CMStatistics 2020: Start Registration
View Submission - CFE
Title: Augmenting the realized-GARCH: The role of signed-jumps, attenuation-biases and long-memory effects Authors:  Ioannis Papantonis - Athens University of Economics and Business (Greece) [presenting]
Elias Tzavalis - Athens University of Economics and Business (Greece)
Leonidas Rompolis - Research Center of Athens University of Economics and Business (Greece)
Orestis Agapitos - Athens University of Economics and Business (Greece)
Abstract: The focus is on the Realized-GARCH framework. We extend the Realized-GARCH to incorporate additional intra-day realized measures capturing different volatility asymmetry sources. We also consider a volatility-of-volatility effect, which can correct for attenuation-biases in measuring Realized Variance (RV), as well as heterogeneous terms of RV, which approximate long-memory properties of variance. These extensions are well-justified by the ongoing literature on Heterogeneous Auto-Regressive (HAR) models. Moreover, we examine the impact of allowing for skewed/leptokurtic distributions on the flexibility of the models to fit returns and variance jointly. Two main conclusions can be drawn from the results of our empirical analysis. First, the model-extensions that we suggest improve both the in- and out-of-sample performance of the model to predict RV, compared to the standard Realized-GARCH. This finding is justified by several goodness-of-fit and prediction-accuracy metrics that we report, as well as by a series of out-of-sample equal-prediction performance tests. Second, allowing for asymmetric/fat-tailed return distributions also seems to play a crucial role in the accurate filtering of the innovations, and consistently helps the identification of the parameters of the volatility process. This further enhances the prediction performance of all the augmented GARCH-type specifications that we consider in our analysis.