Hi everybody ! I am a beginner as a user of Gretl and I would be very pleased if someone could help me with a few details. Especially I am interested in possibilities offered by Gretl to make an asymmetric error correction model. I have not been able to find a way to divide my data, or in fact explanatory variables, on the basis of their sign (positive/negative), so that both positive and negative values of them would be in the model at the same time . Is it possible in Gretl at all ? Sincerely, Hanne _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users

Hi again, Thanks for help ! And yes, I mean that Shin et al -version of asymmetric ECM, in this case with two explanatory variables and a couple of dummies. Unfortunately, I have not managed to realize that, though I do understand the idea of it, as well an idea of that partial sum proces mentioned below. The main problem is that I do not know how to get Gretl to take into account all four (positive and negative values of the both of explanatories) restrictions at the same time. If I use a way Sample - Restrict, based on criterion, only one of limits is taken into account. If I make a model by using the whole data and after it use a way Tests-Linear restrictions and write all needed conditions there, nothing happens. I have done these kind of models earlier with another programme, but this with Gretl is something new for me. I would be so pleased, if someone could help me with this problem. Thanks for your, Hanne On Tue, Apr 3, 2012 at 6:52 PM, artur tarassow <artur.tarassow(a)googlemail.com <mailto:artur.tarassow@googlemail.com>> wrote: Is your question related to the approach suggested, among others, by Shin et al.? If you mean this one, the answer is "yes". Actually you simply need to decompose your series of interest into the positive and negative partial sum processes using e.g. smpl $i > 0 --restrict command and estimate a single-equation ECM as usual. Artur Am 3. April 2012 17:42 schrieb Hanne holopainen <nevalaine1(a)gmail.com <mailto:nevalaine1@gmail.com>>: Hi everybody ! I am a beginner as a user of Gretl and I would be very pleased if someone could help me with a few details. Especially I am interested in possibilities offered by Gretl to make an asymmetric error correction model. I have not been able to find a way to divide my data, or in fact explanatory variables, on the basis of their sign (positive/negative), so that both positive and negative values of them would be in the model at the same time . Is it possible in Gretl at all ? Sincerely, Hanne _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu <mailto:Gretl-users@lists.wfu.edu> http://lists.wfu.edu/mailman/listinfo/gretl-users _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu <mailto:Gretl-users@lists.wfu.edu> http://lists.wfu.edu/mailman/listinfo/gretl-users _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users

Hi again, Thanks for help ! And yes, I mean that Shin et al -version of asymmetric ECM, in this case with two explanatory variables and a couple of dummies.