摘要:We develop a Bayesian framework for estimating beta-pricing models with both traded and nontraded factors, incorporating theory-im
贝塔定价模型的估计与比较
Estimation and Comparison of Beta-Pricing Models
讲座信息
主讲人
冯冠豪教授
香港城市大学
日期和时间
2025年4月10日(周四)
10:30-12:00
地点
综合教学楼D904会议室
论文摘要
We develop a Bayesian framework for estimating beta-pricing models with both traded and nontraded factors, incorporating theory-implied pricing restrictions. The framework facilitates model comparison through marginal likelihoods. Simulations show that models selected via marginal likelihoods outperform those chosen using traditional metrics like cross-sectional R2 and Hansen-Jagannathan distance. Moreover, the selected model achieves superior out-of-sample tangency portfolio performance. Revisiting Kan, Robotti, and Shanken (2013), we demonstrate that the optimal specification is not FF3 but FF3 augmented with nondurable consumption growth. In comparing consumption-based models from Kleibergen and Zhan (2020), we find that all consumption measures weaken significantly when controlling for the market factor. These results highlight the relevance of the model comparison framework developed in this paper.
主讲人简介
冯冠豪教授
香港城市大学
冯冠豪(Gavin Feng)的研究重点是开发方法论解决方案,包括机器学习、贝叶斯统计和金融计量经济学,以应对实证资产定价中的大数据挑战。他的研究成果已发表在《Journal of Finance》、《Journal of Financial Economics》、《Journal of Financial and Quantitative Analysis》、《Journal of Econometrics》和《International Economic Review》等顶级学术期刊上。他是多个外部研究资助项目的首席研究员,例如香港研究资助局的初级研究计划(ECS)和一般研究基金(GRF),以及国家自然科学基金青年科学基金(NSFC)。冯博士的研究受到了业内人士的高度认可,并获得了欧洲国际定量研究协会(INQUIRE Europe)、香港货币与金融研究所(Hong Kong Institute for Monetary and Financial Research)以及AQR洞察奖(AQR Insight Award)等研究奖项的肯定。冯冠豪目前是香港城市大学商业统计学的副教授。他于2017年获得芝加哥大学的博士学位和工商管理硕士学位。
来源:新浪财经