ScienceDirect Publication: Finance Research Letters
Wed, 04 Dec 2019 00:01:56 GMT language
Publication date: Available online 3 December 2019
Source: Finance Research Letters
Author(s): Xinyu Wu, Haibin Xie
This paper proposes a realized EGARCH-MIDAS model with higher moments (REGARCH-MIDAS-SK) which combines the REGARCH-MIDAS model by Borup and Jakobsen (2019) and the REGARCH-SK model by Wu et al. (2019) to model volatility. A key feature of the proposed model is the ability to account for the high persistence of volatility and the time-varying non-Gaussianities of return distribution simultaneously. Empirical results show that the REGARCH-MIDAS-SK model outperforms the REGARCH model as well as the REGARCH-MIDAS and REGARCH-SK models both in terms of in-sample fit and out-of-sample forecast performance.