橫斷面期望報酬異象分析及投資組合管理

Abstract

由於越來越多文獻在探討股票市場橫斷面(cross-section)的期望報酬異象分析，根據Harvey et al. (2016) 提出1967年至今已有數百篇相關文獻探討相關議題，並提出數以百計個可以解釋CAPM 無法解釋的異常報酬的變數，本篇研究試圖從過去文獻所提出對橫斷面的資產期望報酬具有解釋力的變數，來解釋美國股票市場的報酬情形，主要根據Mclean and Pontiff (2016) 所整理的97個因子為依據，包含基本面、評價面、市場資訊及公司事件等各面向變數，使用Fama-MacBeth 迴歸估計並預測報酬的橫斷面性質，使用移動平均10年作樣本內估計，並預測下期股票期望報酬。模型回溯測試每月平均月報酬為3.24%，標準差為5.71%。

Since more and more literatures are exploring cross-section analysis of expected returns in the stock market, there have been hundreds of relevant literatures since 1967 to date, based on Harvey et al. (2016). Hundreds of variables that account for unexplained abnormal returns in CAPM are explored. The present study attempts to explain the payoffs in the US stock market from the explanatory variables that have been proposed in the past to the expected return on assets in the cross-section, Using the Fama-MacBeth regression to estimate and forecast the cross-sectional returns using the 97 factors that are included in Mclean and Pontiff (2016), including the variables of fundamentals, valuation, market information, and corporate actions, 10 years moving average for the sample estimates, and forecast the next stock expected reward. The average monthly compensation for the regression test was 3.24% and the standard deviation was 5.71%.