关键词: 异象, 资产定价, 风险因子, 泛函主成分分析

Abstract: This paper studies five pricing anomalies in China's Ashare markets based on functional principal components analysis (FPCA)Compared with the conventional approaches，the FPCA approach can accurately extract the monotonic and nonmonotonic relationship between firm characteristics and expected returnsThe empirical results show that the sorted panel data of returns mainly includes three principal components：the first one represents the expansion of market premium in the economic sense，which is independent of all anomaly effectsThe second one represents the crosssectional “monotonic” effect of anomalies，in which the robustness of value and momentum anomalies is weakThe third one represents the crosssectional “convexity” effect of anomalies，which is related to the “dispersion effect” caused by investors/ behavior.

Key words: anomaly, asset pricing, risk factors, functional principal component analysis