意見分歧、凸顯性與資產定價異常現象;Disagreement, Salience, and Asset-Pricing Anomalies

NCUIR > School of Management at National Central University > Department of Finance > Research Project > Item 987654321/84674

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/84674

Title:	意見分歧、凸顯性與資產定價異常現象;Disagreement, Salience, and Asset-Pricing Anomalies
Authors:	周賓凰
Contributors:	財務金融學系
Keywords:	意見分歧;賣空限制;樂透關連異常現象;凸顯性;投資人異質性;賣空餘額;;disagreement;short-sale restriction;salience;lottery-related anomalies;investor heterogeneity;short interest.
Date:	2020-12-08
Issue Date:	2020-12-09 10:40:30 (UTC+8)
Publisher:	科技部
Abstract:	「意見分歧」是財務理論很重要的課題, 因為意見分歧可以解釋許多傳統定價理論所無法解釋的現象, 而其中最受重視的變數之一, 就是與交易量有關的種種報種型態與特性(Lo and Wang, 2000)。 2019 年底 Google Scholars 關鍵字搜尋的結果約計三百萬項, 而其中最有名的 Miller (1977) 論文被將近 4000 篇論文所引用, 顯示這個主題所受到的重視。 Miller 一文指出意見紛歧與賣空限制的存在容易導致資產價格被高估, 進而導致後續的較低報酬。後續這理論被擴充解釋諸多現象, 包括價格動能、市場崩盤、泡沫等(見 Hong and Stein, 2007)。然而正如 Hong and Stein (2007) 所言, Miller 的模型為一靜態模型, 因此無法明確說明交易量。相對於其他模型, 該模型乃奠基於部分分析, 非均衡模型, 優點是無須仰賴明確的偏好等假設, 而缺點則是對許多現象(如資產定價異常現象) 無法有較明確的預測與解釋。由於 Miller 的原始模型是敘述性的, 因此很多變數間的關係不容易說明。本三年期計畫的目標有三。1. 將 Miller 的模型數量化, 建構一嚴謹的理論模型。據此模型, 很多變數間的關係(例如賣空限制、賣空餘額與其他影響意見分歧的因素) 將變得比較清楚。2. 我們結合 Miller (1977) 的模型與 Bordalo, Gennaioli, and Shleifer (2012, 2013)的「凸顯性」 (salience) 理論, 證明投資人對具凸顯性的報酬或績效的不同反應, 導致主觀機率上的扭曲評估; 在存在賣空限制下, 對不同凸顯性的報酬或績效資產會有不同程度的過度反應或反應不足。預計此一理論模型可解釋著名的「中短期動能與長期反轉」現象, 以及其他像是 maxing-out effect 等有樂透性質的異常現象。3. 提出一結合投資人異質性與意見紛歧的理論, 解釋上述現象與交易量的關連。初步研究顯示這些樂透關連異常現象(像是 Max effect 等) 的確與交易量呈現顯著的交互關係。我們預計以美國資料為主, 並進而擴充到國際股市的實證分析。 ;Disagreement, or divergence of opinion, plays an important role in finance, because it helps explain various puzzling phenomena unexplained by traditional asset-pricing models, especially those related to trading volume (Lo and Wang, 2000). As of the end of 2019, a search of Google Scholars using the words of “differences of opinion” or “divergence of opinion” yields a total number of 3,000,000 items, among which the seminal paper of Miller (1977) alone earns almost 4000 citations. As Barberis (2018) points out, Miller’s paper was neglected for almost two decades until the late 1990s, because his “imperfect” view contradicted the then mainstream efficient- market view. Miller (1977) shows that differences of opinion, in conjunction with the presence of short-sale restrictions, cause overpricing, which in turn is followed by a lower subsequent return. Miller’s theory has been applied to explain return patterns related to price momentum, market crashes, and even bubbles (see, e.g., Hong and Stein, 2007). However, as it is a static one, it cannot directly speak of trading volume, because it is the “crossing” of opinions among investors that leads to trade, rather then the divergence of opinion itself.Miller’s model has the strength of being simple, because it is based on a partial analysis, rather than an equilibrium analysis, so that it does not rely on strict assump- tions on preferences or behavior. The simplicity has its weakness, because then it is unlikely to provide more specific predictions or explanations for several documented phenomena or anomalies. Hence, the purpose of this 3-year project is threefold:1. As Miller’s original model is descriptive in nature, we provide a theoretical model, so that the relationship among many variables of interest (e.g., short- sale restrictions, short-interest, and disagreement) becomes clear. Based upon the model, some testable hypotheses can be developed and tested.2. We unify MIller’s model and the salience model proposed by Bordalo, Gennaioli,and Shleifer (2012, 2013), so that investors’ misreactions to varying degrees of payoffs or returns generate disagreement, which in turn causing overpricing and underpricing to different assets. We expect the new proposed theory can explain the famous “short-term momentum and long-term reversal” puzzle. The theory also helps explain several lottery-like anomalies such as the maxing-out effect documented by Bali, Cakici, and Whitelaw (2011).3. By augmenting Miller’s (1977) differences-of-opinion theory with investor het- erogeneity, we show that mispricing due to noise traders’ overreaction to salient news will exhibit patterns related to turnover. Our theory is related to two strands of literature: Miller’s disagreement model and the lottery-like features or preferences as proposed by Kumar (2009). The contribution of this project is to provide theoretical and empirical support for turnover-related patterns on lottery-related anomalies. We propose several testable hypotheses to examine whether lottery-related anomalies really exhibit patterns related to turnover. The empirical analysis will be mostly based on the U.S. data, and hopefully extended to international data.
Relation:	財團法人國家實驗研究院科技政策研究與資訊中心
Appears in Collections:	[Department of Finance] Research Project

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	180	View/Open

社群 sharing

Loading...