Mihai Yes, you may try the "GJR- Garch in Mean" either via GUI or gretl console. The estimated results are shown below. Yi-Nung Yang ? list x = const ? garchm(r, 1, 1,x, 1, 0, 0, 0, 0.7, 0, 0, null, null) ================================================================ The coefficient of GJR Term is gamma_i Presample variance: backcast (parameter = 0.7) ------- mean eq & variance eq in this GARCH model -------- e = y - b0 h = omega +alpha1*(e(-1))^2+gamma1*(e(-1))^2*(e(-1)<0)+beta1*h(-1) ================================================================ Using numerical derivatives Tolerance = 1.81899e-012 Function evaluations: 69 Evaluations of gradient: 21 Model 3: ML, using observations 1-500 ll = -0.5*(log(2*pi)+log(h) + (e^2)/h) Standard errors based on Outer Products matrix estimate std. error z p-value ----------------------------------------------------- b0 0.994557 0.0429156 23.17 8.19e-119 *** omega 0.355447 0.0900577 3.947 7.92e-05 *** alpha1 0.262756 0.0804504 3.266 0.0011 *** gamma1 0.491780 0.204598 2.404 0.0162 ** beta1 0.287692 0.115538 2.490 0.0128 ** Log-likelihood -730.5458 Akaike criterion 1471.092 Schwarz criterion 1492.165 Hannan-Quinn 1479.361 2011/3/9 cociuba <cociuba(a)gmail.com&gt;: > Thank you for your quick response. > -I think that it would be helpful for newbie users (like me) to find GIG > manual much easier (like for example a link in the help window from GIG > gui). > - I understand, now, how to calculate TARCH model from gretl (thanks for > the info). > - I am using GRETL 1.9.4, build date 27.02, compiled from sources so > it's almost up to date :-) > > Thank you for your help, Mihai > _______________________________________________ > Gretl-users mailing list > Gretl-users(a)lists.wfu.edu > http://lists.wfu.edu/mailman/listinfo/gretl-users >