Human-Machine Collaboration in Healthcare Investment Integrating Large Language Models

Large language models (LLMs) have rapidly advanced and are increasingly utilized in Finance. These technologies offer innovative decision-making tools for financial investors, particularly as traditional methods often prove insufficient. This study proposes an investment analysis model for the pharmaceutical and biotechnology industry that combines LLMs with human analysts' expertise. By comprehensively analyzing research reports, technical indicators, and corporate governance data, our approach effectively extracts key information such as market sentiment, corporate governance status, and technological innovation, generating investment recommendations. An empirical test with real-world data on 493 listed pharmaceutical companies shows that the GPT-based investment strategy performs excellently in terms of return rates, Sharpe ratios, and maximum drawdowns. Specifically, the cautious investment strategy resulted in a reduction of maximum drawdowns by 9% to 30% and volatility by 5% to 10%. The results indicate that the GPT model significantly improves investment analysis efficiency, accurately identifies high-potential companies, and effectively manages risks. Overall, this study demonstrates the strong potential of human-machine collaboration in addressing the complexity and information density of the pharmaceutical industry, paving the way for intelligent investment analysis.