本文研究分為兩個部分。第一部分以全球發行追蹤標的為各國加權指數之兩倍槓桿型ETF為研究樣本，探討各槓桿型ETF發行日起至2016年3月間ETF的追蹤誤差。再者，分析槓桿型ETF之個別績效，以槓桿型ETF報酬率與追蹤標的指數報酬率差異之絕對值作為追蹤誤差之衡量，各槓桿型ETF平均之追蹤誤差在0.01到2.96個基本點，在45檔槓桿型ETF中，有24檔槓桿型ETF在10%顯著水準下異於零。第二部分進一步分析在10%顯著水準下顯著異於0之槓桿型ETF，造成其追蹤誤差的原因， 研究結果顯示，總費用率、匯率變動率、資產規模變動率、成交量變動率、標的指數成分股數、追蹤標的之地區等原因皆與追蹤誤差皆顯著影響槓桿型ETF追蹤誤差的大小。

本研究旨在探討公司系統性風險與會計變數之關聯性。影響公司系統性風險之因素應包括公司內部因素與公司外部總體經濟因素，但過去文獻並未完全涵蓋到，致使其模式解釋力皆不高。為彌補過去文獻之不足，本研究先以理論推導方式將公司內部與外部因素納入系統性風險模式中，再以實證資料驗證之。 　　模型推導結果顯示，影響系統性風險之因素包括公司盈餘、營運槓桿度、財務槓桿度、帳面價值、股利、市場組合報酬率、無風險報酬率，以及其他總體經濟因素等。理論推導結果產生三大主要命題： 　　1. 在公司前期盈餘為正及當期銷貨成長率為正，以及公司當期之每股盈餘、每股帳面價值及每股現金股利對股價具有正向影響時，公司當期總槓桿程度（營運槓桿度與財務槓桿度之乘積）對系統性風險具有正向影響。 　　2. 在公司前期盈餘為正，以及公司當期之每股盈餘、每股帳面價值及每股現金股利對股價具有正向影響時，公司當期每股現金股利對系統性風險具有正向影響。 　　3. 當公司當期銷貨成長率為正時，營運槓桿度與財務槓桿度為正向相關；但當公司當期銷貨成長率為負時，營運槓桿度與財務槓桿度具有抵換關係。 　　根據上述命題，本研究設立三項假說。第一，公司總槓桿程度對系統性風險具有正向影響，而營運槓桿度與財務槓桿度對系統性風險之影響皆為正向（或負向）。第二，公司發放現金股利對系統性風險具有正向影響。第三，在系統性風險與盈餘皆不變的額外前提下，當銷貨成長率為負時，營運槓桿度與財務槓桿度具有抵換關係;當銷貨成長率為正時，營運槓桿度與財務槓桿度為正相關。 　　實證結果部分支持上述三項假說。首先，公司總槓桿程度、財務槓桿度及現金股利皆對系統性風險具有顯著正向影響。因此，公司可利用降低總槓桿程度、財務槓桿度及減少現金股利之策略來減低系統性風險。其次，市場組合報酬、通貨膨脹率及國民生產毛額成長率等總體經濟因素，對系統性風險皆具有負向顯著影響。此結果說明導致公司系統性風險上升之因素應該包括公司內部與外部因素。因此，公司欲降低風險時，除了利用總槓桿程度、財務槓桿度與股利政策外，尚須考慮其他總體經濟變化。最後，實證結果亦顯示，當公司正處於銷貨成長時期，以追求成長為目標，可能同時面臨高營運風險與高財務風險。然而，在銷貨衰退時，公司卻不必然會以風險控管為目標。因此，營運槓桿度與財務槓桿度並不存在抵換關係。 / This thesis examines the relationship between firm systematic risk and accounting variables. Potential determinants of firm systematic risk theoretically include accounting and macroeconomic variables, but prior research only explored part of them and most models yielded low explanatory power. This research analytically derives and empirically verifies a model of firm systematic risk. 　　The analytical results suggest that determinants of systematic risk at least include earnings, the degree of operating leverage, the degree of financial leverage, book value, dividend, market-portfolio return, risk-free return and other macroeconomic variables. Three main propositions are therefore derived as follows. 　　1. When a firm's prior year earnings and current year sales growth are both positive, if its current book value, cash dividend, and earnings all have a positive effect on its stock price, then its degree of total leverage, defined as the product of degree of operating leverage and degree of financial leverage, has a positive effect on its systematic risk. 　　2. When a firm's prior year earnings is positive, if its current book value, cash dividend, and earnings all have a positive effect on its stock price, then its current cash dividend has a positive effect on its systematic risk. 　　3. When a firm's current year sales growth is positive (negative), its degree of operating leverage is positively (negatively) related with its degree of financial leverage. 　　Three hypotheses are then tested empirically. First, a firm's degree of total leverage has a positive effect on its systematic risk; and its degree of operating leverage and degree of financial leverage both have a positive (or both negative) effect on its systematic risk. Second, a firm's cash dividend has a positive effect on its systematic risk. Third, if a firm's sales growth is positive (negative) without any change in its systematic risk or earnings, then its degree of operating leverage is positively (negatively) related with its degree of financial leverage. 　　The empirical results provide partial support for the above hypotheses. First, the degree of total leverage, degree of financial leverage, and cash dividend each has a positive effect on the systematic risk. Therefore, a firm can reduce its systematic risk by lowering its degree of total leverage, degree of financial leverage and the cash dividend. Second, macroeconomic factors such as the market-portfolio return, inflation and GNP growth have a negative effect on the systematic risk. Hence, a firm attempting to control its systematic risk should consider the changes of macroeconomics besides the leverage and dividend policy. Finally, a firm with growing sales takes a high degree of operating leverage and financial leverage, but a firm does not necessarily take a high (low) degree of operating leverage and a low (high) degree of financial leverage as target when its sales are declining. In other words, these two leverages have no offset relationship.

自從槓桿型指數基金於各股票市場發行後，各國主管機關皆紛紛發出聲明表示，該商品並不適合長期持有；因此，該類型投資商品的公開說明書皆會註明不宜長期投資。然而，本研究實證結果發現，持有期間長短並非主要風險來源。雖然，如大家所知，槓桿型指數基金多是以「日」為單位追蹤指數，而導致複利效果 (Compounding Effect) 使基金長期報酬與槓桿倍數不同。 根據算出上述的報酬差異(Return Difference)可以發現不論是正向2倍或是反向1倍皆與台灣50報酬率的標準差有統計上顯著關係。反向1倍皆與台灣50報酬率的標準差有顯著負相關；反之，正向2倍與台灣50報酬率的標準差有顯著正向相關。然而，從已實現乘數(Realized Multiple)的分佈中可發現，不合理值並不隨投資期間越長而越多。意即儘管投資期間越長，並不一定會導致複利效果越大，而與目標槓桿倍數脫節。再者，隨著投資期間越長，波動度(volatility)的對於報酬差異的解釋力越強；因此，若想長期投資槓桿型指數資金，預測標的波動度的能力更顯為重要。 / When we browse the reports about the inverse and leveraged ETF, most of them emphasize that the LETF is not appropriate to long-term investors. However, in this research, we attempt to demonstrate the main factor of the performance of the leveraged and inverse ETF is not how long the LETF we hold, but the volatility of the underlying index or ETF. Observing the empirical test, no matter how long the investment horizon is, the coefficient of the variance of the Taiwan 50 is statistically significant both in the Taiwan 50 Bear -1X and the Taiwan 50 Bull. However, its effect on the Bear -1X is opposite to that on the Bull 2X. First, the relationship between the volatility and the return difference of Taiwan 50 and the Bear -1X is negative. In contrast, the relationship between the volatility and the return difference of Taiwan 50 and the Bull 2X is positive. However, in accordance with the distribution of the realized multiples, the frequency of either the Bear -1X or Bull 2X was not more and more when the holding period is longer. As a result, our research show the variance has a significant effect on both, no matter how long investors hold. If the volatility is moderate, the return difference may be close to zero; then the LETFs would be a convenient way to investors who desire to magnify the market return. Moreover, due to the increasing explanatory power of the volatility, we may make a further inference that whether the compounding effect is positive or negative depends on the volatility, especially within longer holding period. Therefore, without the great ability to forecast the variance, the LETFs are not recommended to the long-term investors.

多角化經常是許多企業追求持續成長、價值創造所採取的策略之一，然而，大多數的實證結果均發現多角化策略減損了公司價值，而且企業的總價值顯著低於各部門獨立經營應有之價值，存在折價現象，顯示出多角化策略不僅無法創造價值，反而有損於公司價值。如此一來，「多角化」似乎背離了管理者理應採行提昇股東財富之策略。 為何多角化企業會存在折價現象呢？從實證上觀察到多角化企業的財務槓桿程度愈高，價值減損的幅度愈大，似乎兩者存在某種關係。因此，本研究將驗證財務槓桿與多角化企業價值之關係，探究多角化企業折價幅度是否與財務槓桿有關，以瞭解實務中多角化企業普偏負債比率較高的現象是否合理。本研究以美國公開交易公司作為研究對象，樣本期間為1996年至2001年，利用Berger & Ofek（1995）所提出的超額價值作為公司價值的衡量指標，本研究實證結果發現： 一. 不同多角化策略階段下，財富移轉效果之影響存在顯著差異 二. 不同多角化策略階段下，多角化策略與財務槓桿的交互作用效果與公 司價值之關係存在顯著差異 三. 多角化策略確能帶來舉債利益，但隨著多角化程度增加而遞減。 在實證的過程中發現，若以靜態觀察多角化公司價值與財務槓桿間的關係，兩者的確存在顯著負向關係；然而，若捕捉策略調整下的動態過程，發現多角化策略價值變動數與財務槓桿之間為正向關係，意即財務槓桿愈高之企業，進行多角化策略後，價值增加的幅度愈大抑或價值減損的幅度愈小。換言之，財務槓桿並非導致多角化公司折價之因素，反而有利於價值創造。

隱含波動度不對稱效果對選擇權市場參與者是很重要的，因為隱含波動度變大可增加選擇權買方的報酬相對的會減少選擇權賣方的報酬，並且對選擇權避險者來說是一種額外的風險。過去許多文獻皆已證實股票報酬波動度具有不對稱效果，所謂不對稱效果一般是指負向衝擊對報酬波動度增加的影響較正向衝擊大。然而多數研究是以條件變異數作為波動度的衡量，本研究則打算以選擇權之隱含波動度作為波動度的衡量。研究對象為歐洲期貨交易所交易之二十四家德國公司個股選擇權，利用EGARCH模型探討股票價格變動對個股選擇權之隱含波動度不對稱效果，研究期間從2000年2月14日至2001年12月31日。在不對稱效果成立之下，進而探討公司財務槓桿及公司規模對隱含波動度不對稱程度之影響。除以最小平方法模型分析並與Koutmos and Saidi(1995)對照外，更進一步以Panel Data模型加入公司效果或時間效果作為本研究最終目的的分析依據，研究期間從2000年至2001年。 本研究實證結果如下： 1.大多數公司股票選擇權之隱含波動度具有不對稱效果，也就是負向價格變動對隱含波動度增加的影響較正向價格變動大，只有兩家公司例外。 2.以最小平方法模型分析公司財務槓桿對隱含波動度不對稱程度的影響，實證結果與Koutmos and Saidi(1995)不一致，且不能支持Black（1976）所提出槓桿效果能用以解釋隱含波動度不對稱效果之假說，產生遺漏變數偏誤。 3.以Panel Data模型加入公司效果或時間效果之考量，分析公司財務槓桿對隱含波動不對稱程度的影響。實證發現隱含波動度不對稱效果可歸因於財務槓桿假說，此外證實存在時間效果但不存在公司效果。 4.公司規模會影響隱含波動度不對稱程度，兩者呈現正向關係。也就是說規模較大的公司對負向衝擊的反應較規模較小的公司敏感，實證結果與Koutmos and Saidi(1995)一致。

雖然「二次金改」的議題，市場沸沸揚揚討論，但依政府表示，A銀行的特別股私募案堪稱二次金改的「成功典範」，未來亦將牽動金控集團的倂購動向。A銀化在二次金改的推動中，A銀行首當其衝，被鎖定為第(2）項目標:94年底前官股金融機構數目至少減為6家中被減的一家。其中財政部扮演了重要的穿針引線的角色-催化並促成ㄧ家以消費金融起家的B金控高價倂購A銀行，得標金額比底價多一百多億元，到底特別股價格的合理性如何，以及特別股的特色、私募案的影響與背後意義，均為本文核心議題。 到底B金控認購特別股的價格高與否，或許現在言之過早。不過B金控開啟了「既入主A銀行，錢留在A銀行等於自己的錢」的不同思維，這種謀略不再是一般金錢數字的設算，才有了這次金融史上僅見的併購案。不僅B金控的原始大股東未花一分錢，且因B金控以融資方式認購A銀行，在未來三年中每年只花6.58億元，即可擁有1.28兆資產之百年大行庫A銀行。之後，B金控又與美、日兩大外資策略聯盟，引進資金，改善財務結構，對B金控而言，無疑是「一大成功的認購策略」。 本文將循序分析B金控異於一般投資者的逆向併購邏輯及過程，並由A銀行私募特別股對B金控的影響與背後意義來佐證「一大成功的認購策略」之論點。 / The “Second Financial Reform” has remained a hotly debated issue in the market. According to the government, however, the private placement of the preferred shares of A Bank is considered to be a “successful example” of the reform. The ill-performing A Bank, troubled by bad loans in recent years, became a second high priority target for the Second Financial Reform. It was among of the at least six government-controlled financial institutes to be reduced to by the end of 2005. The Ministry of Finance (MOF) has played a guiding role, catalyzing B, a holding company based on consumption financing, to acquire A Bank with a bid which was 10 billion higher than the base price. The core issues of this essay include the reasonableness of the base price of preferred shares, the features of preferred shares and the impacts and implications of private placement. It might be too early to say whether the price of preferred shares acquired by B Holding Company is high. However, the company started a different way of thinking by treating the funds it put into the acquired bank as its own”. The acquisition, one of the few in history, would not have materialized without a strategy as distinguished from the regular money counting ideas. Not only the did acquisition of A Bank by B Holding Company through financing cost its original shareholders nothing, it also allowed the company to have the 1.28 –trillion-worth bank by spending only 658million per annum for the next three years. The holding company later formed a strategic alliance with the financial institutes in the U.S. and Japan for supply of funds to improve its financial structure, which was undoubtedly a “very successful acquisition strategy” for the company. This essay conducts a step-by-step analysis of the company’s logic and process of the acquisition as contrary to the general investment mindset and supports the viewpoint of “very successful acquisition strategy” by discussing the impacts and implications of private placement of preferred shares by B Holding Company on A Bank.

本篇論文目的在檢視壽險業資本結構的影響因素，樣本為美國壽險公司。金融及保險市場的特殊性質會造成保險公司有動機去調整控制資本結構與資產風險，以維持公司失卻清償能力的風險在適當的範圍內。我們的實證結果中顯示槓桿比率與資產風險的負向關係的確存在於美國壽險業當中。 此外，一般公司資本結構理論中的公司大小、代理問題皆顯示對壽險公司的資本結構造成影響；公司的評比對公司的槓桿比率有顯著影響；使用獨立代理人銷售業務的公司有較高的槓桿比率與風險；實證結果亦顯示壽險公司的資產風險與投資的集中度及經營險種的集中度有正相關，意味著資產風險可以透過分散投資與險種多樣化的方式來分散降低。 / This study examines the capital structure determinants of life insurance companies in U.S. The decision of capital structure in insurance industry should be a joint decision of capital structure and firm risk. Insurers have incentive to keep firm risk within a safe range, due to the special features in the insurance markets. We find evidence supporting the negative relationship between leverage ratio and firm risk, suggesting that insurers will balance risk and leverage to keep the insolvency risk within a desired range. In addition, general capital structure theories such as size and agency theory are significant in our empirical result. Rating is shown a significant factor of capital structure in life insurance industry. Evidence suggests the independent agents have less incentive to monitor insurers. Firm risk is also shown significant relationship with undiversified investments and concentrated business line, which suggests that life insurers can reduce firm risk by diversifying investment risk and business risk.

給定一個設計矩陣X,當從X刪去一列或數列以後,X的特徵結構可能產生很 大的改變。在本文中,計算帽子矩陣H的高槓桿值,刪去如此有影響力的觀 察值後,X的特徵結構是否有改變,以及探討它的條件數。舉一些特殊的定 理, 討論從X刪去一列或數列之後的條件數。因此,我們也探討近似的條件 數,考慮兩者之間有何關係。 我們計算設計矩陣X的條件數與設計矩陣X 刪去一列或數列後的條件數,及診斷刪去有影響力的列對共線性之影響。 舉二個實例,使用 Matlab軟體計算條件數,分析它們的共線性性質, 以及 討論隱藏共線性與創造共線性的強度何者為強。

早期國外學者指出不動產投資信託基金具有「低風險」與「防禦性」的特質，簡稱為「抗跌性」與「反槓桿效果」，亦即除了與大盤相關性較低之外，市場的負向衝擊對於其報酬的影響比正向衝擊來的小。為了瞭解台灣不動產投資信託基金的價格是否存在過度恐慌，本文分成兩部分著眼：過去的研究大都著手在不動產投資信託基金的價格上，而忽略其為封閉型基金的特性，故本文先從「封閉型基金」角度，以因素分析擷取其共同因子，再利用BGARCH模型，探討共同因子與台灣VIX指數變化間是否存在不對稱效果。第二則是將不動產投資信託基金折溢價進行拆解，分為價格與資產淨值，運用相同方式重新觀察反槓桿效果，探討台灣不動產投資信託基金的價格是否存在過度恐慌。 本研究得到一些與過去文獻不同的結論： 1.若市場出現正向或是負向衝擊，台灣不動產投資信託基金折溢價的波動會上升， 且過去的衝擊持續性強，反應財務資料呈現的波動叢集性。 2.不動產投資信託基金價格具有「槓桿效果」，市場的負向衝擊會額外增加報酬的波動，並無防禦性特質。 3.REITs價格與台灣VIX指數變動具有顯著相關，而資產淨值則無此現象，證明台灣不動產投資信託基金價格相對於資產淨值存在過度恐慌的現象。

本文探討槓桿型/反向型ETF之追蹤誤差是否會因為追縱標的屬於不同產業而有差異。為此先將46檔具有顯著追蹤誤差之槓桿型/反向型ETF樣本分為七類不同的產業，並設定產業作虛擬變數放入多元迴歸模型中，結果為當追蹤指數屬於建築業、零售業及服務業之槓桿型/反向型ETF傾向得到較大的追蹤誤差，而追蹤指數屬於礦產業、製造業、水電業及金融業者，其追蹤誤差較小。追蹤誤差大之產業，其指數波動度也較大，推測指數波動度可能是造成特定產業追蹤績效不佳的原因。另外在本研究中也發現，追蹤礦產業、建築業及製造業指數之槓桿型/反向型ETF，其槓桿型績效優於反向型績效；而追蹤水電業、零售業、金融業及服務業指數之槓桿型/反向型ETF則是反向型績效優於槓桿型績效。其原因推測與ETF存在年限、ETF發行公司和指數成分股個數有關。 / This paper discusses whether the tracking error of leveraged/inverse ETF varies by industry of underlying indexes. We take 46 leveraged/inverse ETFs with significant tracking error into the samples, divide them into 7 different industries and set those industries as dummy variables in multiple regression models. The outcome shows that the tracking error tends to be larger if the underlying index of leveraged/inverse ETF belongs to construction, retails or service industries; otherwise, the tracking error tends to be smaller if it tracks the index from mining, manufacturing, utility or finance industries. The larger tracking errors may result from more volatile indexes of those industries. Besides, we also find that leveraged ETFs outperform the inverse ones if the index belongs to mining, construction or manufacturing industry; on the other hand, inverse ETFs outperform leveraged ones when the index comes from utility, retails, finance or service industry. The possible reasons may be the different characteristics of ETFs, such as the length of ETF’s existence, ETF’s issuer and the number of constituents in underlying index.