Comparing GARCH Models by Introducing Fuzzy Asymmetric Realized GARCH

Estimation of conditional variance has lots of application reflecting economic, especially financial economics, social economics and political economics’ risk and volatility research. Therefore, obtaining accurate estimation of the conditional variance is especially important. Recently Hansen has mo...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:مدلسازی اقتصادسنجی 2018-09, Vol.3 (4), p.37-58
Hauptverfasser: Esmaiel Abounoori, Mohammad Amin Zabol
Format: Artikel
Sprache:per
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Estimation of conditional variance has lots of application reflecting economic, especially financial economics, social economics and political economics’ risk and volatility research. Therefore, obtaining accurate estimation of the conditional variance is especially important. Recently Hansen has modeled the conditional variance and realized volatility simultaneously which is known as Realized GARCH model. In this paper, we introduce a fuzzy coefficient in the Realized GARCH, and then compare this model with GARCH, EGARCH and GJR-GARCH methods as well as the RGARCH model with 2 different criteria of the realized volatility concerning Tehran Stock Exchange Index. The log likelihood value used to evaluate in-sample fitting. According to this criterion, our proposed model has a better fit than the rest of the models. To evaluate the accuracy of prediction of conditional variance, the rolling window method used with two MSE and QLIKE loss functions. The results indicate that our model, the Realized GARCH with fuzzy coefficient has the best performance with both loss functions.
ISSN:2345-654X
2821-2150
DOI:10.22075/jem.2019.17215.1299