enter image description here Referring to the above image, I wanted to know how to get the covariance matrix $\sum$.

My understanding is, $A$, is our transformation matrice, such that

$\begin{bmatrix} X_1 \\ X_2 \\ \end{bmatrix}$ = $\begin{bmatrix} Z_1 \\ Z_2 \\ \end{bmatrix} \cdot A$

How do we get the covariance matrices from here? Would it be like this:

$\sigma_{1,1} = COV(X_1,X_1) = VAR(X_1)$ Since $X_1 \sim N(0,\sigma_1)$, $VAR(X_1) = \sigma_1$ then $\sigma_{1,2} = COV(X_1,X_2) = COV(\sigma_1 Z_1, \rho \sigma_2 Z_1 + \sqrt{1 - \rho ^2} \sigma_2 Z_2) = COV(\sigma_1 Z_1, \rho \sigma_2 Z_1) + COV(\sigma_1 Z_1, \sqrt{1 - \rho ^2} \sigma_2 Z_2)$

Am I along the right track? What do I do after this, if I am?


You're on the right track; you can simplify using $\operatorname{Cov}(aZ_1,bZ_2)=ab\operatorname{Cov}(Z_1,Z_2)$, also the fact that $\operatorname{Cov}(Z_1,Z_2)= 0$ since $Z_1$ and $Z_2$ are independent, and the fact that $\operatorname{Cov}(Z_1,Z_1) = \operatorname{Var}(Z_1)=1$, and so on.

But there is a more direct method. If $X$ denotes the column vector $(X_1, X_2)^T$ and similarly we write $Z=(Z_1,Z_2)^T$, then you've got $X=AZ$. The general formula for covariance matrix is $$\operatorname{Cov}(X)=E[(X-E(X))(X-E(X))^T] \;.$$ Plug in $X=AZ$ and use matrix algebra to simplify this to $$ \operatorname{Cov}(X)=A\left[E(Z-E(Z))(Z-E(Z))^T\right]A^T=A[\operatorname{Cov}(Z)]A^T\;. $$ By assumption, the $Z_1, Z_2$ are standard normals, so that the covariance matrix for $Z$ is the identity matrix, and the covariance matrix for $X$ is the matrix product $AA^T$.


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.