Student t distribution in ARMA-GARCH model

Hi everyone, My director and I are disagreeing on something. I estimate ARMA-GARCH models using the Student t distribution. When I estimate such models, the output (see joint file) shows a ''DoF'' parameter, with an estimated value, a standard error and a t-stat.

Here is the reason of our arguing : I think that this is the value of the ''Degrees of freedom'' of the Student t distribution estimated to ''fit'' the residuals of my model. It is called ''DoF'' after all, and it makes sense with my data. My director says it does NOT represent the degrees of freedom of the distribution. He says it might be a simple test to decide whether or not the Student t distribution is a good fit to the model, that has nothing to do with the number of degrees of freedom of the distribution…

Who is right, who is wrong? What have we missed? Best regards,and thanks in advance, Simon

[Simon Mathieu]

I think that settles it (see joint file). Thanks for the support, SM