本研究利用股票過去日報酬排名或股票日報酬出現正負符號次數的資訊, 提出無無母數動能策略(也就是買入過去報酬平均排名(或報酬出現正符號的次數)高的股 票並同時放空報酬平均排名(或報酬出現負符號的次數) 低的股票), 並探討其獲利性與其背後之成因。研究結果顯示, 此類策略在未來一年有持續而顯著的獲利性, 也無法被現有的資產定價模型所解釋。無母數動能策略的獲利性高於Jegadeesh andTitman (1993) 所提出的價格動能策略以及George and Hwang (2004) 所提出的52 週高價動能策略。本研究認為, 無母數統計量捕捉股票價格中容易被投資人所忽略的「非凸顯」訊息。進一步探究顯示, 無母數動能策略的獲利性的確會隨著樣本股票具有較凸顯特徵與較低套利風險而有較弱的持續性; 而投資人情緒的變動對無母數動能策略獲利性的預測亦有顯著的影響。;Statistic measures such as rank and sign of daily returns reflect the non-salient information embedded in stock prices ignored by investors. Momentum strategies formed by buying stocks with high average ranks (or signs) and shorting those with low average ranks (or signs) thus are pofitable over a year following formation. The profits annot be explained by well-known asset-pricing models, and re stronger than the price momentum proposed by Jegadeesh and Titman (1993) and the 52-week high momentum proposed by George and Hwang (2004). Further analysis reversals that the rank and sign momentum profits are weaker among stocks with higher salient features, are stronger among stocks that are subject to higher arbitrage risk, and exhibit patterns related to fluctuation in investor sentiment.
