Enhancing Crude Oil Price Forecasting through Hybrid VMD–SVR Models: Evidence from WTI Futures across Multiple Horizons

Maaboudi, Reza; Fotros, Mohammad Hassan; Babaali, Erfan

doi:10.22050/pbr.2026.556568.1420

Enhancing Crude Oil Price Forecasting through Hybrid VMD–SVR Models: Evidence from WTI Futures across Multiple Horizons

Document Type : Original Article

Authors

Reza Maaboudi ¹

Mohammad Hassan Fotros ²

Erfan Babaali ³

¹ Associate Professor, Department of Economics, Faculty of Humanities, Ayatollah Boroujerdi University, Boroujerd, Iran

² Professor, Department of Economics, Faculty of Economic and Social Sciences, Bu-Ali Sina University, Hamedan, Iran

³ M.A. Student, Department of Economics, Faculty of Humanities, Ayatollah Ozma Borujerdi University, Lorestan, Iran

https://doi.org/10.22050/pbr.2026.556568.1420

Abstract

West Texas Intermediate (WTI) crude oil is a pivotal benchmark in the global energy market, exerting a decisive influence on economic expectations and national macroeconomic policies. As the primary pricing basis on the New York Mercantile Exchange and numerous energy futures contracts, WTI is subject to persistent and severe price volatility. Such fluctuations, often appearing as abrupt upward or downward shocks, profoundly affect key macroeconomic indicators. These include inflation, economic growth, trade balances, corporate profitability, production costs, and government budgets. Consequently, variations in WTI prices influence oil and gas markets, financial stability, energy security, and even international geopolitical relations. To address these challenges, this study develops a hybrid VMD+SVR framework to model and forecast WTI crude oil futures prices across short-, medium-, and long-run horizons. Empirical findings reveal that across all three horizons, the proposed hybrid model consistently achieves the lowest forecasting errors compared with alternative approaches. Moreover, the Diebold–Mariano and Wilcoxon tests statistically confirm the superior predictive performance of the hybrid VMD+SVR model. These results highlight the importance of integrating advanced adaptive signal decomposition (VMD) with powerful nonlinear learning algorithms (SVR) for accurate oil price forecasting. The proposed approach not only enhances forecasting accuracy but also provides practical insights for policymakers in managing economic risks, stabilizing budgets dependent on oil revenues, and formulating sustainable energy strategies. It opens a new avenue for developing financial forecasting models inspired by advanced signal processing.

Highlights

This study develops a novel hybrid VMD–SVR framework for forecasting West Texas Intermediate (WTI) crude oil prices across short-, medium-, and long-run horizons.
The Variational Mode Decomposition (VMD) effectively separates high-frequency noise from structural trends, enhancing the predictive capability of the Support Vector Regression (SVR) model.
Empirical results demonstrate that the hybrid model outperforms traditional machine learning approaches, reducing forecasting errors by 36.8% and confirming statistical superiority through the Diebold–Mariano test.
By capturing complex nonlinear dynamics and filtering volatility, the model provides practical insights for policymakers, investors, and financial analysts, aiding risk management, investment planning, and strategic energy policy.

This approach represents a robust and adaptive tool for addressing uncertainty in oil markets and contributes to the development of advanced signal-based forecasting techniques in energy economics

Keywords

Crude Oil Price

Forecasting

Variational Mode Decomposition

Machine Learning Models

SVR

Subjects

Oil and Gas Economics and Management

Baffes, J., Kose, M. A., Ohnsorge, F., & Stocker, M. (2015). The Great Plunge in Oil Prices: Causes, Consequences, and Policy Responses. World Bank Policy Research Note, No. 1. Washington, DC: World Bank. https://econpapers.repec.org/paper/eencamaaa/2015-23.htm.

Bai, J., & Perron, P. (2003). Computation and Analysis of Multiple Structural Change Models. Journal of Applied Econometrics, 18, 1–22. https://doi.org/10.1002/jae.659.

Boadi, E. (2025). Bitcoin Price Forecasting Based on a Hybrid Variational Mode Decomposition and Long Short-Term Memory Network. https://doi.org/10.48550/arXiv.2510.15900.

Boussatta, H., Chihab, M., Chihab, Y., & Chiny, M. (2023). Enhancing Oil Price Forecasting through an Intelligent Hybridized Approach. International Journal of Advanced Computer Science and Applications, 14(9), 115–125. https://doi.org/10.14569/IJACSA.2023.0140913.

Chen, T., & Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, 785–794. https://doi.org/10.1145/2939672.2939785.

Cheng, X., Wu, P., Liao, S. S., & Wang, X. (2023). An Integrated Model for Crude Oil Forecasting: Causality Assessment and Technical Efficiency. Energy Economics, 117, 106467. https://doi.org/10.1016/j.eneco.2022.106467.

Constantin, A. M., Davidescu, E. M., & Manta, A. G. (2025). Time Series Forecasting with LightGBM under Data Scarcity: An Application to Romania’s Inland Gas Consumption. Proceedings of the 19th International Conference on Business Excellence 2025, 1518–1531. https://doi.org/10.2478/picbe-2025-0118.

Coroneo, L. & Iacone, F. (2025). Testing For Equal Predictive Accuracy with Strong Dependence, International Journal of Forecasting, 41(3), 1073–1092. https://doi.org/10.1016/j.ijforecast.2024.11.003.

Drachal, K. (2023). Forecasting the Crude Oil Spot Price with Bayesian Symbolic Regression. Energies, 16(1), 4. https://doi.org/10.3390/en16010004.

Faris Hameed, A., & Jumah Al-Thahab, O., Q. (2025). Comprehensive Review on Fast Fourier Transforms Types and Their Applications. International Journal of Science and Research Archive, 16(03), 1146–1152. https://doi.org/10.30574/ijsra.2025.16.3.2663.

Foroutan, P., & Lahmiri, S. (2024). Deep learning Systems for Forecasting the Prices of Crude Oil and Precious Metals. Financial Innovation, 10(1), 1–40. https://doi.org/10.1186/s40854-024-00637-z.

Gers, F. A., Schmidhuber, J. & Cummins, F. (2000). Learning to Forget: Continual Prediction with LSTM. Neural Computer; 12 (10): 2451–2471. https://doi.org/10.1162/089976600300015015.

Huang, L., Yang, X., Lai, Y., Zou, A., & Zhang, J. (2024). Crude Oil Futures Price Forecasting Based on Variational and Empirical Mode Decompositions and Transformer Mode. Mathematics, 12(24), 4034. https://doi.org/10.3390/math12244034

Huang, Z., & Sen, B. (2025). Multivariate Distribution-free Nonparametric Testing: Generalizing Wilcoxon’s Tests Via Optimal Transport (arXiv:2503.12236). arXiv. https://doi.org/10.48550/arXiv.2503.12236.

Imani, M., Beikmohammadi, A., & Arabnia, H. R. (2025). Comprehensive Analysis of Random Forest and XGBoost Performance with SMOTE, ADASYN, and GNUS Under Varying Imbalance Levels. Technologies, 13(3), 88. https://doi.org/10.3390/technologies13030088.

Jia, H., Cao, P., Liang, T., Caiafa, C. F., Sun, Z., Kushihashi, Y., Grau, A., Bolea, Y., Duan, F., & Sole-Casals, J. (2025). Short-Time Variational Mode Decomposition. Computer Science, arXiv. https://arxiv.org/abs/2501.09174.

Kayral, I. E., Aktaş Bozkurt, M., Bejaoui, A. & Jeribi, A. (2025). Hybrid Wavelet-SVR, Machine Learning, and Deep Learning Models for Predicting Clean Energy Markets Amidst Global Events. Neural Computing & Applications, 37, 16781–16823. https://doi.org/10.1007/s00521-025-11345-9.

Kilian, L., & Zhou, X. (2020). The Econometrics of Oil Market VAR Models. Working Paper, No. 8153, Federal Reserve Bank of Dallas. https://doi.org/10.24149/wp2006.

Kiranyaz, S., Avci, O., Abdeljaber, O., Ince, T., Gabbouj, M., & Inman, D.J. (2021). 1D Convolutional Neural Networks and Applications: A Survey. Mechanical Systems and Signal Processing, 151, 107398. https://doi.org/10.1016/j.ymssp.2020.107398.

Langsetmo, L., Schousboe, J. T., Taylor, B. C., Cauley, J. A., Fink, H. A., Cawthon, P. M., Kado, D. M., Ensrud, K. E. (2023). Advantages and Disadvantages of Random Forest Models for Prediction of Hip Fracture Risk Versus Mortality Risk in the Oldest Old. JBMR Plus, 7(8), e10757. https://doi.org/10.1002/jbm4.10757.

Li, J., Hong, Z., Zhang, C., Wu, J., & Yu, C. (2024). A Novel Hybrid Model for Crude Oil Price Forecasting Based on MEEMD and Mix-KELM. Expert Systems with Applications, 246, 123104. https://doi.org/10.1016/j.eswa.2023.123104.

Li, Z., Wang, M., Wang, X., Liu, Z., & Shi, A. (2020). Oil Price Forecasting Based on Variational Mode Decomposition, Relative Entropy and LSTM Neural Network. IOP Conference Series: Materials Science and Engineering, 750(1), 012203. https://doi.org/10.1088/1757-899X/750/1/012203 .

Liu, L., Zhou, S., Jie, Q., Du, P., Xu, Y., & Wang, J. (2024). A Robust Time-varying weight combined model for crude oil price forecasting. Energy, 299, 131352. https://doi.org/10.1016/j.energy.2024.131352.

Nair, V. & Hinton, G. E. (2010). Rectified Linear Units Improve Restricted Boltzmann Machines. Proceedings of the 27th International Conference on Machine Learning, Haifa, 807–814. https://www.cs.toronto.edu/~fritz/absps/reluICML.pdf.

Niu, H., & Zhao, Y. (2021). Crude Oil Prices and Volatility Prediction by a Hybrid Model Based on Kernel Extreme Learning Machine. Mathematical Biosciences and Engineering, 18(6), 8096–8122. https://doi.org/10.3934/mbe.2021402.

Ouyang, Z.; Ravier, P.; Jabloun, M. (2021). STL Decomposition of Time Series Can Benefit Forecasting Done by Statistical Methods but Not by Machine Learning Ones. Energy Proceeding. 5, 42. https://doi.org/10.3390/ engproc2021005042.

Peng, Z. J., Zhang, C., & Tian, Y.-X. (2023). Crude Oil Price Time Series Forecasting: A Novel Approach Based on Variational Mode Decomposition, Time-Series Imaging, and Deep Learning. IEEE Access, 11, 85431–85444. https://doi.org/10.1109/ACCESS.2023.3301576.

Plakandaras, V., Gogas, P., Papadimitriou, T. & Gupta, R. (2017). The Informational Content of the Term Spread in Forecasting The US Inflation Rate: a Nonlinear Approach. Journal of Forecasting, 36(2), 109–122. https://doi.org/10.1002/for.2417.

Qin, Q., & Li, L. (2025). A VMD-based four-stage hybrid forecasting model with error correction for complex coal price series. Mathematics, 13(18), 2912. https://doi.org/10.3390/math13182912.

Ren, X., Xu, W., & Duan, K. (2022). Fourier Transform Based LSTM Prediction Model Under Oil Shocks. Quantitative Finance and Economics, 6(2), 342–358. https://doi.org/10.3934/QFE.2022015.

Scherer, D., Müller, A., Behnke, S. (2010). Evaluation of Pooling Operations in Convolutional Architectures for Object Recognition. In Proceedings of the 20th International Conference on Artificial Neural Networks, 92–101. https://doi.org/10.1007/978-3-642-15825-4_10

Seabe, P. L.; Pindza, E.; Moutsinga, C. R. B.; Aphane, M. (2025). Temporal Attention-Enhanced Stacking Networks: Revolutionizing Multi-Step Bitcoin Forecasting. Forecasting, 7(1), 1–28. https://doi.org/10.3390/forecast7010002.

Singh, P., Jha, M., & Patel, H. (2025). Wavelet-Enhanced Deep Learning Ensemble for Accurate Stock Market Forecasting: A Case Study of Nifty 50 Index. IEEE Access, 13, 89945–89960. https://doi.org/10.1109/ACCESS.2025.3568634.

Wang, L., Chen, H., & Kim, J. (2023). Hybrid machine learning models for oil price prediction: A comparative study. Energy, 268, 126–643. https://doi.org/10.1016/j.energy.2023.126643.

Yuan, F., Huang, X., Jiang, H., Jiang, Y., Zuo, Z., Wang, L., Wang, Y., Gu, S., & Peng, Y. (2025). An XLSTM–XGBoost Ensemble Model for Forecasting Non-Stationary and Highly Volatile Gasoline Price. Computers, 14(7), 256. https://doi.org/10.3390/computers14070256.

Zhang, Y., & Ma, F. (2023). Forecasting Crude Oil Prices with a New Bayesian Combination Approach. Energy Economics, 120, 106521. https://doi.org/10.1016/j.eneco.2023.106521.

Zheng, G., Li, Y., & Xia, Y. (2025). Crude Oil Price Forecasting Model Based on Neural Networks and Error Correction. Applied Sciences, 15(3), 1055. https://doi.org/10.3390/app15031055.

Zhu, S., Xu, M., Wu, J., Wang, Y., Jiang, X., Huang, Z., Wang, Y., & Zhu, Y. (2025). A Study on Crude Oil Price Forecasting Model Integrating CEEMDAN-VMD Multiscale Decomposition with CNN-BiLSTM. Results in Engineering, 27, 106391. https://doi.org/10.1016/j.rineng.2025.106391.