Do Gas Price and Uncertainty Indices Forecast Crude Oil Prices? Fresh Evidence Through XGBoost Modeling

Kais Tissaoui (), Taha Zaghdoudi, Abdelaziz Hakimi and Mariem Nsaibi
Additional contact information
Kais Tissaoui: University of Ha’il, Applied College
Abdelaziz Hakimi: University of Jendouba and V.P.N.C Lab FSJEG
Mariem Nsaibi: University of Ha’il, Applied College

Computational Economics, 2023, vol. 62, issue 2, No 8, 663-687

Abstract: Abstract This study examines the forecasting power of the gas price and uncertainty indices for crude oil prices. The complex characteristics of crude oil price such as a non-linear structure, time-varying, and non-stationarity motivate us to use a newly proposed approach of machine learning tools called XGBoost Modelling. This intelligent tool is applied against the SVM and ARIMAX (p,d,q) models to assess the complex relationships between crude oil prices and their forecasters. Empirical evidence shows that machine learning models, such as the SVM and XGBoost models, dominate traditional models, such as ARIMAX, to provide accurate forecasts of crude oil prices. Performance assessment reveals that the XGBoost model displays superior prediction capacity over the SVM model in terms of accuracy and convergence. The superior performance of XGBoost is due to its lower complexity and costs, high accuracy, and rapid processing times. The feature importance analysis conducted by the Shapley additive explanation method (SHAP) highlights that the different uncertainty indexes and the gas price display a significant ability to forecast future WTI crude prices. Additionally, the SHAP values suggest that the oil implied volatility captures valuable forecasting information of gas prices and other uncertainty indices that affect the WTI crude oil price.

Keywords: Crude oil price; Gas price; Uncertainty indexes; Complex relationship; eXtreme Gradient Boosting; Shapley additive explanation method (search for similar items in EconPapers)
JEL-codes: C53 D81 Q43 Q47 (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
http://link.springer.com/10.1007/s10614-022-10305-y Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:kap:compec:v:62:y:2023:i:2:d:10.1007_s10614-022-10305-y

Ordering information: This journal article can be ordered from
http://www.springer. ... ry/journal/10614/PS2

DOI: 10.1007/s10614-022-10305-y

Access Statistics for this article

Computational Economics is currently edited by Hans Amman

More articles in Computational Economics from Springer, Society for Computational Economics Contact information at EDIRC.
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().