Integrating Sentiment Analysis into Mean-Variance Portfolio Optimization: Theory, Implementation, and Empirical Performance Evaluation

Classical mean-variance (MV) portfolio optimization, while foundational in modern finance, assumes market efficiency and rational expectations, neglecting behavioral biases that significantly affect market outcomes. Recent advancements suggest that incorporating investor sentiment can substantially enhance optimization accuracy and portfolio performance. In this study, we propose a sentiment-driven mean-variance optimization framework, explicitly integrating sentiment analytics derived from news articles into the traditional Markowitz optimization model. We rigorously formulate a sentiment-adjusted MV model, mathematically deriving adjustments to expected asset returns based on quantified sentiment indicators derived through advanced natural language processing (NLP). Our implementation employs a structured pipeline consisting of real-time market data acquisition via Yahoo Finance, sentiment extraction from financial news using retrieval-augmented generation (RAG), and subsequent sentiment quantification to adjust expected returns in the MV optimization model. Empirical analysis demonstrates that sentiment-driven adjustments mitigate estimation errors inherent in purely historical-data-driven models, yielding port- folios with superior risk-adjusted performance (Sharpe ratios improved by 8–12% empirically). Visualized through clear and insightful graphical representations—including optimized portfolio allocations, sentiment-versus-weight comparisons, and an adjusted efficient frontier—the results indicate that integrating sentiment signals leads to more robust and realistic portfolio allocations. This study contributes a comprehensive methodology supported by theoretical rigor, practical implementation details, and empirical validation, underscoring the value of sentiment analysis in quantitative portfolio management and opening path- ways for further behavioral finance integration into quantitative investment strategies.