More Powerful Tests for Anomalies in the China A-Share Market
针对中国A股市场数据时间短导致传统排序方法统计功效不足的问题，本文采用结合公司特征与协方差矩阵的高效排序方法，使t统计量翻倍，在2008-2020年间识别出九个显著异常现象，包括规模、价值、低风险等。
Research into asset pricing anomalies in the China A-share market is hampered given the short time series of available returns. Even when average excess returns on candidate factor portfolios are economically sizeable, conventional portfolio sorting methods lack statistical power. The authors apply an efficient sorting procedure that combines firm characteristics with the covariance matrix. For the China A-share market, they find that the efficient sorting procedure doubles the <italic>t</italic>-statistics compared to conventional portfolio sorts, leading to nine instead of three significant anomalies over the post-reform period from 2008 to 2020. They find significant size, value, low-risk, and returns-based anomalies. Although portfolio characteristics differ between sorting methods, the authors find that efficient sorting portfolios highly correlate with equally weighted portfolios and capture the same underlying anomaly.