Stock former: A Transformer-Based Profit-Driven Model for Financial Time-Series Forecasting in the Indian Stock Market

RANJIT CHOUDHARY

doi:10.70454/JRICST.2026.30301

Authors

RANJIT CHOUDHARY AMITY INSTITUTE OF INFORMATION TECHNOLOGY, AMITY UNIVERSITY, PATNA Author

DOI:

https://doi.org/10.70454/JRICST.2026.30301

Keywords:

Stock former, Transformer, Financial Time-Series Forecasting, NIFTY Bank, Deep Learning, Granger Causality, Profit Optimization, AI in Finance, Self-Attention, Algorithmic Trading

Abstract

This research presents Stock former, a Transformer-based deep learning model for financial time-series forecasting in the Indian stock market. Using hourly data from the top five NIFTY Bank stocks such as HDFC Bank, ICICI Bank, SBI, Kodak Bank, and Axis Bank, the model leverages the self-attention mechanism to capture temporal and inter-stock dependencies [3]. A Granger Causality test is employed to identify the most influential stock before training. The architecture integrates 1D-CNN feature extraction, Transformer encoding, and a profit-oriented Stock Tan Loss function to directly optimize trading performance. Results indicate improved predictive accuracy and robust trading signals, offering a scalable and interpretable framework for AI-driven financial forecasting.

Downloads

Download data is not yet available.

References

[1] T. Fischer and C. Krauss, "Deep learning with long short-term memory networks for financial market predictions," European Journal of Operational Research, vol. 270, no. 2, pp. 654–669, 2018.

[2] W. Lu, J. Li, Y. Li, A. Sun, and J. Wang, "ACNN-LSTM-based model to forecast stock prices," Complexity, vol. 2020, Art. no. 6622927, 2020.

[3] A. Vaswani et al., "Attention Is All You Need," in Advances in Neural Information Processing Systems (NeurIPS), vol. 30, 2017.

[4] H. Zhou et al., "Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting," in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, no. 12, pp. 11106–11115, 2021.

[5] B. Lim, S. Ö. Arık, N. Loeff, and T. Pfister, "Temporal Fusion Transformers for Interpretable Multi-Horizon Time Series Forecasting," International Journal of Forecasting, vol. 37, no. 4, pp. 1748–1764, 2021.

[6] J. Saleena, C. Jessy John, and G. Rubell Marion Lincy, "A Phase-Cum-Time Variant Fuzzy Time Series Model for Forecasting Non-Stationary Time Series and Its Application to the Stock Market," IEEE Access, vol. 12, 2024.

[7] A. Biem, H. Feng, A. V. Riabov, and D. S. Turaga, "Real-Time Analysis and Management of Big Time-Series Data," IBM Journal of Research and Development, vol. 57, no. 3/4, 2013.

[8] C. Xie, A. Bijral, and J. Lavista Ferres, "NonSTOP: A NonSTationary Online Prediction Method for Time Series," IEEE Transactions on Knowledge and Data Engineering, vol. 25, no. 10, 2018.

[9] W. Shi, D. Karastoyanova, Y. Huang, and G. Zhang, "Bidirectional Piecewise Linear Representation of Time Series and Its Application in Clustering," Knowledge-Based Systems, vol. 72, 2023.

[10] Z. Wang, J. Fan, H. Wu, D. Sun, and J. Wu, "Representing Multiview Time-Series Graph Structures for Multivariate Long-Term Time-Series Forecasting," IEEE Transactions on Artificial Intelligence, vol. 5, no. 6, 2024.

[11] S. Saha, F. Bovolo, and L. Bruzzone, "Change Detection in Image Time-Series Using Unsupervised LSTM," IEEE Geoscience and Remote Sensing Letters, vol. 19, 2022.

[12] T. Sharma, S. K. Prasad, S. Prasad, I. Verma, and A. Sharma, "Forecasting Stock Market Volatility Using XGBoost: A Time Series Analysis," in Proceedings of the International Conference, 2024.

[13] G. Li, M. Xiao, and Y. Guo, "Application of Deep Learning in Stock Market Valuation Index Forecasting," in Proceedings of the International Conference, 2019.

[14] M. Zhou, J. Yi, J. Yang, and Y. Sima, "Characteristic Representation of Stock Time Series Based on Trend Feature Points," IEEE Access, vol. 8, 2020.

[15] T. Liu, J. Li, Z. Zhang, H. Yu, and S. Gao, "FD-GRNet: A Dendritic-Driven GRU Framework for Advanced Stock Market Prediction," IEEE Access, vol. 13, 2025.

[16] B. Tanuwijaya, G. Selvachandran, L. H. Son, M. Abdel-Basset, H. X. Huynh, V.-H. Pham, and M. Ismail, "A Novel Single Valued Neutrosophic Hesitant Fuzzy Time Series Model: Applications in Indonesian and Argentinian Stock Index Forecasting," IEEE Access, vol. 8, 2020.