For example, rising buy USD/CHF most likely there will be a chance to rise sell GBP/USD...
It I to that there is a sense to work ONLY on one of pairs. Increase the size of a position and work.Corellation - it is my sore theme 
.
Hi There
Take a look at http://www.mataf.net/en/analysis-correlation.htm this might help you
regards
zup
The attached is an XL file I put together a while back.
It Mirrors Pairs that NORMALLY travel in opposite directions and Matches those that NORMALLY travel in the same direction.
Thank guys, interesting table.
The proposition I made above is just a Forex form of a famous mathmatic theorem, called Ramsey' theorem. I just want to show no matter the forex market is random or regulated it has to obey mathmatic theorem.
I observed 4 currencies, eur usd gbp jpy, they form 6 instruments mutually. Every day if one instrument goes up, I mark 1 on it, otherwise mark 0. Then I can have totally 2^6=64 different market conditions, if the market is random. However, I only observed 26 conditions in 427 observations, and 4 of the 26 only happenned once during the survey, can be considered non-repeatable. I expected only 12 of the market conditions are non-observable, because there have to be market regulations and economics considerations, if the market is random. The chances of the conditions I actually observed are variable.
This concludes my survey, the market is not random.
The proposition I made above is just a Forex form of a famous mathmatic theorem, called Ramsey' theorem. I just want to show no matter the forex market is random or regulated it has to obey mathmatic theorem.
I observed 4 currencies, eur usd gbp jpy, they form 6 instruments mutually. Every day if one instrument goes up, I mark 1 on it, otherwise mark 0. Then I can have totally 2^6=64 different market conditions, if the market is random. However, I only observed 26 conditions in 427 observations, and 4 of the 26 only happenned once during the survey, can be considered non-repeatable. I expected only 12 of the market conditions are non-observable, because there have to be market regulations and economics considerations, if the market is random. The chances of the conditions I actually observed are variable.
This concludes my survey, the market is not random.Hi Niva,
I agree. Markets are not Random. Banks are not throwing darts at charts on the wall. But I do think you need a larger array of Sample Data.
[example only]
What happens if Australia signs a deal with the UK giving Cable Strength and coincidentaly China decides to peg the Yuan giving USD Strength.
What would you expect GBP/USD to be doing. Ignoring it and only looking to the JPY and EUR to decide where it should be? I don't think so.
Here's a link to another little xl work I did a while back.
https://www.mql5.com/en/forum/175375
If you take the time to sudy it you will notice that there is equilibrium across the board. That's why the Triangular Arbitrage idea is based in good thinking but the markets adjust too quickly to make it worth while often enough. 
$0.02 CAD
correlation EA
....but no one Correlation EA.
Bongo
correlation and hedging
I've made a simple EA based on correlation and hedging
Please review this article and drop me a comment:
Hi,
Your simple EA is only selling.
I try to change:
see below;
total = OrdersTotal();
if(total < 1)
{
if(Buy==0)
{
RefreshRates();
OrderSend(Sym_1,OP_BUY,Lots,MarketInfo(Sym_1,MODE_ASK),Slippage,0,MarketInfo(Sym_1,MODE_ASK)+1000*Point,"Hedging",1234,0,Green);
RefreshRates();
OrderSend(Sym_2,OP_BUY,Lots,MarketInfo(Sym_2,MODE_ASK),Slippage,0,MarketInfo(Sym_2,MODE_ASK)+1000*Point,"Hedging",1234,0,Green);
}
else
{
if(Sell==0)
{
RefreshRates();
OrderSend(Sym_1,OP_SELL,Lots,MarketInfo(Sym_1,MODE_BID),Slippage,0,MarketInfo(Sym_1,MODE_BID)-1000*Point,"Hedging",1234,0,Red);
RefreshRates();
OrderSend(Sym_2,OP_SELL,Lots,MarketInfo(Sym_2,MODE_BID),Slippage,0,MarketInfo(Sym_2,MODE_BID)-1000*Point,"Hedging",1234,0,Red);
}
return(0);
Has anyone realised this? We pick any number of instruments, when we trading, there is always a trend co-relation within these instrument.
Take the number 6, for instance, we can propose that within these 6 instruments we can always have at least 3 instruments show same directions, up or down. Within 6 of them, at least 3 go up or down together at a same time point.