In today's energy markets, forecasting energy commodity prices is difficult due to pronounced nonlinearity, nonstationarity, volatility clustering, structural changes, and multi-scale temporal dependencies. These features are of significant importance for financial risk management, portfolio allocation, derivative pricing, and decision-making in energy markets. To address these problems, we proposed a novel hybrid forecasting framework that combines a decomposition method based on kernel regression and nonparametric kernel smoothing with linear and nonlinear forecasting models. The proposed approach decomposes each energy commodity price into a smooth trend component and a stochastic fluctuation component. This enables the long-run market trend to be modeled separately from short-run fluctuations and localized market variability. The proposed framework was evaluated using five major global energy commodity markets, including WTI crude oil, Brent crude oil, natural gas, gasoline, and heating oil. Empirical results demonstrate that the proposed decomposition-based method consistently outperforms the best forecasting models proposed in the literature, direct hybrid models, and traditional standalone forecasting models across multiple statistical measures and tests of accuracy and efficiency. Our results indicated that kernel-based decomposition effectively captures complex nonlinear dependence structures while separating persistent trends from short-run fluctuations, leading to significant improvements in forecasting accuracy, forecast stability, and directional forecasting performance. The results indicated that the proposed framework can serve as a reliable tool for energy market forecasting and financial analytics to model and predict highly volatile commodity markets and other complex financial time series.
Citation: Hasnain Iftikhar, Said Farooq Shah, Fatimah E. Almuhayfith, Paulo Canas Rodrigues. A kernel regression-based hybrid forecasting framework for highly volatile energy commodity markets[J]. AIMS Mathematics, 2026, 11(8): 24877-24912. doi: 10.3934/math.20261000
In today's energy markets, forecasting energy commodity prices is difficult due to pronounced nonlinearity, nonstationarity, volatility clustering, structural changes, and multi-scale temporal dependencies. These features are of significant importance for financial risk management, portfolio allocation, derivative pricing, and decision-making in energy markets. To address these problems, we proposed a novel hybrid forecasting framework that combines a decomposition method based on kernel regression and nonparametric kernel smoothing with linear and nonlinear forecasting models. The proposed approach decomposes each energy commodity price into a smooth trend component and a stochastic fluctuation component. This enables the long-run market trend to be modeled separately from short-run fluctuations and localized market variability. The proposed framework was evaluated using five major global energy commodity markets, including WTI crude oil, Brent crude oil, natural gas, gasoline, and heating oil. Empirical results demonstrate that the proposed decomposition-based method consistently outperforms the best forecasting models proposed in the literature, direct hybrid models, and traditional standalone forecasting models across multiple statistical measures and tests of accuracy and efficiency. Our results indicated that kernel-based decomposition effectively captures complex nonlinear dependence structures while separating persistent trends from short-run fluctuations, leading to significant improvements in forecasting accuracy, forecast stability, and directional forecasting performance. The results indicated that the proposed framework can serve as a reliable tool for energy market forecasting and financial analytics to model and predict highly volatile commodity markets and other complex financial time series.
| [1] |
Y. Wu, W. Ren, J. Wan, X. Liu, Time-frequency volatility connectedness between fossil energy and agricultural commodities: Comparing the COVID-19 pandemic with the Russia-Ukraine conflict, Financ. Res. Lett., 55 (2023), 103866. https://doi.org/10.1016/j.frl.2023.103866 doi: 10.1016/j.frl.2023.103866
|
| [2] |
C. Yan, S. Zhun, A study on time-varying dependence between energy markets and linked assets based on the Russia-Ukraine conflict, Energ. Explor. Exploit., 42 (2024), 1408–1421. https://doi.org/10.1177/01445987241228322 doi: 10.1177/01445987241228322
|
| [3] |
J. Aizenman, R. Lindahl, D. Stenvall, G. S. Uddin, Geopolitical shocks and commodity market dynamics: New evidence from the Russia-Ukraine conflict, Eur. J. Polit. Econ., 85 (2024), 102574. https://doi.org/10.1016/j.ejpoleco.2024.102574 doi: 10.1016/j.ejpoleco.2024.102574
|
| [4] | H. Iftikhar, J. Zywiołek, J. L. López-Gonzales, O. Albalawi, Electricity consumption forecasting using a novel homogeneous and heterogeneous ensemble learning, Front. Energy Res., 12 (2024), 1442502. |
| [5] |
F. Yu, Y. Wang, Empirical study of portfolio rebalancing strategies in the chinese ETF market, Financ. Res. Lett., 102 (2026), 110119. https://doi.org/10.1016/j.frl.2026.110119 doi: 10.1016/j.frl.2026.110119
|
| [6] |
D. Charles, The impact of oil prices on Guyana's real exchange rate: An AlphaFold-decomposition analysis, FinTech Sustain. Innov., 2 (2026), A3. https://doi.org/10.47852/bonviewFSI52025903 doi: 10.47852/bonviewFSI52025903
|
| [7] |
M. E. Hoque, M. Sahabuddin, F. Bilgili, Volatility interconnectedness among financial and geopolitical markets: Evidence from COVID-19 and Ukraine-Russia crises, Econ. Anal. Policy, 82 (2024), 303–320. https://doi.org/10.1016/j.eap.2024.03.015 doi: 10.1016/j.eap.2024.03.015
|
| [8] | J. Nasir, H. Iftikhar, M. Aamir, H. Iftikhar, P. C. Rodrigues, M. Z. Rehman, A hybrid LMD–ARIMA–machine learning framework for enhanced forecasting of financial time series: Evidence from the NASDAQ composite index, Mathematics, 13 (2025), 2389. |
| [9] |
L. S. Dar, M. M. Abdelwahab, M. Aamir, M. Rind, P. C. Rodrigues, M. A. Abdelkawy, A decomposition-driven hybrid approach to forecasting oil market dynamics, Symmetry, 18 (2026), 465. https://doi.org/10.3390/sym18030465 doi: 10.3390/sym18030465
|
| [10] |
Y. Zhang, S. Feng, P. Wang, Z. Tan, X. Luo, Y. Ji, et al., Learning self-growth maps for fast and accurate imbalanced streaming data clustering, IEEE T. Neur. Net. Lear., 36 (2025), 16049–16061. https://doi.org/10.1109/TNNLS.2025.3563769 doi: 10.1109/TNNLS.2025.3563769
|
| [11] | H. Iftikhar, M. Qureshi, J. Zywiołek, J. L. López-Gonzales, O. Albalawi, Short-term PM 2.5 forecasting using a unique ensemble technique for proactive environmental management initiatives, Front. Env. Sci., 12 (2024), 1442644. |
| [12] | B. Fattouh, An anatomy of the crude oil pricing system, Oxford institute for energy studies, 2011. |
| [13] |
I. Riepin, F. Musgens, Seasonal flexibility in the european natural gas market, Energ. J., 43 (2022), 117–138. https://doi.org/10.5547/01956574.43.1.irie doi: 10.5547/01956574.43.1.irie
|
| [14] |
F. Quispe, E. Salcedo, H. Iftikhar, A. Zafar, M. Khan, J. E. Turpo-Chaparro, et al., Multi-step ahead ozone level forecasting using a component-based technique: A case study in Lima, Peru, AIMS Environ. Sci., 11 (2024), 401–425. http://dx.doi.org/10.3934/environsci.2024020 doi: 10.3934/environsci.2024020
|
| [15] |
X. Sun, B. Guo, Y. Yang, Y. Pan, Time for a change! Uprooting users embedded in the status quo from habitual decision-making, Decis. Support Syst., 189 (2025), 114371. https://doi.org/10.1016/j.dss.2024.114371 doi: 10.1016/j.dss.2024.114371
|
| [16] | W. F. Outlook, Short-term energy outlook, US Energy Information Administration (EIA): Washington, DC, USA, 2010. |
| [17] |
S. M. Gonzales, H. Iftikhar, J. L. López-Gonzales, Analysis and forecasting of electricity prices using an improved time series ensemble approach: An application to the peruvian electricity market, AIMS Math., 9 (2024), 21952–21971. http://dx.doi.org/10.3934/math.20241067 doi: 10.3934/math.20241067
|
| [18] |
H. Iftikhar, J. E. Turpo-Chaparro, P. C. Rodrigues, J. L. López-Gonzales, Forecasting day-ahead electricity prices for the Italian electricity market using a new decomposition—combination technique, Energies, 16 (2023), 6669. https://doi.org/10.3390/en16186669 doi: 10.3390/en16186669
|
| [19] |
M. Yu, H. Huang, R. Hou, X. X. Ma, S. Yuan, A deep graph kernel-based time series classification algorithm, Pattern Anal. Appl., 27 (2024), 73. https://doi.org/10.1007/s10044-024-01292-x doi: 10.1007/s10044-024-01292-x
|
| [20] |
J. Huo, Finite-difference solution ansatz approach in least-squares monte Carlo, J. Comput. Financ., 29 (2025), 67–121. https://doi.org/10.21314/JCF.2025.008 doi: 10.21314/JCF.2025.008
|
| [21] |
H. Iftikhar, N. Bibi, P. C. Rodrigues, J. L. López-Gonzales, Multiple novel decomposition techniques for time series forecasting: Application to monthly forecasting of electricity consumption in pakistan, Energies, 16 (2023), 2579. https://doi.org/10.3390/en16062579 doi: 10.3390/en16062579
|
| [22] | S. Jiang, X. Yuan, Does the e-CNY pilot boost consumption upgrading? evidence from China's city-level quasi-experiment, Appl. Econ. Lett., 2026, 1–5. https://doi.org/10.1080/13504851.2026.2634858 |
| [23] |
J. Berrisch, F. Ziel, Distributional modeling and forecasting of natural gas prices, J. Forecasting, 41 (2022), 1065–1086. https://doi.org/10.1002/for.2853 doi: 10.1002/for.2853
|
| [24] |
A. Wan, J. He, X. Zhou, K. Al-Bukhaiti, P. Zhu, TSBMNet-XAI interpretable multi-step offshore wind power forecasting for enhanced energy storage integration, Comput. Elect. Eng., 136 (2026), 111230. https://doi.org/10.1016/j.compeleceng.2026.111230 doi: 10.1016/j.compeleceng.2026.111230
|
| [25] | R. F. Engle, Autoregressive conditional heteroscedasticity with estimates of the variance of united kingdom inflation, Econometrica: J. Econ. Soc., 1982,987–1007. |
| [26] |
T. Bollerslev, Generalized autoregressive conditional heteroskedasticity, J. Econometrics, 31 (1986), 307–327. https://doi.org/10.1016/0304-4076(86)90063-1 doi: 10.1016/0304-4076(86)90063-1
|
| [27] |
M. S. Alam, M. Murshed, P. Manigandan, D. Pachiyappan, S. Z. Abduvaxitovna, Forecasting oil, coal, and natural gas prices in the pre-and post-COVID scenarios: Contextual evidence from india using time series forecasting tools, Resour. Policy, 81 (2023), 103342. https://doi.org/10.1016/j.resourpol.2023.103342 doi: 10.1016/j.resourpol.2023.103342
|
| [28] |
Y. Wang, Y. Wang, M. Zhao, Spatiotemporal evolution and driving factors of green transition resilience in four types of China's resource-based cities based on the geographical detector model, Sustainability, 18 (2026), 391. https://doi.org/10.3390/su18010391 doi: 10.3390/su18010391
|
| [29] | J. D. Hamilton, What is an oil shock?, J. Econometrics, 113 (2003), 363–398. https://doi.org/10.1016/S0304-4076(02)00207-5 |
| [30] |
P. Sadorsky, Modeling and forecasting petroleum futures volatility, Energ. Econ., 28 (2006), 467–488. https://doi.org/10.1016/j.eneco.2006.04.005 doi: 10.1016/j.eneco.2006.04.005
|
| [31] |
F. Khan, H. Iftikhar, I. Khan, P. C. Rodrigues, A. A. Alharbi, J. Allohibi, A hybrid vector autoregressive model for accurate macroeconomic forecasting: An application to the US economy, Mathematics, 13 (2025), 1706. https://doi.org/10.3390/math13111706 doi: 10.3390/math13111706
|
| [32] |
B. A. Memon, Forecasting fossil energy price dynamics with deep learning: Implications for global energy security and financial stability, Algorithms, 18 (2025), 776. https://doi.org/10.3390/a18120776 doi: 10.3390/a18120776
|
| [33] |
M. Vasant, S. Ganesan, G. Kumar, Enhancing e-commerce security: A hybrid machine learning approach to fraud detection, FinTech Sustain. Innov., 1 (2025), A7. https://doi.org/10.47852/bonviewFSI52024882 doi: 10.47852/bonviewFSI52024882
|
| [34] | A. Ebiwonjumi, A comparative study of penalized regression methods in estimating and predicting economic growth in nigeria, FinTech Sustain. Innov., 2026, 1–17. https://doi.org/10.47852/bonviewFSI62026470 |
| [35] |
G. P. Zhang, Time series forecasting using a hybrid arima and neural network model, Neurocomputing, 50 (2003), 159–175. https://doi.org/10.1016/S0925-2312(01)00702-0 doi: 10.1016/S0925-2312(01)00702-0
|
| [36] |
M. Kljajic, V. Mizdrakovic, L. Jovanovic, N. Bacanin, V. Simic, D. Pamucar, et al., Gasoline and crude oil price prediction using multi-headed variational neighbour search-tuned recurrent neural networks, Comput. Econ., 67 (2026), 3087–3122. https://doi.org/10.1007/s10614-025-10967-4 doi: 10.1007/s10614-025-10967-4
|
| [37] |
M. A. Alwadi, Fuel sales price forecasting using time series, machine learning, and deep learning models, Eng. Technol. Appl. Sci. Res., 15 (2025), 22360–22366. https://doi.org/10.48084/etasr.10348 doi: 10.48084/etasr.10348
|
| [38] |
S. Karasu, A novel hybrid model using demand concentration curves, chaotic AFDB-SFS algorithm and Bi-LSTM networks for heating oil price prediction, Electronics, 14 (2025), 4814. https://doi.org/10.3390/electronics14244814 doi: 10.3390/electronics14244814
|
| [39] |
D. Lyu, Z. Wang, A. Kumar, J. Jin, Morality is for social being: The role of morality in social-adjustive functional attitudes toward counterfeit luxury consumption, J. Bus. Ethics, 204 (2026), 959–980. https://doi.org/10.1007/s10551-025-06027-4 doi: 10.1007/s10551-025-06027-4
|
| [40] |
M. Qureshi, A. F. Hashem, H. Iftikhar, P. C. Rodrigues, A hybrid STL-based ensemble model for PM 2.5 forecasting in Pakistani cities, Symmetry, 17 (2025), 1827. https://doi.org/10.3390/sym17111827 doi: 10.3390/sym17111827
|
| [41] |
S. Z. Abbasi, M. M. Abdelwahab, I. Hussain, M. Qureshi, M. Rind, P. C. Rodrigues, et al., A multi-stage decomposition and hybrid statistical framework for time series forecasting, Axioms, 15 (2026), 273. https://doi.org/10.3390/axioms15040273 doi: 10.3390/axioms15040273
|
| [42] |
E. Sofianos, E. Zaganidis, T. Papadimitriou, P. Gogas, Forecasting east and west coast gasoline prices with tree-based machine learning algorithms, Energies, 17 (2024), 1296. https://doi.org/10.3390/en17061296 doi: 10.3390/en17061296
|
| [43] |
A. J. Conejo, M. A. Plazas, R. Espinola, A. B. Molina, Day-ahead electricity price forecasting using the wavelet transform and arima models, IEEE T. Power Syst., 20 (2005), 1035–1042. https://doi.org/10.1109/TPWRS.2005.846054 doi: 10.1109/TPWRS.2005.846054
|
| [44] |
N. E. Huang, Z. Shen, S. R. Long, M. C. Wu, H. H. Shih, Q. Zheng, et al., The empirical mode decomposition and the hilbert spectrum for nonlinear and non-stationary time series analysis, P. Roy. Soc. London Ser. A, 454 (1998), 903–995. https://doi.org/10.1098/rspa.1998.0193 doi: 10.1098/rspa.1998.0193
|
| [45] |
L. Yu, S. Wang, K. K. Lai, Forecasting crude oil price with an emd-based neural network ensemble learning paradigm, Energ. Econ., 30 (2008), 2623–2635. https://doi.org/10.1016/j.eneco.2008.05.003 doi: 10.1016/j.eneco.2008.05.003
|
| [46] |
M. Noor, H. M. Nazir, M. Qureshi, H. Iftikhar, P. C. Rodrigues, A. F. Hashem, A novel hybrid decomposition-based wind speed prediction model: Application to Lahore, Pakistan, IEEE Access, 14 (2026), 66162–66178. https://doi.org/10.1109/ACCESS.2026.3688253 doi: 10.1109/ACCESS.2026.3688253
|
| [47] |
Z. J. Peng, C. Zhang, Y. X. Tian, Crude oil price time series forecasting: A novel approach based on variational mode decomposition, time-series imaging, and deep learning, IEEE Access, 11 (2023), 82216–82231. https://doi.org/10.1109/ACCESS.2023.3301576 doi: 10.1109/ACCESS.2023.3301576
|
| [48] | M. E. Torres, M. A. Colominas, G. Schlotthauer, P. Flandrin, A complete ensemble empirical mode decomposition with adaptive noise, In: 2011 IEEE international conference on acoustics, speech and signal processing (ICASSP), IEEE, 2011, 4144–4147. https://doi.org/10.1109/ICASSP.2011.5947265 |
| [49] |
P. C. Rodrigues, J. Pimentel, P. Messala, M. Kazemi, The decomposition and forecasting of mutual investment funds using singular spectrum analysis, Entropy, 22 (2020), 83. https://doi.org/10.3390/e22010083 doi: 10.3390/e22010083
|
| [50] |
R. Mahmoudvand, P. C. Rodrigues, A new parsimonious recurrent forecasting model in singular spectrum analysis, J. Forecasting, 37 (2018), 191–200. https://doi.org/10.1002/for.2484 doi: 10.1002/for.2484
|
| [51] |
P. C. Rodrigues, O. O. Awe, J. S. Pimentel, R. Mahmoudvand, Modelling the behaviour of currency exchange rates with singular spectrum analysis and artificial neural networks, Stats, 3 (2020), 137–157. https://doi.org/10.3390/stats3020012 doi: 10.3390/stats3020012
|
| [52] |
Y. Ye, J. Zhang, Q. Yang, S. Meng, J. Wang, Complete ensemble empirical mode decomposition with adaptive noise for dynamic response reconstruction of spacecraft structures under random vibration, Insight-Non-Destruct. Test. Cond. Monitor., 65 (2023), 666–674. https://doi.org/10.1784/insi.2023.65.12.666 doi: 10.1784/insi.2023.65.12.666
|
| [53] |
D. Wang, H. Luo, O. Grunder, Y. Lin, H. Guo, Multi-step ahead electricity price forecasting using a hybrid model based on two-layer decomposition technique and BP neural network optimized by firefly algorithm, Appl. Energ., 190 (2017), 390–407. https://doi.org/10.1016/j.apenergy.2016.12.134 doi: 10.1016/j.apenergy.2016.12.134
|
| [54] |
N. Mbuli, M. Mathonsi, M. Seitshiro, J. H. C. Pretorius, Decomposition forecasting methods: A review of applications in power systems, Energy Rep., 6 (2020), 298–306. https://doi.org/10.1016/j.egyr.2020.11.238 doi: 10.1016/j.egyr.2020.11.238
|
| [55] |
F. Guo, S. Deng, W. Zheng, A. Wen, J. Du, G. Huang, et al., Short-term electricity price forecasting based on the two-layer VMD decomposition technique and SSA-LSTM, Energies, 15 (2022), 8445. https://doi.org/10.3390/en15228445 doi: 10.3390/en15228445
|
| [56] |
S. Wu, G. Xia, L. Liu, A novel decomposition integration model for power coal price forecasting, Resour. Policy, 80 (2023), 103259. https://doi.org/10.1016/j.resourpol.2022.103259 doi: 10.1016/j.resourpol.2022.103259
|
| [57] | E. J. Hannan, L. Kavalieris, Regression, autoregression models, J. Time Ser. Anal., 7 (1986), 27–49. https://doi.org/10.1111/j.1467-9892.1986.tb00484.x |
| [58] |
J. P. Kreiss, J. Franke, Bootstrapping stationary autoregressive moving-average models, J. Time Ser. Anal., 13 (1992), 297–317. https://doi.org/10.1111/j.1467-9892.1992.tb00109.x doi: 10.1111/j.1467-9892.1992.tb00109.x
|
| [59] | P. M. Robinson, Robust nonparametric autoregression, In: J. Franke, W. Härdle, D. Martin (eds), Robust and Nonlinear Time Series Analysis, Lecture Notes in Statistics, vol 26, Springer, New York, 1984,247–255. https://doi.org/10.1007/978-1-4615-7821-5_14 |
| [60] |
T. Taskaya-Temizel, M. C. Casey, A comparative study of autoregressive neural network hybrids, Neural Networks, 18 (2005), 781–789. https://doi.org/10.1016/j.neunet.2005.06.003 doi: 10.1016/j.neunet.2005.06.003
|
| [61] | A. F. Di Narzo, J. L. Aznarte, M. Stigler, Package 'tsDyn'. Available from: http://cran.r-project.org. |
| [62] | R. J. Hyndman, G. Athanasopoulos, C. Bergmeir, G. Caceres, L. Chhay, M. O'Hara-Wild, et al., Package 'forecast', 2026. Available from: https://cran.r-project.org/web/packages/forecast/forecast.pdf. |
| [63] |
T. Hayfield, J. S. Racine, Nonparametric econometrics: The np package, J. Stat. Softw., 27 (2008), 1–32. https://doi.org/10.18637/jss.v027.i05 doi: 10.18637/jss.v027.i05
|
| [64] | F. X. Diebold, Comparing predictive accuracy, twenty years later: A personal perspective on the use and abuse of Diebold–Mariano tests, J. Bus. Econ. Stat., 33 (2015). https://doi.org/10.1080/07350015.2014.983236 |
| [65] |
H. Iftikhar, F. Khan, E. A. Torres Armas, P. C. Rodrigues, J. L. López-Gonzales, A novel hybrid framework for forecasting stock indices based on the nonlinear time series models, Comput. Stat., 40 (2025), 4163–4186. https://doi.org/10.1007/s00180-025-01614-5 doi: 10.1007/s00180-025-01614-5
|
| [66] |
H. Iftikhar, M. Khan, J. E. Turpo-Chaparro, P. C. Rodrigues, J. L. López-Gonzales, Forecasting stock prices using a novel filtering-combination technique: Application to the pakistan stock exchange, AIMS Math., 9 (2024), 3264. https://doi.org/10.3934/math.2024159 doi: 10.3934/math.2024159
|
| [67] |
H. Iftikhar, A. Zafar, J. E. Turpo-Chaparro, P. Canas Rodrigues, J. L. López-Gonzales, Forecasting day-ahead brent crude oil prices using hybrid combinations of time series models, Mathematics, 11 (2023), 3548. https://doi.org/10.3390/math11163548 doi: 10.3390/math11163548
|
| [68] |
H. Iftikhar, M. Qureshi, P. C. Rodrigues, M. U. Iftikhar, J. L. López-Gonzales, Daily crude oil prices forecasting using a novel hybrid time series technique, IEEE Access, 13 (2025), 98822–98836. https://doi.org/10.1109/ACCESS.2025.3574788 doi: 10.1109/ACCESS.2025.3574788
|
Figures(2) / Tables(14)
Hasnain Iftikhar, Said Farooq Shah, Fatimah E. Almuhayfith, Paulo Canas Rodrigues. A kernel regression-based hybrid forecasting framework for highly volatile energy commodity markets[J]. AIMS Mathematics, 2026, 11(8): 24877-24912. doi: 10.3934/math.20261000
DownLoad: