Forecasting using SAS Forecast Server, SAS/ETS, and more

Correcting heteroscedasticity without loosing intercept

Reply
Occasional Contributor
Posts: 5

Correcting heteroscedasticity without loosing intercept

[ Edited ]

Hey,

I have access to 9.4 (SAS/ETS 13.2). While correcting heteroscdasticty :

proc panel data=mydata;

model y=x1 x2 x3 x4/FIXONE FIXTWO RANONE RANTWO HAC(BANDWIDTH=NEWEYWEST94);

id firm time;run;

 

The intecept is not appearing in FIXTWO output. 

Based on http://support.sas.com/documentation/cdl/en/etsug/67525/HTML/default/viewer.htm#etsug_panel_details3... , I understand that for the two-way models, the variance and the covariances for the intercept are not corrected.

 

Is there any way I can correct for heroscdsaticity without loosing 'Intercept'

Thanks in advance!

Rohit

SAS Employee
Posts: 17

Re: Correcting heteroscedasticity without loosing intercept

Dear Rohit,

"Losing the intercept" is not an innate property of a RANTWO model with HAC covariance matrix. The intercept is mostly being dropped because of some other form of collinearity unrelated to HAC. If you can send the data and code to me at Bobby.Gutierrez@sas.com I'd be happy to take a closer look and diagnose why the intercept is being dropped.

--Bobby
Ask a Question
Discussion stats
  • 1 reply
  • 302 views
  • 0 likes
  • 2 in conversation