Graph LLM-Based Portfolio Management Algorithm

This paper explores the integration of large language models (LLMs) with graph-based financial networks for quantitative trading. By leveraging GPT-3 for stock network return predictions, we develop a Graph-LLM trading strategy. Experimental results demonstrate that the proposed strategy achieves lower volatility and more stable performance than traditional baselines. Our findings highlight the potential of combining LLMs with financial complex networks to enhance quantitative trading strategies.