How do i store the degr. of freedom in rolling GARCH?

[oska]

Hello,

I have written some code to be able to forecast conditional variance with a GARCH model (with a rolling estimation window). I am using the t-distribution, and for each regression there's a number for the degrees of freedom. My question is, how do i adjust my code so that the degrees of freedom are stored for each corresponding variance forecast?

Code: Select all

series h_1_step=na series r_1_step=na for !i=1 to 562 smpl !i 949+!i equation eq_01.arch(1, 1, tdist) lth c smpl @all eq_01.forecast r_pred se_pred h_pred smpl 950+!i 950+!i series h_1_step= h_pred 'One-step-ahead of conditional variance series r_1_step= r_pred 'One-step-ahead of actual series next smpl @all

Where lth is the original series.

Thank you in advance,

Ola

[EViews Gareth]

eq.@df ?

[oska]

eq.@df ?

Hello Gareth,

I have never done any programming in eviews before, would you mind telling me where to insert the eq@df so that the program would work?

[EViews Gareth]

Wherever it is that you want to use the degree of freedom.

[oska]

Well, I want to store the "T-DIST.DOF" number that appears with every regression estimate of my forecasts. I.e. I want to get a series of 561 degrees of freedoms from each estimate that corresponds with the variances that I have forecasted. I am having trouble understanding how to use the eq.@df to make that happen.

[EViews Gareth]

series dof = eq_01.@df