Stock Price Prediction: A Comprehensive Review of Methods and Trends

Abstract

Stock price prediction is an important problem in financial research. It is related to applications in investment, risk management, and algorithmic trading. However, accurate stock price prediction has a lot of challenges because stock price is affected by various factors such as market noise, non-stationarity, and external factors such as macroeconomic events, corporate news, and investor sentiment. In this review, we provide a comprehensive overview of traditional statistical models, machine learning approaches, emerging deep learning, and multimodal methods for predicting stock prices. We further discuss the recent application of large language models for sentiment extraction and direct stock price prediction. In addition, this review covers commonly used features in stock price prediction, including technical indicators, sentiment measures, and composite features, which are mathematical combinations of different basic features. Moreover, unconventional features such as environmental, social, and governance factors are included. We also discuss key challenges and open research directions, aiming to guide researchers in selecting suitable methodologies and identifying promising opportunities for future exploration.   Received: 11 September 2025 | Revised: 27 November 2025 | Accepted: 23 January 2026   Conflicts of Interest The author declares that he has no conflicts of interest to this work.   Data Availability Statement Data sharing is not applicable to this article as no new data were created or analyzed in this study.   Author Contribution Statement Qizhao Chen: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization.