fixed effects and clustering standard errors - dated panel

[shajjudu]

Hi I canot see heterodasticity under view test (for White test). I bought the student version of Eviews package.

[shajjudu]

Dear Eviews team,

I want to estimate panel models based on 30 banks in one country for 6 years.

I currently estimate panel models with fixed-effects and a time trend in my Eviews programs as follows:
rdi=c(1)+c(2)*ci+c(3)*mi+c(4)*ni+c(5)*bi+c(6)*aci+c(7)*nrc+c(8)*lnbs+c(9)*lv+c(10)*ml+c(11)*lnage

I now want to add four additional features:
* a bank speciic time trend instead of one "global" time trend
* standard errors clustered by bank
* year fixed effects
* I also want to weight observations by the rdi of the banks, if possible

when I have written this formula (addiing yr1....yr6 dummys) I got 'near singular matrix' notification

I would greatly appreciate any help on how to do this.

Thank you very much in advance!

sham

[EViews Glenn]

The built-in heteroskedasticity tests are only for non-panel equations.

If you have 6 years of data you cannot have the period effects and the intercept. Alternately, if you add period fixed effects through the dialog, EViews will handle them as deviations from mean.

For bank clustered standard errors, Select White Period as you coefficient covariance matrix.

Bank specific trends will have to be handled by adding @trend interacted with an @expand by bank. Note that you should again be careful not to induce collinearity.

Weighting will have to be done by hand. Note that it is going to cause you problems for the built-in fixed effects unless the model is for fixed effects on the weighted data.