I did something today that I did not know existed before. I accounted for non-stationarity in my data while estimating Granger Causality tests in a VAR. Last month I added this simple test to my paper to see whether or not two of my variables were probably endogenously determined. I used the simple version of this test because I didn't know that a more robust version actually existed.
Last week, however, I attended a seminar and someone presented on Granger Causality and I couldn't understand how her whole paper could be based on this very simple test that I knew of as just a diagnostic test. So for a few days that question rested on my mind and finally today I did some research and what did I find? I found that I'm very old school. My econometrics knowledge is from the 2006-2008 era. Which, in 2015 is very old. In my econometrics days I didn't even learn how to use panel data.
This is one of the most important reasons for going to academic conferences and networking with others in your field. If you don't, you'll get left behind! Things change everyday day. Processes get upgraded and new ways of doing things are discovered every year.
Of course not every conference will yield new knowledge and sometimes because of our busy schedules we may not be able to read all the new material that gets released. Notwithstanding that though, the effort to continually learn must be made. Or else one day my 9 years as a university student will become useless simply because I didn't stay relevant.
No comments:
Post a Comment