GARCH simulation question

[lbkova]

Dear All,

I have a bit of a problem interpreting the results of a simulation I ran.
I wanted to model (forecast) the volatility of Dax 2 years (500) days in advance using TGARCH with student distribution and TGARCH with bootstraping. To my surprise I found the that the final cumulative return distribution with student distribution is not normal while that made with bootstraping is normal! I would have expected just the opposite! Can you help me with this?

Thank you!

Ps: I attached the files I used...