Undated or dated panel regression?

For requesting general information about EViews, sharing your own tips and tricks, and information on EViews training or guides.

Moderators: EViews Gareth, EViews Moderator

amybakker
Posts: 2
Joined: Tue Apr 19, 2016 9:00 am

Undated or dated panel regression?

Postby amybakker » Tue Apr 19, 2016 9:17 am

Hi everyone,

I'm really new to EViews (9.5) and I hope you can help me to figure out which test would be best to use.
I'm not sure if it is best to perform an OLS regression or to use an undated or dated panel regression.
My sample consists out of 1000 acquisitions between 2005 and 2010. Around half of the sample only did one acquisition in this time frame, where the other half did more than one acquisitions. My dependent variable is the acquiring companies dividends in t=0, where I have around 5 independent/control variables.

In the first column I numbered all the acquisitions from 1 to 1000, and in the second column I gave the firm's firmid's (around 600 different firm's), the third column includes the year of the acquisition, and the following columns the dependent and independent variables.

I would think that a dated panel would be best for my study. In step 3 I fill in the firmid as cross section ID series, and the year as date series, however then I get the following error: Repeated values found in source id series.

When I select the acquisition number as cross section ID series, I get another error: Illegal panel structure DATEID is nested within ACQUISITION.

Could somebody please explain me what these errors mean, and maybe give me tips on what would be the best regression to do in my case.

Thank you so much for taking the time to read this!

Kind regards,

Amy

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: Undated or dated panel regression?

Postby EViews Gareth » Tue Apr 19, 2016 10:17 am

A panel is defined as a structure where each observation is uniquely identified by a single time period and cross-section id (or if an undated panel, two cross-section identifiers)

The first error indicates that they are not uniquely identified - more than one observation have the same date/id combination.

The second error indicates that your two identifiers (time and cross-section) are nested - in essence you have provided too many identifiers.
Follow us on Twitter @IHSEViews

amybakker
Posts: 2
Joined: Tue Apr 19, 2016 9:00 am

Re: Undated or dated panel regression?

Postby amybakker » Wed Apr 20, 2016 1:27 am

Thank you for your quick response.

Would you advise me to perform an OLS regression instead of a panel regression, or is there something that I have to change in my Excel file maybe so that I won't get the second error?

Thanks again.

Amy

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: Undated or dated panel regression?

Postby EViews Gareth » Wed Apr 20, 2016 4:01 am

Whether you should perform a panel regression or not is purely an economic question, and is up to you
Follow us on Twitter @IHSEViews


Return to “General Information and Tips and Tricks”

Who is online

Users browsing this forum: No registered users and 15 guests