DCC GARCH BIVARIATE

[danchris2015]

Hello

I estimating a bivariate DCC GARCH model to analyse sectoral time-varying conditional correlation between CDS and equity returns. The parameters in the CDS Garch Equation are significant but the summation in some case is higher than 1 (although variables are stationary). I attached a case for the beverage sector which is exactly showing this. Does this mean the the model is flawed or that this data cannot be applied to this model? Thank you.