Consider a joint probability distribution. I have a table where the price variations of the two variables and the joint distributions are represented. They are NOT independent.
How do I calculate the covariance? I know the formula E(X*Y)-E(x)E(Y)

but how do I use that practically? Thans!

2. Originally Posted by 0123
Consider a joint probability distribution. I have a table where the price variations of the two variables and the joint distributions are represented. They are NOT independent.
How do I calculate the covariance? I know the formula E(X*Y)-E(x)E(Y)

but how do I use that practically? Thans!
$E(XY) = \int xy ~p(x,y) dx~dy$

$E(X) = \int x ~p(x,y) dx~dy$

$E(Y) = \int y ~p(x,y) dx~dy$

If the distribution is discrete interpret the integrals as sums

RonL