Author(s)

ChaoYan Wei^*, LanLan Li, PangLeYi Chen, MeiHui Huang, HuiLin Wei, KunYao Yao, XuYang Wang, Xin Ya, ChaoHai Wei

Affiliation(s)

Institute of Information Technology, Guangxi Police College, Nanning 530028, Guangxi, China.

Corresponding Author

ChaoYan Wei

ABSTRACT

As financial markets become increasingly complex, the demand for stock price forecasting is growing. To capture both linear trends and volatility in sequences as well as nonlinear dependencies, this paper proposes an ARIMA-GARCH-LSTM hybrid model. First, ARIMA is used to extract linear factors, followed by GARCH to express residual volatility conditions, and finally LSTM to capture deep nonlinear features. Based on the closing prices of the Shanghai Composite Index over 1,027 trading days from 2021 to 2025, RMSE, MAE, and MAPE were used for moving forecasts and multi-indicator estimates. The experiments show that the hybrid model outperforms individual ARIMA, GARCH, or LSTM models in all metrics, confirming its accuracy and robustness. Additionally, the hybrid model demonstrates strong adaptability during periods of high volatility.

KEYWORDS

Hybrid model; Stock price forecast; ARIMA model; GARCH family model; LSTM model

CITE THIS PAPER

ChaoYan Wei, LanLan Li, PangLeYi Chen, MeiHui Huang, HuiLin Wei, KunYao Yao, XuYang Wang, Xin Ya, ChaoHai Wei. Stock price research based on ARIMA-GARCH-LSTM hybrid model. Journal of Trends in Financial and Economics. 2025, 2(2): 36-43. DOI: https://doi.org/10.61784/jtfe3043.

REFERENCES

[1] Wang J, Zhang T, Lu T, et al. A Hybrid Forecast Model of EEMD-CNN-ILSTM for Crude Oil Futures Price. Electronics, 2023, 12(11): 2521.

[2] Ting Y H, Long Y B, Lin D, et al. Application of a Hybrid Model Based on ICEEMDAN, Bayesian Hyperparameter Optimization GRU and the ARIMA in Nonferrous Metal Price Prediction. Cybernetics and Systems, 2023, 54(1): 27-59.

[3] Sweeti S, Surendiran B, Dhanalakshmi R, et al. Forecasting COVID-19 Pandemic Using Prophet, ARIMA, and Hybrid Stacked LSTM-GRU Models in India. Computational and mathematical methods in medicine,2022, 1556025. DOI: https://doi.org/10.1155/2022/1556025.

[4] Liu Qi. Stock Forecasting Based on GARCH-BiLSTM Network. Changchun University of Science and Technology, 2024. DOI: 10.26977/d.cnki.gccgc.2024.000613.

[5] Xu Shuya, Liang Xiaoying. Research on Stock Price Prediction Based on ARIMA-GARCH Model. Journal of Henan Institute of Education (Natural Science Edition), 2019, 28(04): 20-24.

[6] Wang Xiaorui. Application of ARFIMA-GARCH-LSTM Model in the Prediction of Hand, Foot and Mouth Disease in Shanxi Province. Shanxi Medical University, 2023. DOI: 10.27288/d.cnki.gsxyu.2023.000726.

[7] Jiang Min, Zhang Chuyi, Sun Deshan. Stock price prediction based on PCA-GARCH-LSTM model. Software Guide, 2025, 24(01): 43-48.

[8] Song Zhifang. Research on the Prediction of China's Inflation Rate Based on CEEMDAN-GARCH-LSTM Model. North Minzu University, 2023. DOI: 10.27754/d.cnki.gbfmz.2023.000181.

[9] Jiang Mincong. Research on Option Trading Strategies Based on LSTM Model to Predict Volatility. University of Electronic Science and Technology of China, 2024. DOI: 10.27005/d.cnki.gdzku.2024.003958.

[10] Hu Yamei. Stock Price Prediction Based on GARCH Family Models and LSTM Models [C]// Proceedings of the Comprehensive Innovation and Development Academic Forum, Chongqing Dingyun Culture Communication Co., Ltd. School of Statistics and Mathematics, Guangdong University of Finance and Economics; 2023, 106-109. DOI: 10.26914/c.cnkihy.2023.023020.

[11] Yongchao J, Renfang W, Xiaodie Z, et al. Prediction of COVID-19 Data Using an ARIMA-LSTM Hybrid Forecast Model. Mathematics, 2022, 10(21): 4001.

[12] Ma C, Wu J, Hu H, et al. Predicting Stock Prices Using Hybrid LSTM and ARIMA Model. IAENG International Journal of Applied Mathematics, 2024, 54(3): 424-432.