0
$\begingroup$

I have three variables X, Y, Z. and I know $Corr(X,Y), Corr(X,Z),Corr(Y,Z),Var(X),Var(Y),Var(Z)$. and Cov is Covariance, Corr is correlation, Var is variance, $Corr(X,Y)=\frac{Cov(X,Y)}{\sqrt{Var(X)*Var(Y)}}$ Can I get $Cov(XY,Z)$

$\endgroup$
  • $\begingroup$ Is cov covariance? $\endgroup$ – Rohan Dec 5 '16 at 12:50
  • $\begingroup$ Yes, Sorry I didn't state clearly, and Corr is correlation, Var is variance, $Corr(X,Y)=\frac{Cov(X,Y)}{\sqrt{Var(X)*Var(Y)}}$ $\endgroup$ – Lexi Dec 5 '16 at 12:57
0
$\begingroup$

You do not seem to have enough information. It can be shown (I used a developmental version of mathStatica here to automate, but one could do this manually) that:

$$\text{Cov}(XY, Z) = \mathbb{E}[X] \text{Cov}(Y,Z)+\mathbb{E}[Y] \text{Cov}(X,Z)+\mu _{1,1,1}$$

where $$\mu _{1,1,1}=\mathbb{E}\left[\;(X-\mathbb{E}[X]) \; (Y-\mathbb{E}[Y]) \;(Z-\mathbb{E}[Z]) \; \right]$$

You know all the bivariate covariances, but you would still need to know the raw means $\mathbb{E}[X]$, $\mathbb{E}[Y]$ and the product central moment $\mu _{1,1,1}$ or ... the product raw moment $\mathbb{E}[X Y Z]$ and $\mathbb{E}[X]$, $\mathbb{E}[Y]$, $\mathbb{E}[Z]$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.