State-space: signal equations with autoregressive errors
Posted: Fri Jun 15, 2012 8:22 am
Hello everyone,
I was wondering if Eviews can handle State Space models where the signal equation has autoregressive errors.
Specifically, I have in mind a model like the one attached in the picture. [This is from Alan Clayton Matthews and James Stock's paper in the Journal of Economic and Social Measurement, 1999, An Application of the Stock/Watson methodology to the Massachusetts economy].
Here x(t) are the observed coincident variables (vector), so that equations in (1) are the signal equations. Variable s(t) is the state variable.
As you see in (2), errors are autoregressive. Equation (3) is the state variable equation.
Looking at Chapter 35 of the UG II cannot find a specific reference (whether yes or no).
Not sure if this was raised before in the Forum.
Thanks,
Manfred
I was wondering if Eviews can handle State Space models where the signal equation has autoregressive errors.
Specifically, I have in mind a model like the one attached in the picture. [This is from Alan Clayton Matthews and James Stock's paper in the Journal of Economic and Social Measurement, 1999, An Application of the Stock/Watson methodology to the Massachusetts economy].
- Model.JPG (17.87 KiB) Viewed 2720 times
As you see in (2), errors are autoregressive. Equation (3) is the state variable equation.
Looking at Chapter 35 of the UG II cannot find a specific reference (whether yes or no).
Not sure if this was raised before in the Forum.
Thanks,
Manfred