1. To Seminar 2 question ¡­

1. To Seminar 2 question 3c: Have posted batch files "behind" the graph that showed the adaptive capability of the LT forecasts, and the complete lack of adaptation of the GT forecasts. And a better graph.

2. If you want to estimate ARMA models outside ARFIMA, you can do that in the Non-Linear part of PcGive. More transparent perhaps? I have posted a batch file that estimates an ARMA(2,2) on our data, which should be easy to adapt if you want.