Something like "HAC Newey–West" in GARCH?
Posted: Fri Jan 24, 2014 6:28 am
In case of "LS - Least Squares (NLS and ARMA)", you can choose HAC Newey-West standard errors & covariance. Is there any equivalent in GARCH-model? Is it Bollerslev-Wooldridge robust standard errors & covariance? My problem is, that the residuals are autocorrelated and I cannot "remove" it (even with a lagged endogenous variable). I investigate high frequency financial data in EViews 8. Thank you very much in advance.