste your time as I will keep working until the system can be decently profitable, and fully automated. I have spent over 480 hours on this and have tested it on EUR/USD, EUR/JPY, GBP/USD and USD/JPY. As I stated the system is profitable, it is just extremely difficult to trade this “as is” manually, because none of the indicators do exactly what I need them to do. Some do too much, which add info/variables I don't need. Some are horribly programmed, which have loops that lock up the cpu,(I have a quad core proc and mt4 won't use more then one of them at a time) and some do not do enough.

I also do not want to scare anyone away that would like to help out, because this sounds too difficult. The system only uses 4 indicators: Daily H/L/O/C, Stochastics, SSRC, and a Linear Regression Channel. I understand that Stochastics and SSRC are similar, however I use the SSRC as a filter, as I do not think it will be possible to take angles from the regression channel. The system itself is pretty simple, it is just that the indicators need adjusting and coded correctly, and it needs automated.

So if anyone wants to help out, but do not feel that can do the whole thing, I will still provide you with a copy of the end result for your part in helping.

Ubzen
5394
Ubzen  
Mongoose
114
Mongoose  

  Thanks for the advice, however if I just wanted to buy everything, I already have an individual that codes extremely well:

 

 Moderator edit:  URL Removed

 

  Has done multiple indicators for me, a little pricey, but professional.  The reason I did not go to him first is because this is for adjusting multiple indicators, as well as creating multiple custom indicators, and an EA.(as well as helping test it)  If it comes down to it, I will eventually pay to have it all done, but then I certainly won't share it here.  I thought this website was a good place to find a programmer that wanted to help with a project for both the mutual benefit of them and myself.

   By the way he has multiple, profitable EA's that he is willing to sell, however the prices start at a few thousand Euro each, I would rather create something myself.(besides the wife wont allow me to spend that, LOL)  I thought about just learning how to code myself as I do have a programming background, however that would take a few months to learn and I have already spent almost 500 hours on this as it is, but I guess the option is always there.  I will give this thread a week, if there is no interest I will move along and progress it elsewhere.  Thanks for the advice though, always happy when someone tries to help out.

Simon Gniadkowski
Moderator
18018
Simon Gniadkowski  
Mongoose:

 

  Thanks for the advice, however if I just wanted to buy everything, I already have an individual that codes extremely well:

 

URL REMOVED

Look here: https://www.mql5.com/en/users/register

It says . . . "4. Posting of advertising messages is forbidden. "
Simon Gniadkowski
Moderator
18018
Simon Gniadkowski  
Mongoose:

 

  Thanks for the advice, however if I just wanted to buy everything, I already have an individual that codes extremely well:

 

 Moderator edit:  URL Removed

 

  Has done multiple indicators for me, a little pricey, but professional.  The reason I did not go to him first is because this is for adjusting multiple indicators, as well as creating multiple custom indicators, and an EA.(as well as helping test it)  If it comes down to it, I will eventually pay to have it all done, but then I certainly won't share it here.  I thought this website was a good place to find a programmer that wanted to help with a project for both the mutual benefit of them and myself.

Post your data showing that your strategy is consistently profitable and I'll consider your proposal.  I assume you do have back test data ?
Mongoose
114
Mongoose  

  Sorry did not see that as advertising...  He posted a link for me to buy something, is that not advertising?(by the way you did not edit his post)  I simply let him know if I wanted to buy something, I already had a source.  I do not back test as back testing does not include spreads, news, or random events.  I test by trading with a live account, micro account so there is no great loss trading @ 10 cents a pip.  I would consider showing my live account, how much data do you need?

 

EDIT TO ADD:  Here I was in a trade, when you asked, so I closed the trade.  Yeah it is cherry picked, got too have humor in life to enjoy it :) 

 

 

8559055

2013.04.17 05:11

sell

0.30

eurjpy.

129.503

129.353

0.000

2013.04.17 05:36

129.279

0.00

0.00

0.00

6.84

8559202

2013.04.17 05:36

balance

Credit for close #8559055

0.00

Simon Gniadkowski
Moderator
18018
Simon Gniadkowski  
Mongoose:

  Sorry did not see that as advertising...  He posted a link for me to buy something, is that not advertising?(by the way you did not edit his post)  I simply let him know if I wanted to buy something, I already had a source.  I do not back test as back testing does not include spreads, news, or random events.  I test by trading with a live account, micro account so there is no great loss trading @ 10 cents a pip.  I would consider showing my live account, how much data do you need?

The link that ubzen  provided is just the same as the one at the top of the page called  Jobs,  it is a MetaQuotes organised service that brings together customers for and providers of coding services.

A few years across a few instruments would suffice as long as it is all using the same strategy applied in the same way. 

Mongoose
114
Mongoose  

  Fair enough, since I test live, what you are asking, would take a few years, as old as I am, I would rather not wait that long.  By the way, if you have read the original post, and noticed the sell, you would see that the 1:2.85 risk/reward was met, the Stoch did not turn up until 128.990, 50 pips for a risk of 14 pips.  Give the thread a week, if no one is interested, I will go elsewhere, was just trying to present an opportunity, if anyone was interested. 

Simon Gniadkowski
Moderator
18018
Simon Gniadkowski  
Mongoose:

  Fair enough, since I test live, what you are asking, would take a few years, as old as I am, I would rather not wait that long.  By the way, if you have read the original post, and noticed the sell, you would see that the 1:2.85 risk/reward was met, the Stoch did not turn up until 128.990, 50 pips for a risk of 14 pips.  Give the thread a week, if no one is interested, I will go elsewhere, was just trying to present an opportunity, if anyone was interested. 

So back test and get the data quicker . . . a few months of data does not prove consistency.

R:R on it's own is meaningless . . .  you can have a Risk of 10 and a reward of 100 and still have a losing strategy,  conversely you could have a risk of 100 and reward of 10 and have a winning strategy.  

Mongoose
114
Mongoose  

  To back test would I not have to have an EA to begin with... 

Ubzen
5394
Ubzen  
RaptorUK:

So back test and get the data quicker . . . a few months of data does not prove consistency.

R:R on it's own is meaningless . . .  you can have a Risk of 10 and a reward of 100 and still have a losing strategy,  conversely you could have a risk of 100 and reward of 10 and have a winning strategy.  

I don't think he realizes that mql4.com gets allot of these requests from people who provides even stronger details of their manual systems. These requests goes un-fulfilled because most programmers don't have time to program, de-bug, back-test, demo-test and penny-test other peoples systems.

If I could make these requests and someone complied, I wouldn't have spent 3_years on this forum learning, and hours building my own stuff. I guess the absence of time can make people consider smarter ways of shifting work. But smart or not, its un-likely. The people asking for partnerships would never agree to manually-trade for the programmer [something they can do]. But thinks someone programming for them is a fair-deal.

123
To add comments, please log in or register