Predicting Islamic bank stocks using x (formerly twitter) investor sentiment and deep learning models

Purpose This study aims to examine how investor sentiment extracted from X (formerly Twitter) influences the prediction of Islamic bank stock performance using deep learning methods. The objective is to improve understanding of stock behaviour within the Islamic finance sector by integrating sentiment indicators into forecasting models. Design/methodology/approach Sentiment analysis is applied to X data collected during both trading and non-trading hours. Three deep learning architectures, long short-term memory (LSTM), Bidirectional LSTM and a hybrid convolutional neural networks (CNN)-LSTM model, are implemented alongside market control variables to assess forecasting performance across different time horizons. Prediction accuracy is evaluated using root mean square error, mean absolute error and mean square error. Findings Incorporating sentiment variables significantly enhances predictive accuracy. The hybrid CNN-LSTM model achieves the strongest performance, outperforming all other models. The Diebold–Mariano (DM) test confirms that sentiment-augmented models deliver statistically superior forecasts, highlighting the influential role of investor sentiment in predicting Islamic bank stock performance. Practical implications Investor sentiment derived from social media, particularly X (Twitter), substantially improves stock prediction models. This has important implications for investors, analysts and policymakers seeking to understand sentiment-driven market behaviour and strengthen decision-making in Islamic finance. Originality/value In the context of Islamic finance, this study contributes to the emerging literature by integrating real-time sentiment signals from X (Twitter), explicitly distinguishing between trading and non-trading hours, with deep learning models to forecast Islamic bank stock performance. By combining social media sentiment with CNN-LSTM-based architectures, the research offers a novel approach to predicting Islamic financial markets, enhancing forecasting accuracy and addressing a notable gap in the literature by demonstrating the critical role of investor sentiment in Islamic finance.