If u and y have a bivariate normal distribution with zero me
If u and y have a bivariate normal distribution with zero means, show that cov(u^2, y^2) = 2[cov(u, u)]^2.
Solution
cov (u2,v2) = E(u2v2) -E(u2)E(v2)
Cov (u,v) = E(UV)-E(u)E(V) = E(uv) as mean of u and v are 0
Hence 2{ cov (u,v) }2 = 2{E(uv)}^2
But for any variable z, Var (z) = E(z^2) -{E(z)}^2
Hence cov (u2,v2) = E(u2v2) -E(u2)E(v2) = E(u2v2) -(varu)varv
=2{ cov (u,v) }2
![If u and y have a bivariate normal distribution with zero means, show that cov(u^2, y^2) = 2[cov(u, u)]^2. Solutioncov (u2,v2) = E(u2v2) -E(u2)E(v2) Cov (u,v) If u and y have a bivariate normal distribution with zero means, show that cov(u^2, y^2) = 2[cov(u, u)]^2. Solutioncov (u2,v2) = E(u2v2) -E(u2)E(v2) Cov (u,v)](/WebImages/14/if-u-and-y-have-a-bivariate-normal-distribution-with-zero-me-1018832-1761526722-0.webp)