一、报告题目：

Beta Reversal and Expected Returns

二、报告人：

Prof.XU Yexiao, the University of Texas at Dallas

三、报告时间：

2018年07月03日（周二）下午2:30-5:00

四、报告地点：

知新楼B423

五、报告人简介：

Professor Yexiao Xu received his Ph.D in financial economics from Princeton University in 1996. Currently he is an associate professor in the School of Management, the University of Texas at Dallas. In 2001, he won the Smith-Breeden Prize, one of the most prestigious awards in Finance, for his study on idiosyncratic risks. His published and working papers have generated over 5,000 citations.Professor Xu's research specialty covers stock market volatility, the pricing role of idiosyncratic risk, factor models, predictability, mutual fund performance, analyst research, tax and closed-end fund discounts, and adaptive estimators. Currently he is working on a number of topics including, asset pricing test, implied cost of capital,

predictability of idiosyncratic risk, partial factor structure, econometric models for leverage, short sale interest, and many related issues in the Chinese and the Japanese equity markets. Professor Xu has taught Ph.D level theoretical and empirical asset pricing courses, financial econometrics, as well as MBA financial management and investment courses.

六、 报告摘要：

In this paper we show that it is the beta reversal among a small group of stocks that prevents the CAPM beta from predicting individual stocks’ expected returns as doc-umented by Fama and French (1992). These stocks tend to have both large beta and high idiosyncratic volatility. Consequently, even when the CAPM holds period-by-period, the confounding effect of beta reversal diminishes the significance of the CAPM beta in the cross-sectional tests. The cross-sectional explanatory power of beta is restored after we take into account the beta reversal effect in each of the three ways. More important, the market risk premium estimated from cross-sectional regression analysis is almost identical to the historical average of market excess returns. All results are robust with respect to different measures of beta and idiosyncratic volatility as well as different subsamples. We also find that beta reversal is likely to be a result of several factors including the wealth effect, earnings announcement effect, and real option realization.

七、主办单位：

必赢线路检测中心