QuadHedge cacus' way.

cacus · #1 (**permalink**) 06-11-2009, 05:19 PM

Hi there, i'm new in this community and I would really like to share my strat and of course the EA I made for this one.
This is a really simple to execute but really hard to understand strategy. By "understand" i mean that it's pretty hard to get to the concept of "getting profits from divergence/convergence in correlations".
I call this "Hedging the Hedge", or "QuadHedge", since it needs 4 orders opened at the same time to start to work. Those 4 orders are separately managed in 2 couples of orders.
Let me show you an example so i can explain it clearer:

1º Orders COUPLE:
Buy EU
Sell GU

2º Orders COUPLE:
Buy GU
Sell EU

When one of the couples (of orders) gets in profit and reaches the PipsTrigger value (because of a high divergence of correlation between both currency pairs) then both orders of that couple are closed by the EA in profit and the other couple is left opened until the divergence go down again.

In our example:
Suppose the couple Buy EU-Sell GU is in +100 profit, then we close these two...
So, our Balance has raised +100, but we still have a -100 P/L in open orders.
Then we wait till the correlation between both pairs (EU and GU) gets high again (sooner or later it will) and suppose we now have a P/L in open orders of -20, then we close all the remaining orders (2) and of course our Balance has raised 80 net pips.

It uses the difference in correlation (divergence) in EU and GU on Daily and 1H charts. See, in Daily these two pairs develop a very HIGH correlation (about 95%), but, in 1H TF the correlation go down about 28%, so, we get profits from short term divergence in correlation and get covered from loses by long term high correlation (low divergence in the long term).
This kind of method is pretty sure, because if you simply don't make money you are not gonna lose money either... (well yes, just the spread).
And of course, don't try this EA on a non-Hedge-allowed broker.

PipsTrigger is something like a TP stop, but not necessarily the same.
PipsInGain is the net pips you would like to earn on every cycle the EA does.

PipsTrigger, as said before, is the value that limits the maximum profit a couple can reach before closing this single couple up.
PipsInGain is the difference between the first couple closed profit and the current P/L in opened orders you want to wait for this cycle to close up (the whole orders closed up and you got "PipsInGain" pips in profit).
So, suppose you set PipsTrigger in 50 and PipsInGain at 10, this would happen:
The EA opens the QuadHedge and wait for any of the couples to reach PipsTrigger to close that one.
Suppose 1º Couple gets +50 profit and the EA close this up.
Then, the EA will wait until the open P/L reach 50-PipsInGain (10 in our example) value and close the remaining 2 orders to get a profit of 10pips easily and fast, without any risk.
So, understanding the meaning of Divergence and correlation then you may know that you just CAN'T set PipsInGain with exorbitant values because if you do so you will get into a really unknown risky thing.
If you set a PipsInGain and PipsTrigger close enough then you'll have an scalper EA, low gains each operations but more operations (cycles in this case). If you set the values higher then you will got a more efficient EA in profit/cycle relation, more gains per cycle but maybe less operations.
I personally choose 50-10 to get faster gains with less pips in order to avoid any possibility of big loses (if there is any).

Just in case, I attached a pic to show you how I see this method performs.
Follow the explanation below while you watch the pic:

You'll see at picture's up-left region that both couples where opened in a HIGH correlation moment. Then at 14:10 the correlation tend to undevelop, this is what we call "high divergence in correlation", this is when, in the info displayed up-right in the chart we would be seeing some of the couples going up in profit. In this case, the Short GBPUSD and the Sell EURUSD is the couple with profits (as you may clearly see). At a certain point (14:22) the correlation tend to develop again so the "divergence" is going low till 14:47 approx where the correlation reach the higher again.
As you may see, these are the moments to take in count.
HIGH divergence moments are for taking the wining couple in profit, low divergence moments are for taking the remaining orders closed so the net profit would be positive.
I see not very risky moments in this strat. At least not while correlation between two pairs keeps in the time. If a worldwide war (or something like this) affects some of the country in question then we obviously turn the EA off or simply look for another two pairs with good correlation.
Am I clear enough?

Maybe we can work on a discussion on this method here to make things even more interesting.
Let me know what you think in any case.
I hope you like it and i'll be glad to receive some feedback from you all.

If someone wants to discuss about the source or would really like to take a look/colaborate at/in this please PM me and I will have no problems at all.

Regards.

cacus

cacus · #2 (**permalink**) 06-11-2009, 08:52 PM

Wow, sorry didn't find the EDIT thread button, here's a new version with a little change that was possibly generating erratics order closing.

bobsters · #3 (**permalink**) 06-11-2009, 10:27 PM

Looks very interesting! Thanks, i'll try it out! Just waiting for your attachment to be approved (it doesnt let me download it at the moment)

zupcon · #4 (**permalink**) 06-11-2009, 10:34 PM

This is an interesting strategy. Rather than closing the second pair for breakeven or a small loss when the short term correlation reverts back, have you tried just keeping the second pair open, and opening 2 new orders to start the cycle again ?

You'd at least save the spread which might be significant.

prasxz · #5 (**permalink**) 06-12-2009, 01:12 AM

Thanks for share your EA Cacus , that's interesting system
How about a drawdown ?

===================
Forex Indicators Collection

freedom131477 · #6 (**permalink**) 06-12-2009, 01:48 AM

Quote:

Originally Posted by cacus

Hi there, i'm new in this community and I would really like to share my strat and of course the EA I made for this one.
This is a really simple to execute but really hard to understand strategy. By "understand" i mean that it's pretty hard to get to the concept of "getting profits from divergence/convergence in correlations".
I call this "Hedging the Hedge", or "QuadHedge", since it needs 4 orders opened at the same time to start to work. Those 4 orders are separately managed in 2 couples of orders.
Let me show you an example so i can explain it clearer:

1º Orders COUPLE:
Buy EU
Sell GU

2º Orders COUPLE:
Buy GU
Sell EU

When one of the couples (of orders) gets in profit and reaches the PipsTrigger value (because of a high divergence of correlation between both currency pairs) then both orders of that couple are closed by the EA in profit and the other couple is left opened until the divergence go down again.

In our example:
Suppose the couple Buy EU-Sell GU is in +100 profit, then we close these two...
So, our Balance has raised +100, but we still have a -100 P/L in open orders.
Then we wait till the correlation between both pairs (EU and GU) gets high again (sooner or later it will) and suppose we now have a P/L in open orders of -20, then we close all the remaining orders (2) and of course our Balance has raised 80 net pips.

It uses the difference in correlation (divergence) in EU and GU on Daily and 1H charts. See, in Daily these two pairs develop a very HIGH correlation (about 95%), but, in 1H TF the correlation go down about 28%, so, we get profits from short term divergence in correlation and get covered from loses by long term high correlation (low divergence in the long term).
This kind of method is pretty sure, because if you simply don't make money you are not gonna lose money either... (well yes, just the spread).
And of course, don't try this EA on a non-Hedge-allowed broker.

PipsTrigger is something like a TP stop, but not necessarily the same.
PipsInGain is the net pips you would like to earn on every cycle the EA does.

PipsTrigger, as said before, is the value that limits the maximum profit a couple can reach before closing this single couple up.
PipsInGain is the difference between the first couple closed profit and the current P/L in opened orders you want to wait for this cycle to close up (the whole orders closed up and you got "PipsInGain" pips in profit).
So, suppose you set PipsTrigger in 50 and PipsInGain at 10, this would happen:
The EA opens the QuadHedge and wait for any of the couples to reach PipsTrigger to close that one.
Suppose 1º Couple gets +50 profit and the EA close this up.
Then, the EA will wait until the open P/L reach 50-PipsInGain (10 in our example) value and close the remaining 2 orders to get a profit of 10pips easily and fast, without any risk.
So, understanding the meaning of Divergence and correlation then you may know that you just CAN'T set PipsInGain with exorbitant values because if you do so you will get into a really unknown risky thing.
If you set a PipsInGain and PipsTrigger close enough then you'll have an scalper EA, low gains each operations but more operations (cycles in this case). If you set the values higher then you will got a more efficient EA in profit/cycle relation, more gains per cycle but maybe less operations.
I personally choose 50-10 to get faster gains with less pips in order to avoid any possibility of big loses (if there is any).

Just in case, I attached a pic to show you how I see this method performs.
Follow the explanation below while you watch the pic:

You'll see at picture's up-left region that both couples where opened in a HIGH correlation moment. Then at 14:10 the correlation tend to undevelop, this is what we call "high divergence in correlation", this is when, in the info displayed up-right in the chart we would be seeing some of the couples going up in profit. In this case, the Short GBPUSD and the Sell EURUSD is the couple with profits (as you may clearly see). At a certain point (14:22) the correlation tend to develop again so the "divergence" is going low till 14:47 approx where the correlation reach the higher again.
As you may see, these are the moments to take in count.
HIGH divergence moments are for taking the wining couple in profit, low divergence moments are for taking the remaining orders closed so the net profit would be positive.
I see not very risky moments in this strat. At least not while correlation between two pairs keeps in the time. If a worldwide war (or something like this) affects some of the country in question then we obviously turn the EA off or simply look for another two pairs with good correlation.
Am I clear enough?

Maybe we can work on a discussion on this method here to make things even more interesting.
Let me know what you think in any case.
I hope you like it and i'll be glad to receive some feedback from you all.

If someone wants to discuss about the source or would really like to take a look/colaborate at/in this please PM me and I will have no problems at all.

Regards.

cacus

PM is not allowed and i like to have look in code Possible?

cacus · #7 (**permalink**) 06-12-2009, 04:03 AM

Quote:

Originally Posted by zupcon

This is an interesting strategy. Rather than closing the second pair for breakeven or a small loss when the short term correlation reverts back, have you tried just keeping the second pair open, and opening 2 new orders to start the cycle again ?

You'd at least save the spread which might be significant.

Yes, you are correct, and seems like you got the system logic really fast

You indeed can do that and keep the cycles going on but i think that this could be dangerous taking into count that a discorrelation may happen some day and that would result in a huge loss. But, while the daily correlation keeps the same this is pretty sure system

cacus · #8 (**permalink**) 06-12-2009, 04:08 AM

Quote:

Originally Posted by prasxz

Thanks for share your EA Cacus , that's interesting system
How about a drawdown ?

===================
Forex Indicators Collection

I'm currently testing this strat, seems like for now it's just getting profits.
The larger DD i think this can have is the spread (-12) in the case of EU/GU hedge on FXDD (2 and 4 pips spread respectively).
So, seems until now that this is a pretty low risk method.
Thanks you for the interest, don't doubt if you have any more questions.
Regards.

cacus