Thursday, January 31, 2013

A Request for Stata Information

Bernard Njindan Iyke writes:

"Dear Prof Dave
I am investigating the linkages among financial deepening, trade openness and economic growth.

Can you assist me in getting the Stata commands for:
1. Trivariate Panel Granger Causality Test
2. Choi (2006) Unit Root Test
3. Fisher-Johansen Panel Cointegration Test (Maddala and Wu, 1999).
Thank you in advance."

I'm not much of a Stata user myself, so I thought I'd put this out there to see what comments/suggestions readers have for Bernard. 

© 2013, David E. Giles


  1. He should ask this question on Statalist:

  2. Dear Prof Dave

    I am a student and I am learning VAR only

    Could you confirm or deny the following steps at the VAR model?
    1 / See the stationarity of variables (both stationary with ADF)
    2/Detect the optimal Lag (AIC and BIC)
    3/Estimation VAR model
    4 / confirm the hypothesis of "no autocorrelation errors"(Otherwise. What should I do?)
    5/Analyse the impulse response function and variance decomposition
    6/Analyse of causality

    At stata there is a test that I could not understand, about the stability of the model "Eigenvalue stability condition", could you assist me to understand it?

    Thank you very much Mr.

    1. Yes, these are the basic steps. As I said in this post, I'm not a STATA user. (See the other comment above for some help.)

      My guess is that what you mention is a chack on the roots of the characteristic equation - are thay all outside the unit circle. If not, the model is dynamically unstable, and shouldn't be used.