Study Notes - pg 29 Ch 8, Diebold Modeling cycles: MA, AR and ARMA Models

[Dr. Jayanthi Sankaran]

Hi David,

In the above, I do not understand as to how you derive the autocovariance function:

Gamma(tau) = E(y(t),y(t -tau) = E((epsilon(t) + theta*epsilon(t-1)*(epsilon(t - tau) + theta*epsilon(t-tau-1)) =
theta*sigma(square), tau =1, 0 - otherwise - how do you get theta*sigma(square)?

Thanks!
Jayanthi