Page 1 of 1

Using E Views to estimate the Panzar Rosse Model

Posted: Sat Mar 02, 2013 10:22 am
by Aqmar
Hi,

I need some urgent help.

I have pretty much learned how to use E-Views using the book - Using E-Views for Principles of Econometrics by William E Griffiths.

However, basically I am trying to regress the equation known as the Panzar Rosee Model(picture of equation attached with this post) with this post. Essentially, I have to find the values B1, B2 and B3 which are the co-efficients of P1, P2 and P3.

I am using 14 banks over the period from 2005-2011. So, my timeline would be 2005 to 2011 and I would have a total of 9 independent variables and one dependent variable. I know how to regress this equation for perhaps one bank from 2005 to 2011 with 9 independent variables and one dependent variable. However, how will I be able to do this when I have 14 different banks, because through this equation using my sample of 14 banks I plan to assess the degree of competition in the uk banking industry.

For explanations sake, I can regress perhaps the equation GDP = CS+I+GS for UK from 2005-2011, but I cannot do so for perhaps UK, US, Japan combined. Kindly, can you advise me whether when I have more than one bank or country, shall my data then be cross-sectional data?

For further understanding of my problem, I have attached also a screen shot of the data relating to one of my variables, I have 8 other variables like this. I need to find B1, B2 and B3, from the equation applied to the sample 14 banks from 2005 to 2011.

Kind Regards,
Aqmar Chowdhury

Re: Using E Views to estimate the Panzar Rosse Model

Posted: Sat Mar 02, 2013 11:55 am
by trubador
Yes, pooled or panel estimations would be more appropriate for the purpose of your study. The following links might be of help: http://forums.eviews.com/viewtopic.php?f=7&t=74 and http://forums.eviews.com/viewtopic.php?f=7&t=3784