文獻上對於意見分歧在資產訂價上所扮演的角色有著不同的見解。傳統財務學認為意見 分歧的程度代表市場投資人所面臨的“資訊風險"(information risk)或者“估計風 險"(estimation risk)。然而行為財務學則認為由於市場上存在賣空限制,因此意見 分歧愈大市場上的錯價(mispricing)程度也相對愈高。本文嘗試投過投資人情緒的角色 來重新探討這個議題: 當雜訊交易者情緒愈樂觀時,意見分歧愈大代表市場錯價程度也 愈大;然而當雜訊交易者對於市場感到悲觀時,意見分歧愈大則相對反映市場上有著愈 大的“風險"。透過投資組合的分析方式,我們的發現當雜訊交易者的情緒愈樂觀時, 他們會愈積極的進入市場做長部位投資,進而過度推高資產價格。而賣空限制的存在則 會減緩樂觀型錯價的修正,因此資產未來的報酬也會愈低。然而當雜訊交易者的情緒相 對悲觀時,他們並不會積極的進入市場做短部位投資;此時市場上的悲觀型錯價幅度相 對較低,資產價格也更有效率,未來的報酬會愈高。此外,我們也發現當市場上整體情 緒樂觀時,意見分歧並不能提供顯著的共變異效果(covariance effect);然而當情緒 愈悲觀,意見分歧則能反映出顯著的共變異效果。總合來說,本文的結果認為文獻上關 於意見紛歧的觀點所存在的“分歧",並不全然是互斥的觀點,相反的,可以透過投資 人情緒的角色來重新解釋。;Whether differences of opinion cause mispricing or represent risk has long been a controversial issue. We find that the puzzle associated with differences of opinion can be reconciled with investor sentiment. Differences of opinion are accompanied with negative returns during periods of high sentiment, but positive returns during periods of low sentiment. A common factor extracted from variance divergent opinion measures is only priced following pessimistic periods, supporting the notion the positive risk-return relation implied by differences of opinion is rationally driven. Finally, a time-varying negative relation between sentiment and portfolios sorted on divergent opinion further supports our results.