Hello, Could somebody please provide a more explicit solution to Q6i in the Sep 2022 paper? In the examiner's report it says "Suppose we have shown that e_t = x_t + Sum(1,t-1)[(-b)^i x_t-i]". Why is that given? Thanks, Alex
Hi Alex Did you purchase the ASET for CS2, that gives a fuller explanation? This was a tricky question. We need to show the likelihood in terms of b, if we rearrange the defining equation as e_t=X_t-be_t-1 and repeatedly substitute in we get to that formula. We need it as we are finding the likelihood of the error terms, which are normally distributed N(0,sigma^2). Thanks Andrea
Thank you Andrea, I'm re-taking this exam, so I only have the ASET up to April 2022. I've managed to prove e_t = x_t + Sum(1,t-1)[(-b)^i x_t-i] by repeatedly substituting as you mentioned above. Thanks a lot for your help. Alex