Page 1 of 1

Panel Data Fixed Effect Dummy Variable

Posted: Wed Feb 12, 2014 8:34 am
by drmahmudali
Hy,
My data belongs to 22 banks. 17 banks are commercial and 5 banks are Islamic. The independent variable is ROA. Independent variables are 8. Out of which one variable is dummy. The value of dummy is 1 for commercial banks and 0 for islamic banks. The period is from 2010 to 2013. Independent variables also include CPI (Consumer Price Index) and GDP.
By omitting the three variables which is DT (Dummy), CPI and GDP from my equation in Hausman test i found that fixed effect model is useful for me.
my complete equation; roa c l cr cap eff bs dt cpi gdp.
When I apply this equation in fixed effect model the answer was "Near Singular Matrix". But When I omit the DT from my equation I found the result. The 1 and 0 in my dummy variable shows two different type of banks in my data. Please guide why I am unable to find results when using DT in my equation. What is the overall impact of not including this variable in my equation as it is an important hypothesis in my dissertation.

Re: Panel Data Fixed Effect Dummy Variable

Posted: Wed Feb 12, 2014 9:04 am
by startz
Fixed effects and your dummy are perfectly collinear, unless some banks switch from commercial to Islamic.