how to represent SARIMA model in eviews

[RICH2222]

how do i represent SARIMA (0,1,1)(0,1,1) 12 in eviews?

is it equation seas.ls d(rate,1,12) ma(1) sma(1) or

equation seas.ls d(rate,1,12) ma(1) ma(12) ma(13) or

equation seas.ls d(rate,1,12) ma(1) sma(12)

where rate is my data series
thank you

[EViews Gareth]

What frequency is your workfile?

[RICH2222]

it is monthly series

[EViews Gareth]

equation seas.ls d(rate,1,12) ma(1) sma(12)

[RICH2222]

thanks

so can i represent SARIMA (0,1,1)(1,1,1) 12 in eviews as equation seas.ls d(rate,1,12) ma(1) sar(12) sma(12)

and SARIMA (0,1,1)(0,1,2) 12 as equation seas.ls d(rate,1,12) ma(1) sma(24)

[EViews Gareth]

Yes, although an SAR without an AR term is the same as an AR by itself.

[RICH2222]

lastly so if a person represents d(rate,1,12) ma(1) ma(12) ma(13) as SARIMA (0,1,1)(0,1,1) 12 model it is totally wrong in eviews?

[EViews Gareth]

Yes.

[RICH2222]

so what name can i get to this eviews model d(rate,1,12) ma(1) ma(12) ma(13) ?

[EViews Gareth]

There is no ARIMA notation equivalent.

[RICH2222]

i want to generate a forecast plot by using my the model i developed in eviews but i want to generate the plot with my original data (rate series) not the series d(rate,1,12) i used to develop the model. how do i go about it? thank you

[EViews Gareth]

When you forecast from an equation, EViews asks you whether you want to forecast the original form, or the transformed data.

[RICH2222]

is it correct to say because the mean of the residuals of a model is approximately zero then the residuals are random and normal distributed eventhough the prob value of Jarque- bera statistic is less that 5%?
mean of residuals is -0.031182
prob value of Jarque-Bera is 0.
thank you

[startz]

No. The mean has nothing to do with normality.

[RICH2222]

if in a model where you have monthly dummies, you find that some dummies are not significant ie they have pvalues greater than 0.05, is it wise to drop them from the model?