德勤－复旦会计论坛系列274讲座信息
时间：2026年 6月8日（周一）下午13:30
地点：史带楼503教室
主持人: 黄蓉 教授
报告人: Xiao-Jun Zhang（University of California at Berkeley）(张晓君)
题 目: Measuring Analyst Question Quality in Conference Calls: A Machine Learning Approach
摘 要: We develop a novel machine learning-based measure of analyst question quality derived from the questions analysts ask during earnings calls and the intraday market reactions they elicit. We validate this measure by demonstrating that it predicts information-rich responses by management, larger post-call forecast revisions by analysts, and stronger intraday market reactions by investors. Using this measure, we find that question quality increases when macroeconomic uncertainty rises. This relation is stronger for analysts with greater expertise and institutional resources, weaker for overloaded analysts, and amplified when firms reduce voluntary disclosure. Importantly, the forecast accuracy gains from higher-quality questions nearly double when moving from low- to high-uncertainty environments. Our findings suggest that information acquisition in capital markets is dynamic: analysts function as a stabilizing informational force when uncertainty rises.
-
活动讲座
-
新闻动态
-
微信头条
-
招生咨询
-
媒体视角
-
瞰见云课堂
