Please read Dr. Yu’s article in the Global Finance Journal titled, “Attention based dynamic graph neural network for asset pricing.“
Firms do not function in isolation in the marketplace. They are linked to each other through various channels, such as their supply chain networks, board of directors, similar fundamentals and industry sectors, as well as market conditions. These channels allow continuous information exchanges among firms in the financial markets. As a result, the performance of a firm depends not only on its own operation but also on other relevant firms, i.e., the seamless information flow and interconnection among firms affect each other and their market prices. This interconnectedness among firms through contemporaneous links forms the firms’ network. Previous studies suggest that the structure, properties, and dynamism of a network provide important insights into how information flows and shocks propagate across firms and thereby affect their stock prices. To read the full article.
Attention based dynamic graph neural network for asset pricing. Uddin A, Tao X, Yu D. Glob Financ J. 2023 Nov;58:100900. PMID: 37908899 PMCID: PMC10614642 DOI: 1016/j.gfj.2023.100900