演讲题目：A Simple Noise Reduction Method and Its Applications to Asset Pricing Puzzles
报告人：阮军 助理教授
时 间：2019年4月4日（周四） 下午15：00-16：30
地 点：嘉庚一 507
主持人：刘杨树 副教授
参加者：对会计、财务研究有兴趣的师生
摘要：In this research talk, I will briefly review some fundamental issues in asset pricing based on my research over the past years. I argue that many of the asset pricing puzzles can be attributed to the measurement noise in both expected returns and risk in the tests or uses of the asset pricing models. I introduce a simple noise reduction method in a regression context that utilizes the high correlations between the noise components in two noisy proxies for the same fundamental unobserved variable to attenuate the effects of measurement noise on the test of the relation between unobserved variables under consideration. I discuss three important applications of the method in asset pricing. In the first application, I recover a strong time-series relation between market return and market risk, the absence of which has been a puzzle for at least 30 years. In the second application, I resolve an empirical debate over the last 15 years about whether idiosyncratic risk matters by demonstrating a robust time-series relation between market return and idiosyncratic risk. In the third application, I resurrect the pricing role of the CAPM beta, the absence of which is arguably the greatest puzzle in finance over the last 50 years.
