- 學位論文
財務市場短視損失趨避之行為實驗
Experimental Studies on Behavioral Myopic Loss Aversion in Financial Markets
摘要
中文摘要 在財務市場中存在著幾個備受爭論的迷思，其中之一即為Mehra and Prescott 於1985年所提出的股票溢酬迷思 (equity premium puzzle)。其後，陸續有一些學者試圖對這個迷思提出解說，而短視損失趨避 (myopic loss aversion) 這個行為假說即為Benartzi and Thaler 在1995年對股票溢酬迷思所提出的一種可能解釋。簡單的說，短視損失趨避是指若一個人為損失趨避者，則當其檢視投資報酬的次數愈頻繁，將會降低投資人持有風險資產之意願。而此概念也在許多的文獻中，經由行為經濟實驗獲得證實 (Gneezy and Potters, 1997; Thaler et al., 1997; Gneezy et al., 2003; Haigh and List, 2005; Charness and Gneezy, 2010)。 在過去一系列之行為經濟實驗中，以Gneezy and Potters (1997) 為代表，其分析之重點即在於控制受試者報酬資訊回饋之頻率 (information feedback)，以及受試者調整投資金額之彈性 (investment flexibility)，藉以進一步檢視受試者之投資行為是否如短視損失趨避假說所預測。而本論文則試圖在以較符合投資現況的概念下，加入不同的實驗變因，以探討投資者之風險態度是否更符合短視損失趨避之假說。本論文的第一篇和第三篇文章，即有別於過去之實驗設計，分別允許投資者以投資組合的概念來持有風險資產，以及允許投資者得以考量自身的投資偏好以及過去的投資報酬，來選擇不同的資產配置方式。而本論文之第二篇文章，則從投資者投資心理學著手，來看看當人們在面對連續獲利之後，其風險的態度是否亦符合短視損失趨避之預測。 在短視損失趨避之假說下，投資報酬在經由總合評估後，將相對於個別檢視來說更具吸引力。由於，現實的投資環境中，投資者通常不會只有一種風險投資之機會，因此，在論文之第一篇文章 ”Can myopic loss aversion explain the equity premium puzzle?” 中，本研究引用Gneezy and Potters (1997) 之實驗設計，允許存在多種資產之投資組合，以期透過投資機會的整合，來降低負報酬對於投資人心智盤算之負面影響。而實驗結果顯示，透過投資組合的方式，亦可以達到多期單一投資標的整合而使得得到負報酬機率下降的效果。因此，本文推論MLA可能無法適切地解釋股票溢酬迷思。 此外，在Gneezy et al. (2003) 一文中曾指出，在他們的實驗中仍有一個疑點尚待釐清，稱之為”House money effect”。其認為投資人之風險態度，將受到先前為獲利或損失所影響。也就是說，在先得到正報酬的情況下，相較於先得到負報酬而言，投資者較願意持有風險資產。因此，在論文之第二篇文章 ”Risk taking in the face of large gains: an experimental study on individual behavior” 中，本研究同樣引用Gneezy and Potters (1997) 之實驗設計，來檢驗 House money effect對短視損失趨避現象之影響。本研究將試圖增強house money effect的效果，看看受試者在連續3回合為正報酬以及連續6回合為正報酬的情況下，其投資行為是否因此改變。而實驗結果顯示，人們在面對連續獲得之後，其會較願意承擔風險來持有風險資產之說法，在不同的資訊回饋頻率與投資調整彈性頻率下，並未獲得一致的結論。 本論文的第三篇文章 ” Myopic loss aversion and endogenous choice problem”，則探討影響投資者選擇不同資產配置方式的內在可能因素為何，是否因為時間成本考量而選擇定期定額投資；抑或是因為為掌握更多的投資調整彈性而選擇單筆申購。因此，有別於過去以隨機方式來決定受試者的資產配置方式，本篇允許受試者可以依照自己的投資偏好來進行選擇，即當投資標的之選擇不再是外生決定之下，檢驗看看受試者之風險態度是否符合MLA假說。並且加入申購手續費 (front-end load) 來看看受試者是否願意付出相對的成本，來選擇自己偏好的資產配置方式。而實驗結果顯示，即便在存在申購手續費之下，定期定額投資之平均投資金額仍明顯高於單筆申購，並且，從總平均投資金額來看，其皆在統計上獲得顯著支持。
並列摘要
Abstract One of the anomalies’ puzzles in the financial market has been debated for many years, i.e., the equity premium puzzle dubbed by Mehra and Prescott (1985). Many studies attempt to figure out an explanation for this puzzle in different ways. Among the research, Benartzi and Thaler (1995) offer an explanation of the equity premium using two behavioral concepts; loss aversion and mental accounting. In short, the MLA hypothesis supposes that the investor who has an aversion to loss will be unwilling to hold a risky asset when evaluated frequently. Subsequently, some studies use the method of behavioral and economic experiments to investigate whether the subject’s investment behavior is in line with the myopic loss aversion hypothesis (Gneezy and Potters, 1997; Thaler et al., 1997; Gneezy et al., 2003; Haigh and List, 2005; Charness and Gneezy, 2010). Gneezy and Potters (1997) is one of the representative experimental designs for this issue in these experimental literatures. The focus is on the frequency of feedback information and investment flexibility. This dissertation attempts to examine the myopic loss aversion hypothesis with a similar design to Gneezy and Potters (1997) and with some treatments which are consistent with the investment situation. To mirror a real investment situation, subjects were allowed to invest in multiple risky assets in each period and decide the way they invested. We also considered the effect of prior gains on the individual’s attitude to risks based on the psychology of investing. According to the MLA hypothesis, financial consequences evaluated in an aggregated way will make a risky asset become more attractive. In addition, there is not only one opportunity to invest in risky assets. The first essay in this dissertation examines the MLA hypothesis with multiple choices of risky assets. As a consequence, the experimental results show that the subject will be more willing to take risks if they have multiple choices in risky assets and evaluate them simultaneously. Therefore, the conclusion infers that the MLA hypothesis may not appropriate to explain the equity premium puzzle. Gneezy et al. (2003) pointed out that the house money effect may affect the conjecture of MLA. It infers that prior gains can increase the willingness to take risks. In the second essay, this paper amplifies the house money effect on the subject’s attitude toward taking risks to analyze whether the subjects will increase the amount they invest in the risky asset after three and six consecutive gains. The results did not find consistent consequences across treatments to support the hypothesis that prior consecutive gains contribute to an investor being willing to bear the risks. The third essay investigates the endogenous factor that influences investors to hold the assets in a systematic investment plan or a one-time investment. Considering the time limit or the control flexibility, the investors were allowed to choose the way they invested based on their preference. The other treatment also considered in this essay is the front-end load, to see if a charge would affect the investor’s preference of investment. The results exhibit that even if a front-end load charge exists, the investor is still willing to take risks when they evaluate infrequently, in the overall period, the subject exhibit behavior consistent with MLA hypothesis.
參考文獻
延伸閱讀
- 湯子儀（2009）。最適截斷點之觀點下，以財務比率、公司特性、市場變數與公司治理機制建構財務預警模式〔碩士論文，崑山科技大學〕。華藝線上圖書館。https://doi.org/10.6828/KSU.2009.00014
- 陳松男（1999）。在間斷性避險及交易成本下的選擇權評價模型：以實務觀點修正理論。風險管理學報，1(2)，43-53。https://doi.org/10.30003/JRM.199911.0003
- 湯惠雯、馮淑敏、令狐雅萍（2005）。考慮產業訊息下之財務危機預警模式之研究。管理科學與統計決策，2(1)，77-93。https://doi.org/10.6704/JMSSD.2005.2.1.77
- 林憬鳳（2007）。An Empirical Study of Behavioral Finance on Overreaction of Stock Repurchases〔碩士論文，中原大學〕。華藝線上圖書館。https://doi.org/10.6840/cycu200700669
- Her-Jiun Sheu、Hui-Tzu Lee、Lieh-Ming Luo、Chih-Wei Lee（2023）。Do the Loss Patterns of Financial Assets Influence Investors Perceived Risk? From the Perspective of Loss Aversion。財務金融學刊，31(2)，39-59。https://doi.org/10.6545/JFS.202306_31(2).0002