Munich Personal RePEc Archive

ARFIMAX and ARFIMAX-TARCH Realized Volatility Modeling

Degiannakis, Stavros (2008): ARFIMAX and ARFIMAX-TARCH Realized Volatility Modeling. Published in: Journal of Applied Statistics , Vol. 10, No. 35 (2008): pp. 1169-1180.


Download (585kB) | Preview


ARFIMAX models are applied in estimating the intra-day realized volatility of the CAC40 and DAX30 indices. Volatility clustering and asymmetry characterize the logarithmic realized volatility of both indices. ARFIMAX model with time-varying conditional heteroscedasticity is the best performing specification and, at least in the case of DAX30, provides statistically superior next trading day’s realized volatility forecasts.

MPRA is a RePEc service hosted by
the Munich University Library in Germany.