本文擬由馬可夫模型透過狀態機率預測與下期價格預測，描述投資人對未來價格變動的移動路徑，將投資人在資訊充分下的決策情境納入處分效果考量，期對投資人心理認知過程提出更深入的觀察。首先，本文擬檢視修正後處分效果，是否較傳統處分效果對於反應不足現象提出更好的解釋；並驗證其在多、空不同結構下對於市場價格的影響。再者，基於波動行為對於投資人資產配置與風險管理的重要，本文擬檢視處分效果對市場波動的影響，以增加波動內涵的瞭解。其次，處分效果與成交量異常變化為近來熱門議題，然由於過度自信亦可能導致資人過度交易現象，本文擬區別並比較兩種不同型態心理認知對於投資人交易行為的影響。綜合上述議題，本文針對亞洲四小龍及G7 國家之股市與期貨市場作探討，除了廣泛比較股市與期貨市場投資人交易行為差異，並分析已開發國家與開發中國家投資人認知行為的異同。本研究在模型的設計上，針對成交量、價格變動及處分效果之一階及二階動差交互相關聯，提出ADCC(Asymmetric Dynamic Conditional Correlation) MGARCH-MSKST(Multivariate Skewed Student-t distribution) 模型，同時將三者內生性因素納入模型考量，以妥適描述三者間動態相關，對於投資人行為提供更深入的觀察。 The disposition effect is one of the most striking features of individual investor trading. We present the evidences about the mechanism that links trading volume, market returns and investors’ disposition effect simultaneously by proposing an innovative ADCC (Asymmetric Dynamic Conditional Correlation) MGARCH-MSKST (Multivariate Skewed Student-t distribution) model. Despite the abundance of studies on the features of disposition effect in the extent literatures, rarely has this issue been investigated based on investors’ rational expectation behavior. This paper fills the gap by employing the Markov Switching model to provide a more reasonable approach in describing investors’ forward-looking decision making process, which allows us to better capture whether investors exhibit disposition effect about their valuation. Furthermore, we also examine the dynamics for the effect of disposition effect on market volatility and short-term return. In addition, we attempt to distinguish either overconfidence bias or the disposition effect is the drive to explain the excess volume puzzle. Moreover, we also highlight whether the markets has the tendency to be under-reacted by the presence of disposition effects.
