I ran Markov regime switching model with shared garch ....i used the garch model with MS just to remove heteroscedasticity....my main variable are in mean model....so i wanna divide the high and low volatility regime...so can I divide the regime based on switching Sigma which model results have shown by ignoring garch coefficient....i ran the model in OX- metrics.....i also applied simple MS (2) model but residual have shown heteroscedasticity....so i used ms with shared garch