1) Since the process is stationary, the variance is the same in each period. 2) Since e_t is white noise it cannot be correlated to past values such as y_t-1

Could you write which part of the equation is not clear?

i don’t think it would be relevant as we have sigma^2*Z*Z’. We have sigma^2 that is positive and ZZ’ that is positive. So their product is positive and the inequality is proven.

Thank you so much!

at minute 1:34 I substitute the c with c= mu(1-phi)

The expected value will be 0 as 0/(1-phi)=0

The other moments will be the same.