数字技术与经济金融学术午餐会（第28期）

摘要：

We leverage machine learning to introduce novel granular belief dispersion measures to distinguish two key sources of disagreement, i.e., interpretation heterogeneity (IH) and information asymmetry (IA). We find that IH and IA are pivotal components of disagreement, aligning with human decision-based dispersion. While IA pervasively negatively predicts returns, IH exhibits an asymmetric pricing pattern: IH positively predicts returns among underpriced stocks but negatively among overpriced stocks, consistent with the theoretical model of Atmaz(2018) in that disagreement amplifies expectation bias. Our results underscore that the structure of belief dispersion has important heterogeneous effects on market equilibrium. 

主讲人介绍：

刘扬，湖南大学金融与统计学院助理教授，获清华大学金融学博士学位，中央财经大学金融工程学士学位，圣路易斯华盛顿大学奥林商学院访问学者。研究领域为实证资产定价、金融科技和行为金融。其研究成果发表于The Review of Asset Pricing Studies，多篇工作论文入选AFA和CICF等学术会议。作为项目负责人承担国家自然科学基金、教育部人文社科基金各1项。