本研究主要探討投資人的積極程度以及市場的樂觀程度是否會影響動能策略之獲利能力。本研究利用1973至2013年間美國個股進行實證研究，結果驗證了動能策略於樣本期間能有顯著的獲利。進一步的實證結果顯示，規模較小且交易量成長率較低的公司存在極短期(一個月內)反轉的現象。此外，在市場樂觀期間(較多的首次公開發行的公司家數、較高的消費者信心指數或較低的恐慌指數)動能策略之獲利能力較佳且顯著。因此，我們建議投資人能在市場樂觀期間對規模較小的公司進行動能策略，將可得到較高的預期超額報酬。 / The main purpose of this study is to investigate whether the activism of investors and the sentiment of the market can affect the profitability of the momentum strategy. Using individual firms during 1973 to 2013 as the sample, this study reexamines and confirms the profitability of the momentum strategy. The further empirical result shows that firms with smaller size and lower growth rate of trading volume exhibit a very short-term (within one month) reversal effect. In addition, during the optimistic period (years which have more firms conducting initial public offerings, higher consumer confidence index, or lower VIX), the profitability of the momentum strategy is significantly higher than that during the passive period. Therefore, a suggested trading strategy applying momentum strategy to small firms during the high sentiment period may yield a superior performance.

本篇論文藉由行為財務學為橋樑，試圖利用法律經濟分析來檢視證券交易法的制定。結合了行為學、法律以及財務。首先利用認知心理學來介紹投資人行為偏誤的現象。整理出三項投資人處理資訊的偏誤，分別為有限注意、過度自信以及框架效果。在資本市場上投資人因為這些偏誤產生的無效率現象分別有，會計資訊解讀不完整、股價報酬產生動能現象、投資人過度交易、對公司事件宣告產生反應不足的現象、忽略資產配置的重要、處分效果以及偏好現金股利等等的現象。 而在了解行為模型對於投資人的行為預測以後，我們試著去進行法律經濟分析。行為財務學點出了效率市場假說不存在的證據，並試著以心理學的角度來解釋這些現象，但是行為財務學仍然缺乏一個良好的預測模型，它無法預測投資人何時會有反應不足，何時會有過度反應。不過行為財務學仍然提供了一些觀點值得政策制定者考慮。1.投資人偏好是不一致的，可能會因為情境改變而改變。2. 投資人只有有限的理性。3. 投資人會受到資訊揭露形式而有不同的判斷。根據這些觀點，我們以行為學的角度來分析證券交易法對於詐欺、操縱以及資訊揭露的規定。

傳統財務理論認為市場具有效率，在投資者具有理性且追求最大效用的假設下，股價應能迅速且完全的反應所有資訊，但近年來許多學者研究發現一些違反傳統定價理論和效率市場的實證結果。為解釋上述傳統定價理論無法解釋的異常現象，以心理學對投資人決策過程的研究成果為基礎，重新檢視整體市場價格的行為財務學便獲得重視。 本文使用半參數平滑係數分量廻歸模型，利用1988至2007最近20年的月資料，檢視G7各國在不同大盤表現分量上，不同產業股價超額報酬率，是否造成總體經濟指標對大盤未來超額報酬率的邊際影響有所不同？藉以了解各國在不同股市報酬率分量上的資訊傳遞速度與彼此間的差異。此外，利用半參數平滑係數分量廻歸模型，亦可觀察產業超額報酬率如何直接影響未來大盤超額報酬率，不但較傳統最小平方法（ordinary least squares, OLS）更富有彈性，也能觀察在不同分量上的變化情形。 本文發現美國各產業超額報酬率，對未來大盤超額報酬率的直接或間接影響，在不同大盤表現分量上呈現很大差異，未來大盤超額報酬率皆明顯隨著產業超額報酬率的改變而變動；至於其他六國，亦有相似情況，顯示投資人無法即時解讀產業資訊對未來總體經濟的影響，導致產業資訊於產業與大盤間緩慢擴散。 / In this paper, we use semiparametric smooth coefficient quantile regression model to analyze the information diffusion between industries and stock markets. Under different quantile of stock market performances, we examine whether the returns of industry portfolios cause macroeconomic indicator to affect the future stock market performance marginally using data on monthly returns to G7 industry portfolios for the years between 1988 and 2007. We find that the returns of industry portfolios in USA affect the future stock market performance directly or indirectly which display much variously. Moreover, the other counties of G7 also have the same situation. Hence, these findings indicate that investors are unable to understand the influence of industry shocks on macroeconomic activities and information diffuses across investors in different markets only gradually.

本文旨在研究我國銀行長期投資行為和短期投資行為是否會因其過去投資績效所影響，而有賭資效應、過度自信及心理帳戶假說所預期不理性之現象。另外，更進一步探究此三種心理偏誤對投資行為之影響，是否依銀行業務性質的不同而有所差異，因此，也將討論商業銀行和工業銀行之長、短期投資行為。使用1994至2007年的追蹤資料進行迴歸分析，其結果發現茲臚列於下。 1.在控制資金來源(負債)、資產獲利能力(資產報酬率和存放款利差)、股市榮枯以及過去持有的投資存量這些會影響銀行投資行為的因素後，發現前一季長期和短期投資績效愈好，當季銀行長期和短期投資行為就會承擔過多的風險而有賭資效應之現象；在投資績效分為利得和損失的設定下，銀行長期投資行為有賭資效應之傾向，而短期投資行為則否。再者，當銀行面對過去兩季長期投資績效皆為利得時，當季的長期投資風險態度更顯積極，因而銀行長期投資行為有過度自信的現象，而短期投資行為則否。最後，銀行在進行短期投資決策時，無法將前一季長期投資績效納入考量，表示銀行將短期投資視為一獨立帳戶，忽略長期投資的情況，不能綜觀全局，因此短期投資行為有心理帳戶所述現象，而長期投資行為則否。 2.為了釐清賭資效應、過度自信與心理帳戶這些心理偏誤對投資行為的影響，是否依銀行業務性質的不同而有所差異。因此，就銀行類別而言，本文發現商業銀行長、短期投資行為皆有賭資效應與心理帳戶偏誤；工業銀行長期投資行為有賭資效應及過度自信偏誤，短期行為則無任何心理偏誤。

本研究旨在探討投資人情緒與月營收資訊是否能夠解釋台灣股票報酬，以及投資人情緒是否影響對於月營收資訊的解讀。實證結果發現，前期月營收資訊、前期投資人情緒以及兩者乘積對當期股票報酬有顯著的正向影響。表示月營收資訊的揭露能帶給投資人資訊，作為交易時的考量，而前期投資人情緒能夠部分解釋當期股票報酬。兩者乘積表示營收資訊和投資人情緒為同期時，投資人對於資訊的解讀會受到情緒影響，進而影響股票的報酬。進一步探討，當投資人面對月營收成長率高、月營收波動低時，對於其月營收資訊的解讀傾向伴隨著情緒。 / This study wants to discuss whether sales information and investor sentiment could explain Taiwan stock market and whether investor sentiment affects their interpretation on sales information. The empirical results show that the sales information last period, investor sentiment last period and the multiplication of these two are significantly positively correlated to current stock returns. It indicates that the sales disclosure is informative to investors as a consideration while trading and investor sentiment can partially explain the stock returns. The estimate of multiplication which is positive indicates that when sales information and investor sentiment are in the same period, investors’ interpretation on sales information would be affected by their sentiment, and further affecting the stock returns. We further find that when companies have the characteristics of high sales growth rate and low sales volatility, their sales information tends to be interpreted by investors emotionally.

本研究最主要的貢獻為建構一具台灣股市投資人情緒指數並檢測投資人情緒指標與台灣股市的關係。本研究以台灣股票市場為背景，研究期間為2001年1月至2010年12月。利用Baker, Wurgler and Yuan在2009年提出的方法以Volatility Premium, Number of IPOs, First Day Return of IPOs, Turnover Rate四個變數編製台灣股市投資人情緒指數，並探討台灣股市投資人情緒指數變動量與台股大盤報酬之間的領先落後關係。 實證結果發現，在較短的時間，如月資料，台股大盤報酬會影響下一期的台灣股市投資人情緒指數變動量，而在較長的時間，如季資料，台灣股市投資人情緒指數變動量會影響四期後的台股大盤報酬，即短期台灣股市投資人情緒指數變動量為大盤報酬之落後指標，長期則為大盤報酬之領先指標，短期原因為投資人情緒指數受大盤報酬影響，而易有追高殺低現象，長期雖投資人情緒領先大盤報酬的結果在統計上顯著，但經濟上並無顯著意義，另一方面，台股大盤報酬與台灣股市投資人情緒指數變動量間存在正相關，即不能以台灣股市投資人情緒指數變動量預測股市泡沫。 / The main contribution would be the construction of the sentiment index in Taiwan stock markets and examining the relationship between the variation of the sentiment index and Taiwan stock market returns. The background is Taiwan stock markets. The sample period is from January 2001 to December 2010. We use the method in Baker, Wurgler and Yuan (2009) to measure investors’ sentiment and explore the relationship between the variation of the sentiment index and Taiwan stock market returns. The empirical result reveals that in monthly data, Taiwan stock market returns is the leading indicator of the variation of investment sentiment. In a longer term, we mean the quarterly data in this paper, the situation changes. In quarterly data, the variation of the investment sentiment is the leading indicator of the Taiwan stock market returns. In addition, instead of a negative correlation between the stock market returns and our sentiment index, we prove that our sentiment index have a positive impact on stock market returns. Therefore, we could not use this sentiment index to forecast future economic bubbles.

本篇論文藉由文獻探討與迴歸分析探討法人與散戶的投資偏好與行為的不同而造成的報酬上的差異，本文共收集了54篇探討法人與散戶投資行為和選股偏好的文獻並進行整理，其54篇文獻細分成三個方向探討投資人的投資行為：交易策略、認知與情緒偏誤和訊息內含。由文獻的整理中發現，法人與散戶的投資報酬差異確實與交易策略和選股偏好相關，相較於散戶，法人具有資訊與資源上的優勢，所以他們較能夠在面臨投資和選股決策時，做出正確的決定。然而，散戶在資訊的取得上相較法人處於劣勢，所以在做決策時較為不明確，並且由統計數據來看，散戶的部位通常與法人為相反的，所以散戶通常為法人提供流動性，並且因此得到較差的報酬。 而行為財務學的角度來看，法人和散戶皆有不理性的投資行為，而這些不理性的行為皆會為他們招致較低的投資報酬，而法人有較高的投資報酬率並非因為他們理性，而是因為相較於散戶，法人的不理性程度較為低的原因。既然由行為財務學的觀點來看大家皆是不理性的，便推論法人與散戶的報酬差異是來自選股的偏好，在第四部分以台灣經濟新報658家上市公司的資料進行迴歸分析以探討台灣法人、散戶與外資的選股偏好，結果顯示法人和散戶偏好依據其週轉率、公司規模、本益比、股價波動度與股利殖利率有所不同，而外資法人的偏好則與國內法人類似。 / This paper surveys the literatures relating to the investment preference and performance between institutional and individual investors in order to determine the reason of return disparity. 54 papers are surveyed to understand the preference and performance through three aspects: trading strategies, cognitive and emotional biases, and information content. Disparity of returns is due to trading behaviors and stock preferences. Institutional investors tend to be informed and make correct decision when trading. However, individual investors tend to invest in the opposite direction to institutions and provide liquidity for them. Therefore, institutional investor have better performance due to their less irrational behaviors and correct selection of underlying objects. In the fourth part, using data of 658 listed equities from Taiwan Economic Journal from Taiwan’s stock market, we investigate the relation between investors’ ownership and financial indicators. The regression analysis shows that the stock preferences between individual and institutional investor are different. Results indicate that institutional and individual investors have distinct preferences based on turnover rate, size, price to earnings per share ratio, stock volatility, and dividend yield. Foreign institutions’ stock selection preference is similar to domestic institutions.

為了針對實證股價與交易量的資訊，能否作為預測未來政治事件結果提供依據，此篇論文探究四個政治事件，分別是2014年台北市長選舉國民黨黨內初選、2014年台北市長選舉、2016年總統選舉國民黨更換候選人以及2016年總統選舉。作者使用異常交易量作為判斷投資人是否將政治事件視為投資機會，並觀察個股與投資組合之累積異常報酬率是否能作為預測依據，最後以事件後五日內的異常累積報酬判斷投資人是否在事前過度反應或是反應不足。 　　本研究發現，在政治事件發生前六十日內，大部分政治相關的個股交易量皆顯著異於前一年之交易量。與勝選相關的股票在事件前60日至31日有顯著的正累積異常報酬。相較之下，與敗選相關的股票中，僅有和候選人有緊密政治連結的股票有顯著負累積異常報酬，政策相關的股票並無顯著負累計異常報酬。最後，在現行交易制度下，正異常報酬伴隨著正異常交易量，而負異常報酬卻因放空限制等因素，無顯著正異常交易量。 / This paper analyzes the last four political events, which includes KMT's Taipei Mayoral Primary, Taipei Mayor Election, KMT Presidential Candidate Replacement and Taiwan Presidential Election. We use trading volume to detect whether investors join the market due to potential political investment opportunity. Then we examine the CAR tendency with the political event results to identify its forecast ability. Last, we detect CAR within 5 days later to find if investor overreact or underreact before the event day. We find that the CAR meets voters’ political anticipation before the event window. Investors believe they can time the market through these events and gain profit. Furthermore, stocks relevant to those elects experience positive CAR. In contrast, stocks relevant to those also-rans do not experience significant returns. The only fortuneteller is the company, which has close relationship to the defeated candidate, telling with negative CAR. Because of short-sale constrains, the trading volume are not larger than before even it is a good chance to gain profit in the political events.

本研究針對台灣股票市場自民國51年2月9日至民國91年12月31日，對股票報酬率與風險溢酬進行實證研究，並且依據Fama與French的方法論，依照台灣發行量加權股價指數(TAIEX)的標準進行資料採樣，以平均股票報酬模式(average stock return model)、股利成長模式(dividend growth model)與盈餘成長模式(earning growth model)來進行估計，研究結果發現台灣股票市場的風險溢酬相較於以美國為主的已開發市場，呈現較高水準的風險溢酬與高度的變異性，此外，在經由變異數調整後，依據基本分析方法(股利成長模式與盈餘成長模式)所衡量的風險溢酬較具一致性，並且股利成長模式為較優秀的風險溢酬估計式，其原因包括其在各個投資期間有較低的標準差，以及在數值上呈現了一致性。 另外，在實質股利成長率與盈餘成長率的預估上，無論是一年期或二年期的預測力都相當低。股利股價比(Dt/Pt)與盈餘股價比(Yt/Pt)在近年呈現下滑的趨勢，根據價值分析理論，暗示未來股票報酬率有可能降低，然而，依據淨值市值比(BE/ME) 效應，則出現與前述推論不一致的結果。 不過，三種模式的風險溢酬皆高於平均收益報酬率A(Yt/Bt-1)，顯示投資報酬率低於資金成本，違反投資淨現值大於零的原則，因此依賴基本分析對台灣這類新興市場進行分析可能存在適用上的問題。為加強對台灣股票市場的解釋，本研究另加入其它觀點與行為財務學角度的解釋，來說明台灣股票市場高風險溢酬與高變異的特性。 / Abstract This thesis is an empirical analysis of the risk premium of Taiwan’s equity market. The sample period covers 1962/2/9 - 2002/12/31. My methodology is based on the average stock return model, dividend growth model, and earning growth model suggested by Fama and French (2002). In contrast to the United States, which is a developed market, my empirical results indicate that Taiwan’s equity market is characterized by high risk premium and high volatility. In addition, fundamental analyses (dividend growth model and earning growth model) yield similar results. But the dividend growth model has the best forecasting ability of the three models because of its high consistency and low variability. The real dividend growth rate and the real earning growth rate have low predictability of one- and two-year growth rates in the future. The Dt/Pt and Yt/Pt ratio decrease in recent years, implying that future equity premium will decrease. However, the BE/ME ratio does not behave similarly. Surprisingly, the equity premium of all three models is grater than the average real income return on book equity. The expected stock return exceeding the expected income return on book equity implies that typical corporate investments have negative net present value. Thus, applying fundamental analysis to an emerging market such as Taiwan may not be suitable. I try to provide sensible explanations for my findings on the Taiwan’s equity market by referencing viewpoints expressed in the behavioral finance literature and other literatures.

估價平滑相關研究大多採取Quan and Quigley（1991）提出的理性行為解釋估價師的部分調整行為。然而財務行為學認為行為人通常具有非理性的行為，因此本研究以保守性心理偏誤解釋部分調整行為，探討非理性心理對於估價師之影響。為了捕捉估價師個體的行為資訊，本研究以台灣不動產投資信託的重估價結果作為研究資料，檢視估價師對市場新資訊和過去估值的處理行為。本研究首先應用部分調整模型進行實證分析，結果發現台灣估價師確實具有部分調整行為，相對較低估市場價值資訊的權重，而此類保守性偏誤即為總體估價平滑效果的展現。此外進一步以Tobit Model進行市場價值資訊權重的影響因素分析，結果發現市場資訊的權重會受到心理效果而改變，當估價方法價值差異率越大時，估價師會降低市場資訊的權重，展現了對模糊事件的趨避傾向。捷思法可以幫助估價師提升效率，但過度依賴捷思法亦有可能引發錯誤和偏差；本研究發現模糊趨避效果顯示估價師更重視市場資訊的可信度，似較符合估價證據之要求，故非理性心理偏誤對估價師言非全為負面影響。 / Previous appraisal smoothing studies were based on the appraiser rational behavior assumptions which developed by Quan and Quigley (1991) and found that appraisers do insufficiently react to market fluctuate. However, this paper starts with behavioral finance theory to investigate the partial adjustment behavior of Taiwan appraisers, using the reappraisals of Taiwan Real Estate Investment Trust (T-REITs) to show each appraiser’s behavioral intention at disaggregate level which could be diversified away at aggregate level. This study tests the partial adjustment model to observe the weights of new market information and previous estimates of the same property reappraisal. The results show that Taiwan appraisers more rely on the previous estimates, namely the conservatism bias, which will induce the appraisal smoothing at aggregate level. Further, the Tobit Model is employed to test the influence factors of appraisers’ weight on new market information, and the results show weights on new market information will be affected by appraisers’ heuristics. There is a negative effect to weights on new market information which stands for the ambiguity aversion of appraisers. It shows appraiser do concern about the reliability of market evidence which is much accord with the appraisal normative model.