When you specify a seasonal MA (1) term is this ESTIMATE Q=(12) ? Does this matter if you have an additive rather than a multiplicative model (in the later case you have a spike at lag 13 I am told)? In some sources I see suggestions that a multiplicative model has spikes at 12 and 13 and additive only at 12 (in theory both have a spike at 11 but this is often not signficant in practice). Thus a multiplicative model would be Q =(12,13) and an additive model would be Q = (12)
I have the same basic question with seasonal differencing. When you specify a difference in the IDENTIFY statement is there any difference if the seasonal effect is multiplicative versus additive? I think a seasonal difference that is multiplicative is (12,13) but an additive model is (12)
I have the same basic question with seasonal differencing. When you specify a difference in the IDENTIFY statement is there any difference if the seasonal effect is multiplicative versus additive? I think a seasonal difference that is multiplicative is (12,13) but an additive model is (12)
