Hello,<div><br></div><div>I&#39;m writing a master thesis about volatility forecasting using GARCH and GJR models (with Normal, stud-t and GED distributions). I need to prepare out-of-sample forecasting for that models. Thus, I&#39;ve tried to prepare scripts using gig package. </div>
<div><br></div><div>I&#39;ve got the following script for GARCH (1,1) with GED distribution:</div><div><br></div><div><div>include gig.gfn</div><div>open file.gdt</div><div>model = gig_setup(RETURN,1,const)</div><div>gig_set_dist(&amp;model, 2)</div>
<div>gig_estimate(&amp;model)</div><div>#gig_plot(&amp;emodel)</div><div>#gig_dplot(&amp;emodel)</div><div><font color="#ff0000">series e = $uhat</font></div><div><font color="#ff0000">series h = $h</font></div><div>dataset addobs 1</div>
<div>a0 = $coeff[2]</div><div>a1 = $coeff[3]</div><div>b1 = $coeff[4]</div><div>series hfc = h</div><div># set future errors to their expectation</div><div>e = misszero(e)</div><div># forecast the variance</div><div>hfc = a0 + a1 * e(-1)^2 + b1 * hfc(-1)</div>
<div>smpl 165 ;</div><div>print e h hfc --byobs</div></div><div><br></div><div>However, the commands  <span style="color:rgb(255,0,0)">series e = $uhat </span>and<span style="color:rgb(255,0,0)"> </span><span style="color:rgb(255,0,0)">series h = $h</span>   don&#39;t work. The information is that: &quot;The statistic you requested is not available&quot;</div>
<div><br></div><div>I&#39;ll be really grateful if you can write me how to get &#39;uhat&#39; and &#39;h&#39; series in that case. Or maybe it is some another way for preparing out-of-sample forecasting for GARCH and GJR models. </div>
<div><br></div><div>Best regards,</div><div>Marta</div><div><br></div><div><br></div><div><br></div>