Hedge and Correlation Strategy | Page 11 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Correlation + Momentum Strategy? 6 replies

MT4 MTF Correlation table and Correlation oscillator indicators 155 replies

InsomiaFX Correlation Double Hedge EA 66 replies

Looking for EA for Correlation Strategy 6 replies

Hedge strategy (not a hedge fund) 13 replies

Attachments: Hedge and Correlation Strategy
Exit Attachments

Hedge and Correlation Strategy

  • #201
  • Edited Jul 1, 2009 2:33am Jun 30, 2009 11:48pm | Edited Jul 1, 2009 2:33am
Quoting T.E.Lawrence
New trade:

Attachment
Looked like perfect trade, right? -120 pips.

Now I see we need to incorporate money management and stop losses. Waiting for a bad trade to change into the good one doesn't work - the pairs cross back and you're still left with many red pips; even after 12 hours, a 5min TF trade doesn't look like something is going to change.
  • #202
  • Jul 7, 2009 7:46pm Jul 7, 2009 7:46pm
Here's an interesting link about correlations.
  • #203
  • Jul 8, 2009 5:42am Jul 8, 2009 5:42am
Hi all.
I've been developing an indicator that might help with this.

http://www.forexfactory.com/showthread.php?t=180981

It will show the size of the price difference between the 2 currencies in pips, as well as the relative directions of the pairs. You can set it to zero them out at a specified time (I choose 00:00 every day).
My thinking is that once you've established the typical behavior of the variance...or "spread", then you can enter the trade in relative safety.

It is very much a work in progress. I know Swingman and others wrote alot of code to try and work with this issue, but I think it was all too complicated. (for me anyway)

I went back over 2008 and 2009 to date, and (if my data is right, and there's no guarantee that it is) it seems that 70% of the time the two (GBPUSD and EURUSD) will come back together.

So, if the indicator is showing a 100 pip gap, or variance between the two, and it shows GBP is down -50 and EUR is down -25, (result=gap -25) and you know from past data that typically it will spread out -25 (made up numbers for illustration purposes), then you could safely buy GBPUSD and sell EURUSD.

Seems to work in theory, but I need to test it in the real world.

I'll try and turn it into an EA to simplify forward testing.

All the best
John
  • #204
  • Jul 8, 2009 5:46am Jul 8, 2009 5:46am
John, this is wonderful, I've been hoping to see something like this. Is there any chance you could post the compiled code (I think it is .ex4?) for this? I always have a hard time putting that code into a workable indicator.

Thank you again, and I look forward to forward testing with you.

Quoting Jm1941
Hi all.
I've been developing an indicator that might help with this.

http://www.forexfactory.com/showthread.php?t=180981

It will show the size of the price difference between the 2 currencies in pips, as well as the relative directions of the pairs. You can set it to zero them out at a specified time (I choose 00:00 every day).
My thinking is that once you've established the typical behavior of the variance...or "spread", then you can enter the trade in relative safety.

It is very much a work in progress. I know Swingman and others...
  • #205
  • Jul 8, 2009 6:21am Jul 8, 2009 6:21am
No problem, I will post the compiled code when I get home tonite.
  • #206
  • Jul 8, 2009 6:43am Jul 8, 2009 6:43am
Quoting Jm1941
No problem, I will post the compiled code when I get home tonite.
Thanks much.
  • #207
  • Jul 8, 2009 3:31pm Jul 8, 2009 3:31pm
Here you go.
Enter your pairs in the inputs and the timeframe and the number of bars youd like to display (it'll take a while for them to fill up).

There are some issues with it, and I'm not done yet, but its something to play around with for now. Keep in mind it will start from the current bar and then move forwards, so when you put it on initially you'll only get one bar.

Also don't forget to enter the date you want to "zero" them at

Best,
John.
Attached File(s)
File Type: ex4 JM_Histov006.ex4   4 KB | 627 downloads
File Type: mq4 JM_Histov006.mq4   3 KB | 883 downloads
  • #208
  • Jul 8, 2009 3:40pm Jul 8, 2009 3:40pm
Got it, very nice job. What is the "more bars" input?

Also, I just left the defaults, except made the time frame 60 (assuming that is 1 hour) and so I assume these numbers will reset daily at 00 GMT, correct?

Quoting Jm1941
Here you go.
Enter your pairs in the inputs and the timeframe and the number of bars youd like to display (it'll take a while for them to fill up).

There are some issues with it, and I'm not done yet, but its something to play around with for now. Keep in mind it will start from the current bar and then move forwards, so when you put it on initially you'll only get one bar.

Also don't forget to enter the date you want to "zero" them at

Best,
John.
  • #209
  • Jul 8, 2009 7:36pm Jul 8, 2009 7:36pm
Hi,
More bars *should* pertain to the amount of bars on the histogram that will be visible. So, if set to one, you will only ever see one bar. Set to 20, you'll get 20 bars, and then they'll pop off the chart as new ones come one.

Or maybe none of the above, I'm new at this

Yeah 60 would be 1 hour, 240=4 hours etc. However, since it's working from the tick data as it comes in, I actually don't think this matters at all.
  • #210
  • Jul 8, 2009 7:46pm Jul 8, 2009 7:46pm
Oops. Theres a problem with the date entry criteria. You can set the date in the mq4 file for now. I've rest it to today's 00:00 (FXDD).

Will try to post a fix shortly, but you can test with this one for now.
  • #211
  • Jul 8, 2009 7:47pm Jul 8, 2009 7:47pm
Here you go.
Attached File(s)
File Type: mq4 JM_Histov007.mq4   3 KB | 588 downloads
File Type: ex4 JM_Histov007.ex4   4 KB | 444 downloads
  • #212
  • Jul 8, 2009 8:41pm Jul 8, 2009 8:41pm
Fixed the date entry and a little problem with the math.
Please let me know if you encounter any problems.
Attached File(s)
File Type: ex4 JM_Histov008.ex4   4 KB | 664 downloads
File Type: mq4 JM_Histov008.mq4   4 KB | 938 downloads
  • #213
  • Jul 8, 2009 9:38pm Jul 8, 2009 9:38pm
Quoting Jm1941
Hi all.
I've been developing an indicator that might help with this.
when you are counting the spread from a static time point you will see only two parallel line that will never cross may be we should investigate the cointegration methods? But i think it's useless as well... it's good only on a static history. dumn, on a real data all of those methotods are useless...

The cointegration coefficient is counted on the history vectors. The vectors are normalized price data, based on highs and lows, so it will not work for real data anyway...
  • #214
  • Jul 8, 2009 9:59pm Jul 8, 2009 9:59pm
Why would you only see 2 parralell lines?
The price of one pair is not moving exactly with the price of the other pair, hence the varying spread value.

This histogram shows you exactly the variation between the two pairs.
If you buy one pair and sell the other when the value is 100 (for example), if/when the value goes to zero, you would have made 100 pips...

They're crossing quite a bit these days, which is why I'm interested in this approach.
  • #215
  • Jul 8, 2009 11:18pm Jul 8, 2009 11:18pm
Quoting Jm1941
Why would you only see 2 parralell lines?
It was not about your indicator. Have not tried it yet.

Quoting Jm1941
This histogram shows you exactly the variation between the two pairs.
If you buy one pair and sell the other when the value is 100 (for example), if/when the value goes to zero, you would have made 100 pips...
are you sure the value will return to zero?
  • #216
  • Jul 9, 2009 12:23am Jul 9, 2009 12:23am
Quoting decypher
It was not about your indicator. Have not tried it yet.



are you sure the value will return to zero?
If we were sure it would return to zero we would be millionaires soon! Disclaimer: There is risk in forex trading.
  • #217
  • Jul 9, 2009 4:23am Jul 9, 2009 4:23am
Sorry decypher, thought you were talking to me.

You're right, we don't *know* that it will return to zero, but it does more often than not. Last night it went from +22 to -79. Currently at -62.

So, my plan is to watch it and gather data, and build a stategy to suit...

Will keep you posted.
John
  • #218
  • Jul 9, 2009 4:35am Jul 9, 2009 4:35am
We definitely need to specify and clarify this strategy. Trading on the basis of 1st post, after few weeks, in my opinion it can only bring loss. Pairs tend to drift for long period of time and when they cross over you're left with red pips - on 5min chart even with 130 pips loss.

There are times when everything goes well and other times when you're losing without any clue why.
  • #219
  • Jul 9, 2009 6:38am Jul 9, 2009 6:38am
I'm not understanding the numbers that are presented on the indicator. Right now I'm showing (Eur/Usd 85) and (GBP/USD 167) for a -82. Shouldn't it be a plus 82?
  • #220
  • Jul 9, 2009 7:10am Jul 9, 2009 7:10am
Hi Dreamliner,
Yes if those are your numbers then it should be +82 not -82.
I thought I had that fixed. I'll have to take another look at it tonite.
2 traders viewing now