- Version
- Download 0
- File Size 1.42 MB
- File Count 1
- Create Date July 24, 2025
- Last Updated July 24, 2025
Predicting Stock Prices In Vietnam’s Financial Market Using ARIMA And Neural Network Models: A Comparative Study
This study investigates the effectiveness of two popular time series forecasting models—ARIMA (AutoRegressive Integrated Moving Average) and LSTM (Long Short-Term Memory neural network)—in predicting stock prices within Vietnam's financial market. Using daily closing price data from five major Vietnamese stocks (FPT, VIC, VCB, GAS, and MWG), the research evaluates each model's prediction accuracy over a six-month testing period. The models were trained using data from January 2022 to December 2023 and assessed using two key error metrics: Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE).
The results reveal that the LSTM model consistently outperforms the ARIMA model across all selected stocks, achieving significantly lower MAE and RMSE values. While ARIMA provides acceptable performance on stocks with more stable price trends, it struggles with volatile or nonlinear series, where LSTM demonstrates superior adaptability and robustness. These findings highlight the potential of neural network models as effective tools for stock price forecasting in emerging markets like Vietnam.
The study provides practical insights for investors and financial analysts, suggesting that incorporating deep learning techniques into forecasting strategies can lead to more informed investment decisions. Future research is recommended to explore hybrid models, real-time forecasting, and the integration of external variables for enhanced predictive performance.
Author: MSc. Vo An Hai A research paper submitted for the degree of Doctor of Business Administration (D.B.A.) in Finance European International University, Paris, France