Panel data structure, possible missing Ids?
Posted: Sun Jun 30, 2013 9:27 am
Dear all
Although I am comfortable with time series, I have never worked with panel data. I have followed EviewsGareth's posts on stacking data in excel, so I believe I have done it correctly. The only difference in my dataset form most other panel regression example i've found online is that I have around 2000 time periods per country (daily bond data). essentially i want to analyse the impact of rating announcements on bond spreads, so I plan to perform the following fixed country effects regression
spread= c+ spread(t-1) + interest rate + rating dummy.
I will not have access to eviews for a couple of days but i would very much like to know if I have prepared the data correctly for my analysis. Here is a sample of the stacked data (just austria, belgium and cyprus). Consider I have 17 countries in the entire dataset.
Thanks you in advance!
Although I am comfortable with time series, I have never worked with panel data. I have followed EviewsGareth's posts on stacking data in excel, so I believe I have done it correctly. The only difference in my dataset form most other panel regression example i've found online is that I have around 2000 time periods per country (daily bond data). essentially i want to analyse the impact of rating announcements on bond spreads, so I plan to perform the following fixed country effects regression
spread= c+ spread(t-1) + interest rate + rating dummy.
I will not have access to eviews for a couple of days but i would very much like to know if I have prepared the data correctly for my analysis. Here is a sample of the stacked data (just austria, belgium and cyprus). Consider I have 17 countries in the entire dataset.
Thanks you in advance!