論文摘要內容： 我們利用大規模全球個股的資料，研究來自長短期之不同形成期及持有期反向策略之績效。與之前的研究發現不同，我們發現全球市場的反向策略在各種期間中都有實質的獲利。 經過 「略過一個月」、「排除極端值個股」後，反向策略仍然是有效的。 同時在不同市場情境下，全球市場的反向策略仍具有一致的獲利性。 此外，我們發現反向策略的報酬主要來自於過度反應的效果。 最後，我們發現不論是Fama and French (FF, 1993)的三因子模型或是 Connor and Korajczyk (CK, 1986)的主成分分析因子模型皆無法解釋反向策略的報酬。 Abstract Using global data, this study investigates the performance of contrarian strategies for various ranking and holding horizons ranging from one month to three years. In contrast to previous findings, we find the contrarian strategies are substantially profitable in the global market on all horizons. After conducting the robustness checks such as skipping one month and trimming extreme observations, the market still exhibits contrarian. While there is evidence that the contrarian strategies perform consistently well no matter the aggregate market is bull or bear during the ranking or the holding period. In addition, the contrarian profits are mainly attributable to the overreaction effect. Finally, we find that neither the Fama-French (1993) three-factor model nor the Connor-Korajczyk (1986) five-factor model explains the contrarian profits.
