News

Seminar (Dept. of Finance)

Copyfrom:Finance Time:2024-05-24

Title: Approaching the mean-variance efficiency in a high dimension: Which firm characteristics matter?

Speaker: Ti Zhou, Assistant Professor of Finance, School of Business, Southern University of Science and Technology

Time: 10:00-11:30, May 24, 2024(Friday)

Venue: Room 1008, Mingde Business Building

Language: English/Chinese


ABSTRACT:

Numerous firm characteristics can predict the cross-section of stock returns. In this paper, we investigate which characteristics provide independent information to achieve mean-variance efficiency in the presence of other competing characteristics. We develop a novel regularized projection portfolio approach to tackle this high-dimensional problem. First, we show that the efficient portfolio weights can be represented as the coefficients of a linear regression of a known constant on the excess returns of base assets. We then propose a regularized regression method that can produce portfolio weights even when the dimension of the base assets far exceeds the sample size. We show theoretically that the expected return and variance of the estimated portfolio converge to those of the true efficient portfolio at the rate of the square root of log(N)/T. Applying our approach to 187 characteristic-managed U.S. equity portfolios in an out-of-sample analysis, we find that the resulting portfolio weights are sparse and well behaved. The portfolio strategy outperforms a variety of benchmark strategies, delivers an annualized Sharpe ratio of 3.32 without transaction costs and 1.21 with transaction costs, and generates significant risk-adjusted returns across a variety of factor models. Over 80% of the 187 characteristics are redundant. Announcement return appears to be the most important characteristic, and characteristics related to momentum, value and growth, and trading frictions play a dominant role in forming the efficient portfolios.

SHORT BIOGRAPHY:

Ti Zhou is an Assistant Professor of Finance in School of Business, Southern University of Science and Technology. He obtained his Ph.D.in Finance from the Business School in the Hong Kong University of Science and Technology. His research interests center on Asset pricing, Option-implied information, Portfolio choice, Big data and machine learning in finance. He has published papers on Journal of Financial and Quantitative Analysis, Journal of Empirical Finance and etc.

官方微信

RMBS made the Top-50 list of MBA,
EMBA and EE programs——The Financial Times

Renmin University of China

@Business School, Renmin University of China 京ICP备05066828号-1