Evaluating the Performance of a Hybrid Model in Improving Key Risk Management Metrics in Financial Markets

Morteza  Vahdati; Mohammad Reza Karimi; Abdolrahim  Rahimi

Authors

Morteza Vahdati Department of Media Management, ST.C., Islamic Azad University, Tehran, Iran Author https://orcid.org/0009-0002-0114-0833
Mohammad Reza Karimi Department of Media Management, ST.C., Islamic Azad University, Tehran, Iran https://orcid.org/0000-0003-2560-070X
Abdolrahim Rahimi Department of Cultural Management and Planning, ST.C., Islamic Azad University, Tehran, Iran Author https://orcid.org/0000-0001-7861-8634

Keywords:

Hybrid model, financial risk management, FinBERT, long short-term memory, investor sentiment, cryptocurrency, value at risk, expected shortfall

Abstract

This study aimed to design and evaluate a hybrid financial–behavioral model integrating technical market indicators, volatility features, and social-media sentiment to improve key risk-management metrics in Bitcoin and Ethereum markets. This quantitative, ex post facto, developmental–applied study used hourly and daily financial data for Bitcoin and Ethereum from January 2019 to December 2025, together with English-language posts from Twitter/X containing relevant cryptocurrency keywords and hashtags. Financial variables included prices, trading volume, returns, volatility measures, and more than 50 technical indicators. Textual data were processed using natural language processing procedures, and sentiment scores were extracted through a finance-specific BERT model. The proposed FinBERT–LSTM model combined financial, technical, and behavioral features. Its performance was compared with naïve persistence, ARIMA, GARCH, random forest, financial LSTM, and sentiment-only models using chronological train–validation–test partitions and walk-forward validation. Predictive accuracy, directional classification, value-at-risk calibration, maximum drawdown, expected shortfall, and risk-adjusted performance were evaluated. The hybrid model significantly outperformed all benchmark models, achieving the lowest mean absolute error, root mean squared error, and mean absolute percentage error, as well as the highest directional accuracy, F1 score, and area under the curve. Diebold–Mariano tests confirmed significantly lower forecasting errors than the financial LSTM and sentiment-only models (p < 0.001). The hybrid strategy also produced lower annualized volatility, downside deviation, maximum drawdown, value at risk, and expected shortfall, while yielding higher Sharpe, Sortino, and Calmar ratios. Kupiec and Christoffersen tests indicated adequate value-at-risk coverage and independence at the 95% and 99% confidence levels. Ablation analyses further showed that removing sentiment, technical, engagement, or volatility features significantly weakened predictive and risk-management performance. Integrating technical, temporal, and behavioral information within a hybrid deep-learning architecture improves both forecasting accuracy and the management of downside and tail risk in cryptocurrency markets.

References

[1] K. Hrytsiv and J. Kartašova, "Risk in Financial Decision-Making: A Conceptual Framework for Investors and Corporate Managers," 2025, doi: 10.15837/aijes.v19i1.7145.

[2] A. O. Scott, P. Amajuoyi, and K. B. Adeusi, "Theoretical Perspectives on Risk-Management Strategies in Financial Markets: A Comparative Review of African and U.S. Approaches," 2024.

[3] H. Hassanpour, M. Vaezz Barzani, A. Moarefi Mohammadi, and S. M. Emami Jomeh, "A Comparative Analysis of the Epistemological Foundations of Classical and Behavioral Finance," Islamic Economics Knowledge, vol. 11, no. 22, pp. 57-74, 2020.

[4] J. Almeida and T. C. Gonçalves, "A Systematic Literature Review of Investor Behavior in Cryptocurrency Markets," Journal of Behavioral and Experimental Finance, vol. 37, p. 100785, 2023, doi: 10.1016/j.jbef.2022.100785.

[5] A. Jalilvand and M. Rostami Nowruzabad, "Financial Literacy, Investor Sentiment, Risk Perception, and Investment Intention: Evidence from the Tehran Stock Exchange," Investment Knowledge, vol. 7, no. 27, pp. 141-169, 2018.

[6] S. Arabi Monfared, "Examining the Effect of Illusory Financial Literacy on Behavioral Biases among Stock-Market Investors in Shiraz," in Third International Conference and Fourth National Conference on New Findings in Management, Psychology, and Accounting, Tehran, Iran, 2023.

[7] G. Gol Arzi and S. Piri, "The Effect of Investors' Emotional Decisions on Crises in the Tehran Stock Exchange," Financial Management Strategy, vol. 9, no. 37, pp. 99-116, 2022.

[8] M. Ramesheh, A. Jannati Manzari, and M. Dadpanah, "Investor Sentiment and Financing Decisions in the Tehran Stock Exchange," Financial Management Perspective, no. 34, pp. 173-193, 2021.

[9] X. Huang, H. Li, and N. Wei, "Big Data-Driven Behavioral Economics: Analysis of Decision Biases in Financial Markets," 2025, doi: 10.70339/vyet1339.

[10] O. Sattarov and J. Choi, "From Prediction to Profit: A Comprehensive Review of Cryptocurrency Trading Strategies and Price-Forecasting Techniques," IEEE Access, vol. 12, pp. 87039-87063, 2024, doi: 10.1109/ACCESS.2024.3417449.

[11] V. Ladyzhets and S. Terenchuk, "Models and Methods of Technical Analysis of Financial Markets," vol. 48, pp. 47-52, 2021, doi: 10.32347/2412-9933.2021.48.47-52.

[12] C. Cappello, A. Congedi, S. De Iaco, and L. Mariella, "Traditional Prediction Techniques and Machine-Learning Approaches for Financial Time-Series Analysis," Mathematics, vol. 13, no. 3, p. 537, 2025, doi: 10.3390/math13030537.

[13] W. A. Addy, A. O. Ajayi-Nifise, B. G. Bello, S. T. Tula, O. Odeyemi, and T. Falaiye, "Machine Learning in Financial Markets: A Critical Review of Algorithmic Trading and Risk Management," International Journal of Science and Research Archive, vol. 11, no. 1, pp. 1853-1862, 2024, doi: 10.30574/ijsra.2024.11.1.0292.

[14] J. S. Dote-Pardo, M. C. Cordero-Díaz, M. T. Espinosa Jaramillo, and J. Parra-Domínguez, "Leveraging Artificial Intelligence for Enhanced Decision-Making in Finance: Trends and Future Directions," Journal of Accounting Literature, 2025, doi: 10.1108/JAL-02-2025-0100.

[15] V. Pillai, "Integrating AI-Driven Techniques in Big-Data Analytics: Enhancing Decision-Making in Financial Markets," 2023, doi: 10.18535/ijecs/v12i07.4745.

[16] F. Fereydouni, R. Darabi, and A. A. Anvar Rostami, "Application of Artificial Intelligence in Predicting Income Smoothing," Financial Accounting and Auditing Research, vol. 12, no. 45, pp. 103-134, 2020.

[17] O. M. Ebadati, M. A. Jafari, and N. Davoudifar, "Stock-Price Prediction Using Hybrid GA-SVM Algorithms," Advances in Finance and Investment, vol. 2, no. 5, pp. 1-22, 2021.

[18] S. Sarafraz, F. Sefati, and A. Ghiasvand, "Stock-Price Prediction Using Hybrid Market Indicators and a Neuro-Fuzzy Model," in International Conference on New Ideas in Management, Economics, and Accounting, Barcelona, Spain, 2016.

[19] M. Mashari, H. Didehkhani, K. Khalili Damghani, and E. Abbasi, "Designing a Hybrid Intelligent Model to Predict Golden Points in Stock Prices," Investment Knowledge, vol. 8, no. 29, pp. 45-66, 2019.

[20] V. Vafaei Ghaeni and A. M. Kimiagari, "One-Day-Ahead Stock-Price Prediction Using a Hybrid Model," Investment Knowledge, vol. 8, no. 30, pp. 313-328, 2019.

[21] S. Faridi, M. Madanchi Zaj, A. Daneshvar, S. Shahverdiani, and F. R. Roodposhti, "Portfolio Rebalancing Based on a Combined Method of Ensemble Machine Learning and Genetic Algorithm," Journal of Financial Reporting and Accounting, vol. 21, no. 1, pp. 105-125, 2023, doi: 10.1108/JFRA-11-2021-0413.

[22] D. Sabancı, S. Kılıçarslan, and K. Adem, "Stock-Market-Price Forecasting Using Metaheuristic Algorithms and Multivariate Adaptive Regression Splines," International Journal of Intelligent Computing and Cybernetics, vol. 16, no. 4, pp. 847-866, 2023, doi: 10.1108/IJICC-02-2023-0030.

[23] M. Korablyov, O. Fomichov, D. Antonov, S. Dykyi, I. Ivanisenko, and S. Lutskyy, "Hybrid Stock-Analysis Model for Financial-Market Forecasting," 2023, doi: 10.1109/CSIT61576.2023.10324069.

[24] A. K. Bukhanets, D. A. Dudnik, and O. V. Kosnikova, "Hybrid Machine-Learning Models for Financial-Market Analysis," 2024, doi: 10.47576/2949-1908.2024.1.1.003.

[25] F. Afzal, F. Afzal, M. Kamran, and H. Pan, "Stock-Market Predictive Modeling Using a Recurrent Neural Network with Long Short-Term Memory Architecture," International Journal of Islamic and Middle Eastern Finance and Management, 2025, doi: 10.1108/IMEFM-12-2024-0632.

[26] S. Panigrahi, G. Kukreja, and S. Kumaraswamy, "Forecasting Stock-Market Behavior in BRICS Economies Using Artificial Neural Machine-Learning Models," Journal of Business and Socio-economic Development, 2025, doi: 10.1108/JBSED-05-2025-0134.

[27] Z. Mao and C. Wu, "Stock-Price-Index Prediction Based on an SSA-BiGRU-GSCV Model from the Perspective of Long Memory," Kybernetes, vol. 53, no. 12, pp. 5905-5931, 2024, doi: 10.1108/K-02-2023-0286.

[28] W. Wang and Y. Wu, "Risk Analysis of the Chinese Financial Market Using a Novel Hybrid Volatility-Prediction Model," Mathematics, vol. 11, no. 18, p. 3937, 2023, doi: 10.3390/math11183937.

[29] T. Kaur, K. Rani, P. Thakur, and N. S. Talwandi, "Enhanced Decision-Support System for Financial Risk Assessment Using Hybrid Fuzzy Logic and Machine Learning," 2024, doi: 10.1109/ICCICA60014.2024.10584834.

[30] B. Bunkar and K. Ramaiah, "A Serial Mediation Model for Investigating the Intention to Use Algorithmic Trading Platforms among Retail Investors in India," VILAKSHAN: XIMB Journal of Management, vol. 21, no. 2, pp. 263-280, 2024, doi: 10.1108/XJM-12-2023-0233.

[31] W. C. Hsu, M. C. Wang, and H. I. Ting, "Reducing Emotional Bias in Investment Decisions: The Role of GPT-4 in Financial Analysis," Asia-Pacific Journal of Business Administration, pp. 1-26, 2025, doi: 10.1108/APJBA-03-2025-0181.

[32] L. Bani Fatemeh and S. Ghaffari, "Artificial Intelligence, the Semantic Web, and Their Interaction: A Review," Scientific Quarterly, vol. 10, no. 21, pp. 61-85, 2021.

[33] X. Ren, W. Jiang, X. Sun, and S. Wang, "Investor Sentiment from Images: A Few-Shot Learning Investigation," Journal of Accounting Literature, pp. 1-30, 2025, doi: 10.1108/JAL-04-2025-0199.

[34] K. Nyakurukwa and Y. Seetharam, "Leveraging Sentiment Connectedness for Portfolio Optimisation: A Minimum-Connectedness Approach," Review of Behavioral Finance, pp. 1-10, 2025, doi: 10.1108/RBF-03-2025-0125.

[35] S. Nayak, "Developing Predictive Models for Financial Stability: Integrating Behavioral Analytics into Credit-Risk Management," 2024, doi: 10.47363/JAICC/2024(3)E173.

[36] M. Lobato, J. Rodríguez, and H. Romero-Perez, "Overseas Listing and Risk: The Case of Single-Listed Chinese ADRs," Managerial Finance, pp. 1-14, 2025, doi: 10.1108/MF-10-2024-0753.