state-space model

[Emily89]

hi, I am working on a dynamic factor model. I try to use the state-space model to figure out a common factor for GDP of different countries.
The model I use is:

Y(t) = A*Y(t-1) +B*F(t) +u(t)
F(t) = C*F(t-1) + e(t)

The F(t) is the dynamic factor model(unobservable), I enter the following code in the eviews 7:

@signal uk = c(11)*sv1 + c(21)*uk(-1) + [var=exp(c(1))]
@signal germany = c(12)*sv1 +c(22)*germany(-1) + [var=exp(c(2))]
@signal france = c(13)*sv1 +c(23)*france(-1) + [var=exp(c(3))]

@state sv1 = c(31) * sv1(-1) + [var=exp(c(32))]

But the standard error, z statistics and probability are all "NA". What's wrong with the code? Could anyone help?

Thank you!

[startz]

hi, I am working on a dynamic factor model. I try to use the state-space model to figure out a common factor for GDP of different countries.
The model I use is:

Y(t) = A*Y(t-1) +B*F(t) +u(t)
F(t) = C*F(t-1) + e(t)

The F(t) is the dynamic factor model(unobservable), I enter the following code in the eviews 7:

@signal uk = c(11)*sv1 + c(21)*uk(-1) + [var=exp(c(1))]
@signal germany = c(12)*sv1 +c(22)*germany(-1) + [var=exp(c(2))]
@signal france = c(13)*sv1 +c(23)*france(-1) + [var=exp(c(3))]

@state sv1 = c(31) * sv1(-1) + [var=exp(c(32))]

But the standard error, z statistics and probability are all "NA". What's wrong with the code? Could anyone help?

Thank you!

Nothing appears to be wrong with the code. You might try
(1) Changing the model some. Perhaps constants in each equation?
(2) Try different starting values for the parameters?