State Space Formulation in Eviews 5

[Libes]

Hi,

I am trying to carry out temporal disaggregation of quarterly GDP series into monthly series using 2 exogenous monthly variables (IP, RS). I have come up with 2 formulations. There are 2 versions of each formulation. These are my Eviews codes:

1st Formulation (1st Version)

@signal GDP_q=(GDP_m1+GDP_m2+GDP_m3)/3+[Var=exp(c(1))]

@state GDP_m3=c(5)+c(6)*IP+c(7)*RS+[Var=exp(c(2))]

@state GDP_m2=c(5)+c(6)*IP(-1)+c(7)*RS(-1)+[Var=exp(c(3))]

@state GDP_m1=c(5)+c(6)*IP(-2)+c(7)*RS(-2)+[Var=exp(c(4))]


1st Formulation (2nd Version)

@signal GDP_q=(GDP_m1+GDP_m2+GDP_m3)/3+[Var=exp(c(1))]

@state GDP_m3=c(5)+c(6)*IP+c(7)*RS+[Var=exp(c(2))]

@state GDP_m2=GDP_m3(-1)+[Var=exp(c(3))]

@state GDP_m1=GDP_m2(-1)+[Var=exp(c(4))]


2nd Formulation (1st Version)

@signal GDP_q=(GDP_m1+GDP_m2+GDP_m3)/3+[Var=exp(c(1))]

@state GDP_q_aux=c(6)+c(7)*IP_q+c(8)*RS_q+[Var=exp(c(2))] ' GDP_q_aux - an auxiliary variable equal to GDP_q wich is necessary to estimate the model

@state GDP_m3=c(6)+c(7)*IP+c(8)*RS+[Var=exp(c(3))]

@state GDP_m2=c(6)+c(7)*IP(-1)+c(8)*RS(-1)+[Var=exp(c(4))]

@state GDP_m1=c(6)+c(7)*IP(-2)+c(8)*R(-2)+[Var=exp(c(5))]


or 2nd Formulation (2nd Version)

@signal GDP_q=(GDP_m1+GDP_m2+GDP_m3)/3+[Var=exp(c(1))]

@state GDP_q_aux=c(6)+c(7)*IP_q+c(8)*RS_q+[Var=exp(c(2))]

@state GDP_m3=c(5)+c(6)*IP+c(7)*RS+[Var=exp(c(3))]

@state GDP_m2=GDP_m3(-1)+[Var=exp(c(4))]

@state GDP_m1=GDP_m2(-1)+[Var=exp(c(5))]

I am not sure which state space formulation is correct. What do you think?

Cheers,

Andy