Changing Determinant Driver and Oil Volatility Forecasting: A Comprehensive Analysis  
Author Jiqian Wang


Co-Author(s) Feng Ma; Qin Luo


Abstract Academic research relies on the exogenous drivers to improve the accuracy of oil volatility forecasting. Following the existing literature strand, this study collects the 62 exogenous drivers that can reflect the movements of oil demand, oil supply, oil inventory, macroeconomic fundamental, financial indicators, and uncertainty. Using the various predictive regressions, our empirical results indicate that the dimension reduction regressions, especially principal component analysis regression (PCA), can successfully predict the both WTI and Brent oil volatility at the one-month ahead forecast horizon, whereas the shrinkage methods outperform peers for the medium- and long-term forecast horizon. Furthermore, the unsupervised learning method (PCA) can achieve superior forecasting performance during the period of oil price decrease, while the supervised learning methods, i.e., shrinkage method, can significantly improve the volatility accuracy during the period of increase in oil price. Additionally, the empirical results reveal that the movements of the Kilian index, World industrial production index, global economic conditions index, U.S. steel production, Chicago fed national activity index, capacity utilization for manufacturing, U.S. default yield spread and MSCI emerging market index can extremely drive the oil volatility.


Keywords Exogenous drivers, Oil volatility forecasting, Shrinkage method
    Article #:  DSBFI23-88
Proceedings of 2nd ISSAT International Conference on Data Science in Business, Finance and Industry
January 8-10, 2023 - Da Nang, Vietnam