R^2 in GARCH model

[ldnghi]

Hi

I'm examine GARCH model in Viet nam Stock Market.
But its R^2 is small (=0.13).
Does anyone help me to explain the role of R^2 in GARCH model?
Can I use this model in my research? If not, how can I do to have the better model?

And one more question:
If I have several models that satisfy 2 conditions: all estimated parameter are significant, no autocorrelation in residuals. How can I choose the best model?

Thanks a lot