E-Thesis 26 views
ESSAYS ON FORECASTING STOCK MARKET VOLATILITY / ABDULKARIM ALHEJAILI
Swansea University Author: ABDULKARIM ALHEJAILI
DOI (Published version): 10.23889/SUThesis.71310
Abstract
This thesis aims to enhance the predictive accuracy of volatility models by examining the predictive power of key uncertainty indicators and introducing a global uncertainty measure designed to forecast future volatility across international stock markets. It comprises three self-contained empirical...
| Published: |
Swansea University
2025
|
|---|---|
| Institution: | Swansea University |
| Degree level: | Doctoral |
| Degree name: | Ph.D |
| Supervisor: | Shabi-Ul-Hassan, S., Tsakou, K., and Shabi, S. |
| URI: | https://cronfa.swan.ac.uk/Record/cronfa71310 |
| Abstract: |
This thesis aims to enhance the predictive accuracy of volatility models by examining the predictive power of key uncertainty indicators and introducing a global uncertainty measure designed to forecast future volatility across international stock markets. It comprises three self-contained empirical chapters, each addressing a distinct dimension of volatility forecasting.Chapter Two examines the predictive power of local implied volatility (IV) and economic policy uncertainty (EPU) indicators on forecasting aggregate volatility across international market. A central contribution lies in presenting the first comprehensive cross-country of these measures. The chapter also examines whether U.S. uncertainty measures provide superior forecasting performance compared to local uncertainty indicators. The findings reveal that local IV is consistently the most effective predictor of future volatility, particularly during periods of heightened uncertainty, while the predictive strength of U.S. indicators weakens considerably once local measures are accounted.Chapter Three introduces a new Global Implied Volatility (GIV) index, constructed using principal component analysis to combine the IV information from 13 international equity markets and two commodity markets gold and oil. This chapter evaluates the performance of the GIV within the Heterogeneous Autoregressive (HAR) framework and shows that it significantly outperforms the VIX in forecasting implied volatility. These results challenge the prevailing reliance on the VIX as a global uncertainty benchmark and underscore the benefits of incorporating cross-national information.Chapter Four extends the analysis by assessing the ability of the GIV to forecast RV across 28 international stock markets. The results demonstrate that the GIV improves out-of-sample forecast accuracy in most markets. Its predictive performance is further strengthened when combined with the local IV index. The study also introduces a new Global Variance Risk Premium (GVRP), which shows superior predictive power for international equity returns, particularly at medium and longer horizons compared to the U.S VRP. |
|---|---|
| Keywords: |
Volatility forecasting, Implied Volatility (IV), Global Implied Volatility (GIV), Realized volatility (RV), Economic policy uncertainty (EPU), Variance Risk Premium (VRP), Heterogeneous Autoregressive (HAR) model. |
| College: |
Faculty of Humanities and Social Sciences |
| Funders: |
Jouf University |