Applied Data Mining for Forecasting Using SAS

Appendix B

For and MA1:

Covariance at lag 0 (variance of y):

Cross terms drop because of the independence assumption: E[ε_tε_t–k] = 0    ∀k n.e.0

Covariance at lag 1:

Cross terms drop.

γ₁ = −θσ²

Covariance at lag 2 (and all higher lags):

Cross terms ...