Confusion in model representation

Question

Will the functional form of a Moving Average model having coefficients h = [1 0.2 -0.5] be

y(t) = 0.2e(t-1) - 0.5e(t-2)

Or

y(t) = 1+ 0.2e(t-1) - 0.5e(t-2)

score 2 · Accepted Answer · answered Sep 09 '14 at 03:30

2

There is also the current error term, so

$$y_t = e_t + 0.2e_{t-1}-0.5e_{t-2}$$

answered Sep 09 '14 at 03:30

While this is flagged because of its brevity, I think a short answer covers it in this case. – Glen_b Sep 09 '14 at 03:59
@Glen_b I left the question for a day, believing that someone would type this obvious answer. No one did, and I didn't want yet another question left in the unanswered pile just due to its triviality, so I posted this as an answer rather than as a comment. – Alecos Papadopoulos Sep 09 '14 at 04:15
I think you're right to do so. You answer was flagged (presumably automatically by a script), I voted it as okay, and left that comment so other people contemplating the issue would see my reasoning (and hopefully agree that it is okay as it is). – Glen_b Sep 09 '14 at 04:19
@AlecosPapadopoulos: Thank you for your answer. But I have a question which is that the 1 value in h stands for $e_t$ or is it just a convention to write 1? What if h =[0.5 0.2 -0.5] instead of having 1? – SKM Sep 09 '14 at 16:32
The values in the $h$ vector are the values of the coefficients of the corresponding error term. – Alecos Papadopoulos Sep 09 '14 at 16:36

1 Answers1