Systematic Review of Deep Learning and Hybrid Architectures for Forecasting Forex and Precious Metals: Trends, Performance, and Challenges
Article Sidebar
Main Article Content
Abstract
The non-stationary, volatile, and non-linear characteristics of financial markets make forecasting foreign-exchange and precious-metal prices methodologically challenging. This study systematically reviews the development, performance, and limitations of deep learning and hybrid architectures used for financial time-series forecasting. Literature published between 2020 and 2025 was identified through selected scholarly databases and screened using predefined inclusion, exclusion, and quality-assessment criteria. Twenty empirical studies were retained and analyzed through a task-stratified narrative synthesis covering architectural functions, input-fusion strategies, forecasting targets, evaluation metrics, and practical limitations. Among the included studies, 13 employed hybrid, ensemble, decomposition-based, or multimodal architectures, whereas seven examined standalone or comparative deep learning models. Eight studies explicitly incorporated non-price information, including technical indicators, macroeconomic variables, financial news, sentiment, and limit-order-book data. Evaluation practices were heterogeneous: 15 studies primarily reported regression-error metrics, three emphasized classification, directional, or economic-performance measures, and two used mixed evaluation schemes. Hybrid models frequently outperformed standalone baselines within individual studies; however, cross-study superiority could not be established because of differences in datasets, targets, frequencies, horizons, validation procedures, and metrics. Only two studies explicitly assessed economic usefulness, revealing a gap between statistical accuracy and practical financial value. The review is limited to 20 English-language studies published during 2020–2025. Its principal contribution is a task-aware mapping of architectural developments, data-fusion strategies, evaluation practices, and unresolved gaps in the generalizability and practical utility of financial forecasting models.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
By signing this statement, I hereby assign and transfer to JCSIS all exclusive copyright rights relating to the work identified above. These rights include, without limitation, the authority to publish, reproduce, revise, adapt, distribute, transmit, market, sell, and otherwise use the work and any associated materials worldwide, either in full or in part, in any language and through any existing or future form of electronic, printed, digital, or other media. JCSIS may also authorize or license third parties to exercise any of these rights. understand that these exclusive rights will be vested in JCSIS from the date on which the article is formally accepted for publication. As the copyright owner, JCSIS will have the exclusive authority to approve, license, or permit the reproduction, distribution, and other uses of the article. Nevertheless, all proprietary rights that are separate from copyright, including patent rights and rights relating to any process, method, or procedure described in the work, will remain with the author. I further acknowledge that JCSIS permits authors to reuse and share their published articles in accordance with the terms and conditions of the applicable Creative Commons license.
References
M. S. M. Bhuiyan et al., “Deep learning for algorithmic trading: A systematic review of predictive models and optimization strategies,” Array, vol. 26, no. November 2024, p. 100390, 2025, doi: https://doi.org/10.1016/j.array.2025.100390.
F. Dakalbab, A. Kumar, M. A. Talib, and Q. Nasir, “Advancing Forex prediction through multimodal text-driven model and attention mechanisms,” Intell. Syst. with Appl., vol. 26, no. December 2024, p. 200518, 2025, doi: https://doi.org/10.1016/j.iswa.2025.200518.
L. Boongasame, P. Viriyaphol, K. Tassanavipas, and P. Temdee, “Gold-Price Forecasting Method Using Long Short-Term Memory and the Association Rule,” J. Mob. Multimed., vol. 19, no. 1, pp. 165–186, 2023, doi: https://doi.org/10.13052/jmm1550-4646.1919.
Y. Zhao, Y. Guo, and X. Wang, “Hybrid LSTM–Transformer Architecture with Multi-Scale Feature Fusion for High-Accuracy Gold Futures Price Forecasting,” Mathematics, vol. 13, no. 10, 2025, doi: https://doi.org/10.3390/math13101551.
M. Salim and A. Djunaidy, “Development of a CNN-LSTM Approach with Images as Time-Series Data Representation for Predicting Gold Prices,” Procedia Comput. Sci., vol. 234, pp. 333–340, 2024, doi: https://doi.org/10.1016/j.procs.2024.03.007.
M. Nabipour, P. Nayyeri, H. Jabani, S. S., and A. Mosavi, “Predicting Stock Market Trends Using Machine Learning and Deep Learning Algorithms Via Continuous and Binary Data; a Comparative Analysis,” IEEE Access, vol. 8, pp. 150199–150212, 2020, doi: https://doi.org/10.1109/ACCESS.2020.3015966.
M. S. Islam and E. Hossain, “Foreign exchange currency rate prediction using a GRU-LSTM hybrid network,” Soft Comput. Lett., vol. 3, no. October 2020, p. 100009, 2021, doi: https://doi.org/10.1016/j.socl.2020.100009.
D. Alaminos, M. B. Salas, and A. Partal-Ureña, “Hybrid ARMA-GARCH-Neural Networks for intraday strategy exploration in high-frequency trading,” Pattern Recognit., vol. 148, no. March 2023, 2024, doi: https://doi.org/10.1016/j.patcog.2023.110139.
A. Saghafi, M. Bagherian, and F. Shokoohi, “Forecasting Forex EUR/USD Closing Prices Using a Dual-Input Deep Learning Model with Technical and Fundamental Indicators,” Mathematics, vol. 13, no. 9, pp. 1–17, 2025, doi: https://doi.org/10.3390/math13091472.
S. Mou, Q. Xue, J. Chen, T. Takiguchi, and Y. Ariki, “MM-iTransformer: A Multimodal Approach to Economic Time Series Forecasting with Textual Data,” Appl. Sci., vol. 15, no. 3, p. 1241, Jan. 2025, doi: https://doi.org/10.3390/app15031241.
M. Nabipour, P. Nayyeri, H. Jabani, A. Mosavi, E. Salwana, and S. S., “Deep Learning for Stock Market Prediction,” Entropy, vol. 22, no. 8, p. 840, Jul. 2020, doi: https://doi.org/10.3390/e22080840.
S. B. Primananda and S. M. Isa, “Forecasting Gold Price in Rupiah using Multivariate Analysis with LSTM and GRU Neural Networks,” Adv. Sci. Technol. Eng. Syst. J., vol. 6, no. 2, pp. 245–253, Mar. 2021, doi: https://doi.org/10.25046/aj060227.
M. Ayitey Junior, P. Appiahene, and O. Appiah, “Forex market forecasting with two-layer stacked Long Short-Term Memory neural network (LSTM) and correlation analysis,” J. Electr. Syst. Inf. Technol., vol. 9, no. 1, pp. 1–24, 2022, doi: 1https://doi.org/0.1186/s43067-022-00054-1.
M. Z. Abedin, M. H. Moon, M. K. Hassan, and P. Hajek, “Deep learning-based exchange rate prediction during the COVID-19 pandemic,” Ann. Oper. Res., vol. 345, no. 2, pp. 1335–1386, 2025, doi: https://doi.org/10.1007/s10479-021-04420-6.
T. Fischer, M. Sterling, and S. Lessmann, “Fx-spot predictions with state-of-the-art transformer and time embeddings,” Expert Syst. Appl., vol. 249, p. 123538, Sep. 2024, doi: https://doi.org/10.1016/j.eswa.2024.123538.
P. Foroutan and S. Lahmiri, “Deep learning systems for forecasting the prices of crude oil and precious metals,” Financ. Innov., vol. 10, no. 1, 2024, doi: https://doi.org/10.1186/s40854-024-00637-z.
S. Ahmed, S.-U. Hassan, N. R. Aljohani, and R. Nawaz, “FLF-LSTM: A novel prediction system using Forex Loss Function,” Appl. Soft Comput., vol. 97, p. 106780, Dec. 2020, doi: https://doi.org/10.1016/j.asoc.2020.106780.
A. Rostamian and J. G. O’Hara, “Event prediction within directional change framework using a CNN-LSTM model,” Neural Comput. Appl., vol. 34, no. 20, pp. 17193–17205, Oct. 2022, doi: https://doi.org/10.1007/s00521-022-07687-3.
K. Ito, H. Iima, and Y. Kitamura, “LSTM forecasting foreign exchange rates using limit order book,” Financ. Res. Lett., vol. 47, p. 102517, Jun. 2022, doi: https://doi.org/10.1016/j.frl.2021.102517.
Y. Mao, Z. Chen, S. Liu, and Y. Li, “Unveiling the potential: Exploring the predictability of complex exchange rate trends,” Eng. Appl. Artif. Intell., vol. 133, p. 108112, Jul. 2024, doi: https://doi.org/10.1016/j.engappai.2024.108112.
L. Huo, Y. Xie, and J. Li, “An Innovative Deep Learning Futures Price Prediction Method with Fast and Strong Generalization and High-Accuracy Research,” Appl. Sci., vol. 14, no. 13, p. 5602, Jun. 2024, doi: https://doi.org/10.3390/app14135602.
G. Taneva-Angelova, S. Raychev, and G. Ilieva, “A Framework for Gold Price Prediction Combining Classical and Intelligent Methods with Financial, Economic, and Sentiment Data Fusion,” Int. J. Financ. Stud., vol. 13, no. 2, pp. 1–25, 2025, doi: https://doi.org/10.3390/ijfs13020102.
M. Sarıkoç and M. Celik, “PCA-ICA-LSTM: A Hybrid Deep Learning Model Based on Dimension Reduction Methods to Predict S&P 500 Index Price,” Comput. Econ., vol. 65, no. 4, pp. 2249–2315, Apr. 2025, doi: https://doi.org/10.1007/s10614-024-10629-x.
A. Hadizadeh, M. J. Tarokh, and M. M. Ghazani, “A novel transformer-based dual attention architecture for the prediction of financial time series,” J. King Saud Univ. Comput. Inf. Sci., vol. 37, no. 5, p. 72, Jul. 2025, doi: https://doi.org/10.1007/s44443-025-00045-y.
Y. Liang, Y. Lin, and Q. Lu, “Forecasting gold price using a novel hybrid model with ICEEMDAN and LSTM-CNN-CBAM,” Expert Syst. Appl., vol. 206, p. 117847, Nov. 2022, doi: https://doi.org/10.1016/j.eswa.2022.117847.
A. Hermawan, L. Lunardi, Y. Kurnia, B. Daniawan, and Junaedi, “Optimizing Convolutional Neural Networks with Particle Swarm Optimization for Enhanced Hoax News Detection,” J. Inf. Syst. Eng. Bus. Intell., vol. 11, no. 1, pp. 53–64, Mar. 2025, doi: https://doi.org/10.20473/jisebi.11.1.53-64.
Y. Kurnia, E. D. Kusuma, L. W. Kusuma, Suwitno, and W. Apridius, “Perbandingan Naïve Bayes dan CNN yang Dioptimasi PSO pada Identifikasi Berita Hoax Politik Indonesia,” bit-Tech, vol. 6, no. 3, pp. 340–352, Apr. 2024, doi: https://doi.org/10.32877/bt.v6i3.1225.
R. Kaushik, S. Jain, S. Jain, and T. Dash, “Performance evaluation of deep neural networks for forecasting time‐series with multiple structural breaks and high volatility,” CAAI Trans. Intell. Technol., vol. 6, no. 3, pp. 265–280, Sep. 2021, doi: https://doi.org/10.1049/cit2.12002.
A. Casolaro, V. Capone, G. Iannuzzo, and F. Camastra, “Deep Learning for Time Series Forecasting: Advances and Open Problems,” Information, vol. 14, no. 11, p. 598, Nov. 2023, doi: https://doi.org/10.3390/info14110598.
Y. Zhang, W. Yang, J. Wang, Q. Ma, and J. Xiong, “CAMEF: Causal-Augmented Multi-Modality Event-Driven Financial Forecasting by Integrating Time Series Patterns and Salient Macroeconomic Announcements,” in Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2, New York, NY, USA: ACM, Aug. 2025, pp. 3867–3878. doi: https://doi.org/10.1145/3711896.3736872.