Hello R-Community,
I actually don´t know how to implement the following:
I have quartaly Panel- Data and want to run a Granger- Test. I´ve already
done a univariate autoregression with grangertest() from package lmtest.
Because the Granger test is designed to handle pairs of variables, and may
produce misleading results when the true relationship involves three or more
variables,
a similar test involving more variables can be applied with vector
autoregression.
And a bivariate (panel-)VAR(1) model seems appropriate to me.
( I have read Chapter2 of B.Pfaffs book Analysis of integrated and
cointegrated Time Series in R and the application in package vars, because
didn´t find it in Panel-Context)
But I guess it should be possible to run my estimation with the tools of the
vars-package in consideration of (group)correlated Residuals?
Is the usage of vars-package a good idea? Or do I need o run a dynamic panel
model with pgmm from plm-package?
Because I can´t give you an reproduceable example, I try to give you a
formal idea of my model:
The formal Model of the bivariate VAR(1) as extension from the Basismodel
looks like this:
[R(i,t) ;
C(I,t)]
=
[a(1,i);
a(2,i)]
+[b1* R(i,t-1);
b2*R(i,t-1)]
+[d1* C(i,t-1);
d2*C(i,t-1)]
+[ß1* X(i,t);
ß2*X(i,t)]
+[e1(i,t);
e2(i,t)]
Where
t= time dimension,
i= Country
[a(1,i); a(2,i)] = Vector of individual constants
[b1* R(i,t-1); b2R(i,t-1)] +[d1* C(i,t-1); d2C(i,t-1)] = Vectors
of lagged endogeneous
[ß1* X(i,t); ß2X(i,t)]= Vector with other exogeneous Variables (not lagged)
[e1(i,t); e2(i,t)]= Vector of residuals
And the question is if R(i,t) granger cause C(I,t) and the other way around.
The Realization in R of the simultaneous regression and afterwards the Wald
test is a big question to me.
Thanks for your time and your hints!
Katie
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
