Correlated currency pair deviation

Hello, I am working on strategy that would require me to be equipped with the following knowledge. I know there is no black and white answer.

The correlated pairs move with or opposite of one another but some times they deviate from this correlation (statArb).
My question is, is there anyway i can estimate how much they deviate in terms of pips. (i know the percentile the have for correlation but I want some thing that I take two pairs and back the deviation in terms of pips.

Thanks.

@theCyberGuy
i frequently use correlating pairs, they are a good indicator

Like you said there is no Black and White answer here its all about testing , some pairs correlate better than others.
Take NAS100 Correlates with US2000 very well

but then when i’m trading EUR AUD, i use GBPUAD to correlate in the same direction
and the opposite direction is AUDUSD and AUDCHF
now these correlate to a lesser degree, but when they line up it’s a pretty clear signal as to the direction of the EURAUD

and i have the M1’s correlate
and then i have the M30 correlate as well
now when all these line up, i know what the EUR AUD is doing

so… how do you get to this point
MATE… YOU SIMPLY HAVE TO TEST DIFFERENT PAIRS AND SEE HOW THEY ALIGN WITH EACH OTHER
there really is no other way

1 Like

Thanks for the reply. I just dont want to see if they are correlated. Its one of the 4 module in my trading system. So i need to know on average ( by how many pips) do they DEVIATE.

i understand , and my advice was TEST IT OVER TIME.
get some correlating pairs that you’re working with and test them over time and measure the deviations and then average them.
TEST IT

1 Like

Ohh boy!! I knew it but was in denial. Lets see if someone else have some creative approach.

Use this:

https://www.oanda.com/forex-trading/analysis/currency-correlation

yeah, for sure,
but let’s say you find someone who gives your answer and they say they tested it over , say 3 years while taking sample points at a frequency of 1 Sample per hour

ok… that’s great
but how do you KNOW 100% the data is reliable
Fact is… You don’t

it sux , i know

1 Like

Yes you are in denial. There is simply no way to answer this question other than for you to make a hypotheses, figure out how to test that hypotheses, test and analysis the resultant data. This is our job.

@_bob hey bob, You’re from W.A. right ? hehe
You’ll appreciate this…

I just got off the phone from a company (and it’s good for everyone to know this… even if it off topic)

it’s a company called EZY DIGITAL MARKETING
indian sounding guy,
Basically the pitch is… they send 4000 SMS messages each day and build you a client base

THEY CLAIM TO BE BASED IN W.A.

They charge Between $330 - $680 a month

the guy couldn’t pronounce HUME HWY correctly hehe he called it HUMEY HIGHWAY
so i said… You’re not in Australia are you ?
he swore black and blue, he was

apparently this is the address

now, let’s have a look at google maps hehe
You gotta love it
NICE UTE MATE :stuck_out_tongue: (The company car i imagine hehe)

Where does the BS Stop huh

he actually tried to convince me, HE WORKS FROM HOME AND IT’S A LEGITIMATE BUSINESS.

bob, i don’t suppose you’re near this guy and you could go over and give him a smack in the head for me hehe ?

A venture of SPEEDY INNOVATIONS based in Kolkata, West Bengal, India.

Walks like a duck.

Boddington, just down the road bro. Gold mining town. Not much room for Indian scam artists there. Shouldn’t be to hard to find!

1 Like