Deriving a formula from an ARIMAX Model

[PimTue]

Hello all,

I am finishing my master thesis and have included an ARIMAX model in the research. Now I would like to make some manual calculations based on the results from the ARIMAX Model in SPSS. I know that the results are not all significant but this is not a concern because a simulated example is used so the significance is not that important.

My question, for now, is how to derive a formula from this model (and which formula to use), and to use the results of the model in SPSS to make some manual calculations for future values of revenue.

Does anyone know a solution for this? Suggestions would be greatly appreciated!

 

[hlsmith]

I haven't done much of this, but once you think you have the formula - can you use the model coefficients to attempt to recover the values in the estimated forecasts?

@noetsi - any comments?

 

[noetsi]

What do you mean by ARIMAX? Multivariate Arima?

What formula are you trying to derive? The functional form of the predictors?

Given that this method, if that is what you mean by it, is not used all that much in my experience I don't think the functional form or showing the model is all that important (to show that is). You are interested if you predict correctly and if there is a relationship.

 

[PimTue]

Yes, an ARIMA model with multiple independent variables. I am trying to derive the formula of this ARIMAX or multivariate ARIMA model with the given parameters or estimates from the model in SPSS, because my supervisors of the university would like to see me make some manual calculations for future values to see if I understand the model well enough.

I have found some general notations of the formula for the ARIMAX, however, I don't exactly know how to use this to make calculations based on my results.

I am using an ARIMAX (0,1,1) model and the parameters/estimates are given in the picture above...