Artificial Intelligence in Stock Market Trading - A Comprehensive Survey of Models

Abstract

With financial markets becoming increasingly complex and volatile, traditional statistical and human-driven approaches are proving inadequate for capturing the nonlinear and dynamic nature of stock trading. Artificial Intelligence (AI) has therefore emerged as a transformative force, employing advanced algorithms and deep learning to identify hidden patterns, forecast prices, and inform trading decisions. This paper presents a comprehensive bibliometric survey of 9,088 scholarly works spanning 1971 to 2025, offering the most detailed review of AI applications in stock market trading to date. Using the SPAR-4-SLR framework and bibliometric tools such as Biblioshiny and VOS viewer, the study maps intellectual contributions, identifies key research clusters, and analyses collaboration networks across the field. The results reveal dominant methodologies including neural networks, long short-term memory (LSTM), reinforcement learning, and hybrid approaches, while also highlighting the growing importance of Explainable AI (XAI) and ESG-aligned frameworks. Contributions from East Asian institutions, particularly in China, stand out, although significant inputs from Europe and North America are also observed. Despite these advances, challenges persist in areas such as interpretability, real-time adaptability, and the integration of alternative data sources like sentiment analysis and satellite imagery. Future research directions emphasize the development of quantum AI, reinforcement learning-based adaptive systems, and ethical regulatory frameworks that ensure responsible innovation. By bridging theory and practice, this study provides an intellectual roadmap and practical recommendations for researchers, practitioners, and policymakers. Overall, the findings underscore the urgency of advancing transparency, robustness, and interdisciplinary collaboration to ensure AI-driven trading systems contribute to sustainable financial innovation and trustworthy decision-making.