|摘要: ||本研究將深入分析獨特性風險與預期報酬的關係。此議題過去已有大量的文獻認 同，但兩者間的關係近期卻變得難以解釋。獨特性風險的困境(Idiosyncratic volatility puzzle) 近期由Ang, Hodrick, Xing and Zhang (2006, 2009)提出，他們指出獨特性風險與 預期報酬成顯著負向關係，推翻過去呈現正向關係的理論，隨即也引發許多相關研究。 一部份的研究重點在於獨特性風險的衡量方式，另一研究方向即是希望找到此負向關 係的解釋原因，獨特性偏態是近期被關注的研究重點。投資人可能接受股票具高獨特 性波動卻是低預期報酬的原因，除了某些投資人尋求高波動，樂透型的報酬也是吸引 投資人投資的因素之一；而在控制偏態的特性後，獨特性波動與預期報酬間的負向關 係變得不顯著。由此觀點加以延伸，本文將估算獨特性風險的高階動差，偏態與峰態 的特性應顯著影響預期報酬。再者，本文也使用動態模型估計獨特性風險，在考量偏 態與峰態的條件下，將消弭獨特性風險與預期報酬間的負向顯著。本文進一步將研究 樣本擴展至效率性較低之新興市場，樣本資料的深度與廣度都將提升以詳細分析獨特 性風險的困境。以上結果對於市場參與人將有相當的實質助益。|
The relationship between the idiosyncratic risk and expected return is investigating in this research with detailed analyzing. The idiosyncratic volatility related to the expected return are proved by considerable quantities literatures, however, what kind of relationship becomes a puzzle recently. The ‘idiosyncratic volatility puzzle’ proposed by Ang, Hodrick, Xing and Zhang (2006, 2009) generates substantial interest among researches. One way of the researches are focus on measurement of idiosyncratic volatility, another focus is try to explain the negative relationship in various ways. The idiosyncratic skewness is discussed lately. Investors may accept lower average returns on stocks with high idiosyncratic volatility, not because they seek higher volatility, but because they have a preference for stocks with lottery-like return properties. The negative relationship between the idiosyncratic volatility and expected return is much smaller and insignificant after controlling for expected idiosyncratic skewness. Based on this viewpoint, we measure the idiosyncratic risk with higher moments. Stocks with high skewness and kurtosis would be expected to have great reward, these characteristics should have significantly related to expected return. Then, as considering the time-varying properties in estimating idiosyncratic risk, the negative relationship may not significant while controlling the skewness and kurtosis characteristics. Further, the discussion is extending to emerging markets with relative weak efficiency than developed markets. Finally, the sample is extent by both the width and length of sample for deeply analyzing the idiosyncratic puzzle. These findings are truly valuable to market participations.