InsomiaFX Correlation Double Hedge EA | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Hedge and Correlation Strategy 4,244 replies

MT4 MTF Correlation table and Correlation oscillator indicators 155 replies

Double it Then Double it Again - Swing Trading 4H 0 replies

need a programmer to program an alert for double tops and double bottoms 10 replies

Double Tops & Double Bottoms - How do you tackle them? 0 replies

Attachments: InsomiaFX Correlation Double Hedge EA
Exit Attachments

InsomiaFX Correlation Double Hedge EA

  • #41
  • Jun 2, 2013 7:02pm Jun 2, 2013 7:02pm
I can't read code, so I have no idea what this EA does.
I've modified the code by going through it and removing the "m"s, so it matches my broker's. EA still doesn't open any trades though.

But aside from that, I am trying to understand the reasoning behind this and what has to happen before the strategy opens or closes a trade.

I am having a problem understanding your spreadsheet Insomnia.
From what I see, it goes like this

AUDJPY Long and
CADJPY Short are opened as a pair (1A and 1B)

At a later time
AUDJPY Short and
CADJPY Long are opened as a pair (2A and 2B)

The 2 longs are closed for some reason at the same time (1A and 2B)
This leaves you with 2 open positions CADJPY short and AUDJPY short
You wait for the price to drop 50 pips then close these positions.

From your earlier posts, I understood that the AUDJPY and CADJPY were opened and closed as a pair. Same with the other 2. This does not seem to happen in your spreadsheet.

What I don't understand is where the correllation factor comes into play
If I could understand the mechanism behind this EA, I could probably do a spreadsheet to test the feasibility with past data
Please Do Not PM Me With Coding Enquiries
  • #42
  • Jun 3, 2013 12:05am Jun 3, 2013 12:05am
As I couldn't get the EA to work on my current demo account I have downloaded an Exness Demo.
How long do you normally have to wait for it to open a trade? Does it only trade at certain times of the day?
Please Do Not PM Me With Coding Enquiries
  • #43
  • Jun 3, 2013 2:11am Jun 3, 2013 2:11am
I think that the problem I am having with getting this EA to work is to do with the DLL

Quote
Note: The rar contains a dll which simply boosts the MT4 process priority to high. Its not needed if you don't trust DLLs. Just uncomment the related code in the EA. If you like to use it, put the dll into the same folder as terminal.exe. Needs Windows 7 and .NET Framework 4.0. Source on request.

I have XP and I don't know how to uncomment the related code in the EA.
Please Do Not PM Me With Coding Enquiries
  • #44
  • Edited 11:24am Jun 3, 2013 11:08am | Edited 11:24am
Hi Gumrai,

the new DLL is now needed as it calculates the correlation for us. If someone knows how to code the correlation function in MQL4, we can remove the dll again.
I'm not a good MQL4 programmer Maybe test with a friend who has Win7 or use VMware to run a virtual machine with Win7.

I will rewrite the EA to include the correlation as an indicator to decide when to place orders. Currently the EA places orders immediately, means at a random correlation. Thats not working very well, still makes profits. But we miss opportunities and have pairs hanging around negative for too long.

In my Excel all 4 orders were placed at the same time. You can now modify the correlation for 1B and 2A. Once we simulate a profit, the Closure Sequence shows us how we have to close the 4 pairs to avoid a margin call due to negative balance (has to be done manual in Excel, reference the highest loss from the correct number in F2-5 etc). The spreadsheet or the screenshot simulates how a correlation effect with 1B and 2A turns Equity into profit. In this is example that happens when the correlation is -506 points (ticks). Now we add up the correlation difference for each pair and end up with 25 points profits which are 1000$ with 40 lots. In this case, all orders would be closed.

The next EA will wait for a specific correlation number (-1/1 or 0.5/0.5 0/0 ..) and place all 4 orders again if you like to trade with 2 Pairs. Otherwise same for 1 Pair only.

Below screenshots of one demo account which runs Pair 1 only (swap positive), around 14% Equity profits within 3 days

Attached Image (click to enlarge)
Click to Enlarge Name: 1.png Size: 63 KB


Attached Image (click to enlarge)
Click to Enlarge Name: 2.png Size: 91 KB


I closed the last pair manually for fun. 12.000 profits while I had my morning coffee at 4pm ..
  • #45
  • Jun 3, 2013 5:23pm Jun 3, 2013 5:23pm
Quoting InsomiaFX
Hi Gumrai, the new DLL is now needed as it calculates the correlation for us. If someone knows how to code the correlation function in MQL4, we can remove the dll again. I'm not a good MQL4 programmer Maybe test with a friend who has Win7 or use VMware to run a virtual machine with Win7. I will rewrite the EA to include the correlation as an indicator to decide when to place orders. Currently the EA places orders immediately, means at a random correlation. Thats not working very well, still makes profits. But we miss opportunities and have pairs...
I managed to get the EA to work. Well, it wasn't working but after I had shut down my computer for a while, when I started it up again, the EA was working. More about that later maybe.

Insomnia, I am quite a logical person and so I always like to understand the logic behind things.
I find that I cannot understand the logic in your spreadsheet.
The way i see it, it is impossible for all 4 orders to have been opened at the same time.
If AUDJPY was entered long at 97.795, the short would also be entered at 97.795 (ignoring the spread). But it wasn't, the short was entered at 97.483,
There was also 30+ pips difference between the long and short entries with CADJPY

Likewise with the closing, it is impossible that all were closed at the same time.
AUDJPY long was closed at 90.00, yet the short was closed at 89.494
Again, 50 pips difference between the closing of CADJPY long and short.

The only way that this is logical is if pairs 1 and 2 are entered at different times and the longs and shorts are also exited at different times.

Regarding the closure sequence
Shouldn't it look more like this ?

Attached Image (click to enlarge)
Click to Enlarge Name: Closing cropped.jpg Size: 54 KB


No matter in what order the trades are closed, the equity will always remain at 101,000
Please Do Not PM Me With Coding Enquiries
  • #46
  • Edited 6:36pm Jun 3, 2013 6:23pm | Edited 6:36pm
Hi Gumrai,

great that the EA works now. Before you test it too much, wait for my next release. It will contain correlation as a decision maker to place orders. Also I use only 1 Pair for now to make things easier.

With the spreadsheet, don't worry about the preset open prices. Just change them to anything you like. To make the sheet easy I set a 50 pip difference between AUDJPY and CADJPY (check B10). Attached a more simplified 1 Pair version. You would change only values in the blue boxes.

The whole point of this Excel thing is to:

1) simulate Drawdowns from correlation differences
2) See how the correlation needs to change for a pair with a set start price to reach a user defined TP. Assumed is that only 1B changes and 1A stays static. Not very realistic, but possible and makes it easy to see the differences and simulate correlation values.
3) See the value of each order if the pair stays open very long and the prices moved away a lot

My sheet is not "precise" in any way. Just good to get a basic idea. I don't know who's right with the closure sequence. I will setup a demo account and close each order manually and record each step so we can be sure of how is really works. Guess its too simple for me to get it lol

Attached Image (click to enlarge)
Click to Enlarge Name: Excel.png Size: 21 KB


Attached File(s)
File Type: xlsx InsomiaFX_Correlation.xlsx   11 KB | 504 downloads
  • #47
  • Jun 4, 2013 2:30am Jun 4, 2013 2:30am
Quoting InsomiaFX
Hi Gumrai, great that the EA works now. Before you test it too much, wait for my next release. It will contain correlation as a decision maker to place orders. Also I use only 1 Pair for now to make things easier. With the spreadsheet, don't worry about the preset open prices. Just change them to anything you like. To make the sheet easy I set a 50 pip difference between AUDJPY and CADJPY (check B10). Attached a more simplified 1 Pair version. You would change only values in the blue boxes. The whole point of this Excel thing is to: 1) simulate...
I don't know if it is so great that the EA works now
Not being a coder, I had to remove parts of the code and that may have messed it up.
It started working during a volatile part of the trading sessions
I don't know whether due to slippage or maybe my internet connection too slow, but some trades were closed with profit and some with loss. Also because AUDCAD was mainly rising, only Long AUDJPY Short CAD pairing ever made profit.
Problem is that with the pair re-entered as soon as they closed, less profit was made than the accumalating loss on the other 2 trades. This is because of the additional spread being paid each time.

Similarly, your trades that made 17%, because they are net AUDCAD made that profit because the AUDCAD rose. If you had cut out all the opening and closing trades, you would have doubled your profits.

I look forward to seeing your correllation findings. As you say, opening new trades immediately after closing is not great
Please Do Not PM Me With Coding Enquiries
  • #48
  • Jun 4, 2013 2:59am Jun 4, 2013 2:59am
Hi Gumrai,

the screenshots of the demo account which made profits had nothing to do with AUDCAD.

12 Pairs were closed due to correlation profits.

3x 1B had the profit
3x 1A had the profit

Thats a perfect example of the correlation swings. Exactly how it should be.
  • #49
  • Jun 4, 2013 6:08am Jun 4, 2013 6:08am
Quoting InsomiaFX
Hi Gumrai, the screenshots of the demo account which made profits had nothing to do with AUDCAD. 12 Pairs were closed due to correlation profits. 3x 1B had the profit 3x 1A had the profit Thats a perfect example of the correlation swings. Exactly how it should be.
Yes, but you are making profits because the underlying pair (AUDCAD) was rising. It is just fluctuations that allow you to enter and exit positions. The problem is that every time that you close and open new positions, you are losing on the spread.
If AUDCAD drops significantly, you will be in a position with such a high loss that no correlation discrepancies will allow you to to exit with a profit unless AUDCAD returns to the same level that it was when you entered the trade. If AUDCAD continues to fall, it will wipe out your account.
I note that since your last trades were opened in your above post AUDCAD has fallen and so has probably wiped out the profit that you made and is now in loss overall.
What is the situation with your account at the moment?
Please Do Not PM Me With Coding Enquiries
  • #50
  • Jun 4, 2013 6:52am Jun 4, 2013 6:52am
Gumrai, we talk in circles I think. Spread losses are normal, every trade has them. Personally, I don't care about spreads. My background is grid trading with a huge amount of open orders.

Any AUDCAD changes are unimportant if you use the 2 Pair Double Hedge. In a double hedge the drawdown will reach a maximum and stay at that maximum if things go very wild. Its a total offset into any direction. AUDCAD changed around 8% within the past 2.5 years. Thats not even worth to think about. Correlation can mix up a pair by 100% within hours. Thats where drawdowns come from.

The demo account is closed as I create a new demo account whenever I develop a new EA version. Long term testing will start once the EA is finished and someone checks the code for any code/logic errors.
  • #51
  • Jun 4, 2013 7:06am Jun 4, 2013 7:06am
Quoting InsomiaFX
Gumrai, we talk in circles I think. Spread losses are normal, every trade has them. Personally, I don't care about spreads. My background is grid trading with a huge amount of open orders. Any AUDCAD changes are unimportant if you use the 2 Pair Double Hedge. In a double hedge the drawdown will reach a maximum and stay at that maximum if things go very wild. Its a total offset into any direction. AUDCAD changed around 8% within the past 2.5 years. Thats not even worth to think about. Correlation can mix up a pair by 100% within hours. Thats where...

You didn't answer my question. The account that made 17,000$ profit, what is it at now?
You can't answer because you have closed the account? Why have you closed the account?
Please Do Not PM Me With Coding Enquiries
  • #52
  • Jun 4, 2013 7:19am Jun 4, 2013 7:19am
Quoting InsomiaFX
Gumrai,. Any AUDCAD changes are unimportant if you use the 2 Pair Double Hedge..
As I was responding to a post that you showed your profit from Long AUDJPY and Short CADJPY. 2 pair double hedge is not applicable.
Please Do Not PM Me With Coding Enquiries
  • #53
  • Jun 4, 2013 4:03pm Jun 4, 2013 4:03pm
Hi Gumrai,

I like to answer your post by quoting myself: "The demo account is closed as I create a new demo account whenever I develop a new EA version."

The reason it that a new EA deals with a trading account in a different way and thus would falsify any conclusions about the correctness of the system.
Gumrai, you are not a programmer like me. We think different. You may be right, I do not decline that. But I develop code which works. Always did.

It seems that after 7000 views this thread becomes a fight between two traders with a different but respectable mindset. Within the last days I spent more time to respond here than to improve the EA and its C DLL.

I have to act for the interest of my investors and will not continue to reply to this thread. I have shown evidence that my system generates profits and thats all needed.

I thank you all for your time to read through this, test the code I did provide, your feedback and ideas. I have not been proven wrong but reserve my right to keep any further developments private based on the lack of positive feedback.
  • #54
  • Jun 4, 2013 5:26pm Jun 4, 2013 5:26pm
InsomniaFX,

such a shame that one argumentative poster on the thread has dampened your very generous spirit of sharing. i for one was silently checking the thread with great interest and anticipation of the system, and i'm sure others were too, perhaps we should be more vocal. i do sincerely hope you will reconsider
  • #55
  • Jun 4, 2013 6:58pm Jun 4, 2013 6:58pm
I was not intending to fight with you Insomia. I was trying to be helpful. I was just pointing out what I perceived as flawed logic with the strategy in its current form.
With the 2 pair strategy, the relative strength between AUD and CAD is most certainly important.
You may not believe me, but I believe that every winning set up is basically due to AUDCAD rising about 12 pips. I also believe that this is true no matter whether it is 1A or 1B that makes the profit.
Maybe I am wrong, but what if I am right? Don't you think that would be an important piece of information to know with your ongoing development?

Please just do me a favour. Maybe add to your EA to open and close a separate small long trade with AUDCAD at the same times or maybe just print on the chart the value of AUDCAD at the time of opening and closing positions.
This would enable you to directly compare and see if the movement with AUDCAD has any effect on the success of the strategy.

Quote
I have not been proven wrong but reserve my right to keep any further developments private based on the lack of positive feedback.

I am sorry that my feedback is not welcome, it was intended to be useful to you.
Please Do Not PM Me With Coding Enquiries
  • #56
  • Jun 4, 2013 7:15pm Jun 4, 2013 7:15pm
Quoting blamshakk
InsomniaFX, such a shame that one argumentative poster on the thread has dampened your very generous spirit of sharing. i for one was silently checking the thread with great interest and anticipation of the system, and i'm sure others were too, perhaps we should be more vocal. i do sincerely hope you will reconsider
You may see me as argumentative.

You have been silently checking the thread with great interest and anticipation of the system

I have been spending time and energy on research and testing the theory in an excel spreadsheet while you have been waiting to exploit the end product of Insomia's work without any input whatsoever.
Yet you have the nerve to be critical of ME??!!

If my research exposes possible flaws, do you think that I should not mention it?
You see that my input is negative and argumentative.
At least I have made some input and if it proves to be useful, then that input is positive.

Please put in some effort on this project yourself before condemning me.

One thing that I do agree with you, is that Insomia has been kind enough to share and it would be a shame if he stopped posting now.
If he is to stop posting because he doesn't like what I have to say, then I will gladly stop posting.
He only has to say "GumRai, please stop posting in this thread, I am not interested in your concerns"
Please Do Not PM Me With Coding Enquiries
  • #57
  • Jun 4, 2013 7:22pm Jun 4, 2013 7:22pm
fair enough, you have been analysing and spending time on this, but i think your combative tone has disrupted the progress of this thread. do not take the criticism personally, but you can see that your tone has demotivated the OP from further sharing his work on here, rather than the concerns you had outright.


Quoting Gumrai
{quote} You may see me as argumentative. You have been silently checking the thread with great interest and anticipation of the system I have been spending time and energy on research and testing the theory in an excel spreadsheet while you have been waiting to exploit the end product of Insomia's work without any input whatsoever. Yet you have the nerve to be critical of ME??!! If my research exposes possible flaws, do you think that I should not mention it? You see that my input is negative and argumentative. At least I have made some input and...
  • #58
  • Edited Jun 5, 2013 5:13am Jun 4, 2013 7:43pm | Edited Jun 5, 2013 5:13am
Quoting blamshakk
fair enough, you have been analysing and spending time on this, but i think your combative tone has disrupted the progress of this thread. do not take the criticism personally, but you can see that your tone has demotivated the OP from further sharing his work on here, rather than the concerns you had outright. {quote}
I admit that my tone may have been a bit combative at times, especially my more recent posts.
I am sorry about that.

But I did ask what the current situation was with the account that he showed with the 14% gain.
I asked because I checked the value of AUDCAD at the time that the last trades were opened.
Now as the value of AUDCAD has failed to rise by 12 pips since that entry point, I believe that the trades would still be open. I also believe that the account would now have dropped by more than $30,000.
I was rather frustrated when Insomia announced that he had closed the account and not even given the balance it was at when he closed it.
To be honest, I felt that Insomia was avoiding the issue and not willing to admit that the trades had gone well into loss.
If the balance had fallen, and I'm almost positive that it did, then I would hope that Insomia will realise that my theories may have some substance.
Please Do Not PM Me With Coding Enquiries
  • #59
  • Jun 5, 2013 5:25am Jun 5, 2013 5:25am
I know that you don't want to see this.
I tried the 2 pair strategy Long AUDJPY and Short CADJPY
I closed the positions manually when they were over $25,000 negative and the EA immediately re-opened them
The new trades did make a profit, closed and then re-opened.

Attached Image (click to enlarge)
Click to Enlarge Name: AUDJPY CADJPY cropped.jpg Size: 82 KB


The current trade is losing over $38,000 at the moment, amounting to a total $63,000 loss in less than 2 days.
If you look at the AUDCAD chart for the period of these trades, you will see that it has been falling.
Goes someway to vindicating my observations.
Please Do Not PM Me With Coding Enquiries
  • #60
  • Jul 27, 2013 2:13am Jul 27, 2013 2:13am
Mr. Gum,

Understood you are trying to maximize profit here. However, by manually over-riding the EA you are defeating the purpose of the EA itself. Not everyone can arbitrarily manually intervene and know exactly when to do so. Your intervention may have been from experience, it also could have been luck.

InsomniaFX has invested a lot of thought, energy and time into this concept which I and others find fascinating.

By definition any correlation and any hedge trade requires and equal for equal (as much as possible) proxy instrument to trade against with. His logic sounds strong and well as his coding ideas.

To insist that your "method" or observation is more profitable does not add to the value of the thread or topic.

Let's say that you are absolutely right. And by manually intervening at the right time and by taking off correlated and hedged positions you would be more profitable all the time (if you could be right 100%).

The thread is titled..."Correlation Double Hedge EA".

Not ... "Correlation Double Hedge EA (with manual intervention to make it more profitable)".

Hopefully InsomniaFX decides to come back to the thread and share his EA and his observations.

Correlation and/or Hedge strategies mixed with grid trading is an interesting sort of hybrid trading model.
0 traders viewing now