From unstructured to dated panel ?

[friteman]

Hello,

Sorry to bother you with another question about panel data but I have the following problem in Eviews 7.
In the dataset below, I want to create a 3D Panel where the dimensions are :
1) each loan -facililyid
2) different banks - bank (2 in the file)
3) each quarter - time (between 2003Q1 and 2003Q4).

I am able to import the data and get an unstructured unbalanced panel but I cannot move it to a dated panel. In the unstructured framework, I can run regression selecting the sample (for exemple 1 bank and 1 quarter) but even if I select the two banks I don't have access to the panel options (and so fixed effects) in the Equation Estimation box and I need thus effects :(
I am also able to create a balanced panel with the two first dimensions but I cannot find the transformation to add the time dimension in my panel data.

Thank you for your help

[EViews Gareth]

You don't have enough variety. A two-dimensional panel, using facilityid and bank perfectly describes the data. As EViews says, the time dimension is redundant.

[friteman]

Hi,

Thank you for your fast answer.
In my full sample I have 28 banks, 8 years (so 32 quarters) and more than 10 000 loans. I will try to do the same thing with more banks, loans and years to see if it works and I will keep you updated if I have a problem.

Thank's again

[friteman]

Hi,

So I was able to get a dated panel structure (with facilityid and bank (5) for the 2 first dimensions and time (12) for the 3rd one).
However I cannot get the panel options when I try to run a regression (I still need fixed effects). It's normal if I use the 3 dimensions but even if I focus on 1 quarter (using smpl 2005q1 2005q1 - so I have a 2 dimensions panel) the option does not appear in the regression box.

Am I missing something ?

Thx a lot

[EViews Gareth]

Panel estimation options are only available in workfiles structured as 2D panels.