Forecasting Volatility in ETFs using Realized Volatility Jump Models(HAR, HAR-J, HARQ, HARQ-J)
Subject Areas : Risk Management
Shiva Hallaji
1
,
Mahdi Madanchi Zaj
2
,
Fereydoun Ohadi
3
,
Hamidreza Vakilifard
4
1 - Department of Financial Management, Islamic Azad University Science and Research Branch,Tehran,Iran
2 - Department of Financial Management, Central Tehran Branch, Islamic Azad University, Tehran, Iran.
3 - Department of Industrial Engineering, Karaj Branch, Islamic Azad University, Karaj, Iran.
4 - Department of Accounting, Islamic Azad University Science and Research Branch,Tehran,Iran
Keywords: Keywords:Forecasting, Volatility, Jump,
Abstract :
Abstract This study undertook an investigation into the forecasting of volatility pertaining to jump risks within exchange-traded funds (ETFs) in Tehran. The investigation was conducted employing data sourced from six stock and fixed income funds over the financial period spanning from 2019 to 2022. The data encompassed high-frequency daily observations at intervals of 15 minutes. In the course of this research, three principal categories of heterogeneous autoregression (HAR) models were employed to prognosticate volatility. These models incorporated jumps within an econometric framework, realized through the estimation of HAR models.The findings of the study demonstrated that, among the diverse range of funds considered, a heightened predictive capacity for volatility was discernible within the realm of stock funds. Notably, the results underscored the HARQ-J model as the most efficacious among the employed HAR models. This model demonstrated proficiency in both the modeling and forecasting of realized volatility (RV), a conclusion drawn based on assessment using metrics such as the mean square error (MSE) and the quasi-likelihood (QLIKE) criteria. Furthermore, compelling evidence substantiated the superiority of models predicated on bipower variation models when it comes to predicting RV. This observation holds the promise of more accurate prognostications and enhanced volatility estimation in comparison to their HARQ counterparts.
