In this thesis we prove several results on the structure of solutions to optimal transportation problems. The second chapter represents joint work with Robert McCann and Micah Warren; the main result is that, under a non-degeneracy condition on the cost function, the optimal is concentrated on a $n$-dimensional Lipschitz submanifold of the product space. As a consequence, we provide a simple, new proof that the optimal map satisfies a Jacobian equation almost everywhere. In the third chapter, we prove an analogous result for the multi-marginal optimal transportation problem; in this context, the dimension of the support of the solution depends on the signatures of a $2^{m-1}$ vertex convex polytope of semi-Riemannian metrics on the product space, induce by the cost function. In the fourth chapter, we identify sufficient conditions under which the solution to the multi-marginal problem is concentrated on the graph of a function over one of the marginals. In the fifth chapter, we investigate the regularity of the optimal map when the dimensions of the two spaces fail to coincide. We prove that a regularity theory can be developed only for very special cost functions, in which case a quotient construction can be used to reduce the problem to an optimal transport problem between spaces of equal dimension. The final chapter applies the results of chapter 5 to the principal-agent problem in mathematical economics when the space of types and the space of available goods differ. When the dimension of the space of types exceeds the dimension of the space of goods, we show if the problem can be formulated as a maximization over a convex set, a quotient procedure can reduce the problem to one where the two dimensions coincide. Analogous conditions are investigated when the dimension of the space of goods exceeds that of the space of types.

In this thesis we prove several results on the structure of solutions to optimal transportation problems. The second chapter represents joint work with Robert McCann and Micah Warren; the main result is that, under a non-degeneracy condition on the cost function, the optimal is concentrated on a $n$-dimensional Lipschitz submanifold of the product space. As a consequence, we provide a simple, new proof that the optimal map satisfies a Jacobian equation almost everywhere. In the third chapter, we prove an analogous result for the multi-marginal optimal transportation problem; in this context, the dimension of the support of the solution depends on the signatures of a $2^{m-1}$ vertex convex polytope of semi-Riemannian metrics on the product space, induce by the cost function. In the fourth chapter, we identify sufficient conditions under which the solution to the multi-marginal problem is concentrated on the graph of a function over one of the marginals. In the fifth chapter, we investigate the regularity of the optimal map when the dimensions of the two spaces fail to coincide. We prove that a regularity theory can be developed only for very special cost functions, in which case a quotient construction can be used to reduce the problem to an optimal transport problem between spaces of equal dimension. The final chapter applies the results of chapter 5 to the principal-agent problem in mathematical economics when the space of types and the space of available goods differ. When the dimension of the space of types exceeds the dimension of the space of goods, we show if the problem can be formulated as a maximization over a convex set, a quotient procedure can reduce the problem to one where the two dimensions coincide. Analogous conditions are investigated when the dimension of the space of goods exceeds that of the space of types.

Submitted by Marcelo de Castro Orefice (marcelo.orefice@gmail.com) on 2017-02-21T03:26:05Z No. of bitstreams: 1 Dissertação Marcelo Orefice.pdf: 869159 bytes, checksum: 0309054015ee724b48e5ea290305b527 (MD5) / Rejected by Renata de Souza Nascimento (renata.souza@fgv.br), reason: Marcelo, boa tarde Só deverá submeter o trabalho, após incluir a ficha catalográfica enviada pela biblioteca. Sua apresentação ocorreu em 2017 - Alterar São Paulo 2016 para São Paulo 2017. Att on 2017-02-21T16:23:16Z (GMT) / Submitted by Marcelo de Castro Orefice (marcelo.orefice@gmail.com) on 2017-02-22T14:45:12Z No. of bitstreams: 1 Dissertação Marcelo Orefice com Ficha Catalográfica.pdf: 782628 bytes, checksum: 19ae6d2959fe63578398457a9f044ee9 (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2017-02-22T15:27:53Z (GMT) No. of bitstreams: 1 Dissertação Marcelo Orefice com Ficha Catalográfica.pdf: 782628 bytes, checksum: 19ae6d2959fe63578398457a9f044ee9 (MD5) / Made available in DSpace on 2017-02-22T17:40:14Z (GMT). No. of bitstreams: 1 Dissertação Marcelo Orefice com Ficha Catalográfica.pdf: 782628 bytes, checksum: 19ae6d2959fe63578398457a9f044ee9 (MD5) Previous issue date: 2017-02-07 / In this dissertation, we discuss the practice of portfolio pumping in Brazil. Although the topic is recurrent in other countries, few studies provide this analysis for the Brazilian case. The statistical study is elaborated in three stages: first, we considered Brazilian investment funds‟ shares for the period from September 2011 to June 2016, estimating daily abnormal returns of those funds based on the Ibovespa, considering and not considering the adjusted beta of the portfolios of those funds. Our results suggest that the practice of portfolio pumping is more frequent at the end of months ex-semester than at the end of semesters. When we consider the beta adjusted to calculate abnormal returns of the funds, we found a greater significance for the existence of this practice. In the second step, the funds were ordered based on their performance in the previous period (by month, semester, and year), which resulted in few relevant results for the analysis of the topic, despite what is proposed by the principal-agent problem literature. In the last step, we analyzed the practice of portfolio pumping in stocks traded on BM&F Bovespa, ordering them by their participation in the portfolios and by their Market Cap. The results indicated that the stocks with greater presence in the portfolios of the investment funds have higher abnormal returns at the end of the periods, reinforcing the thesis that this increase in stock prices in those moments may be a consequence of a deliberate action taken by the managers of those funds / Nesta dissertação, discutimos a prática de portfolio pumping para o caso brasileiro. Embora o tema seja frequente em outros países, são poucos os estudos que realizam essa análise para o Brasil. O estudo estatístico foi realizado em três etapas: na primeira, consideramos o valor das cotas de fundos brasileiros de investimento em ações para o período de setembro de 2011 a junho de 2016, calculando o retorno anormal diário desses fundos com base no Ibovespa, com e sem a consideração do beta ajustado das carteiras desses fundos. O resultado observado sugeriu que a prática de portfolio pumping é mais frequente ao final dos meses ex-semestre do que ao final dos semestres. Quando consideramos o beta ajustado para o cálculo do retorno anormal dos fundos, verificamos maior significância para a existência dessa prática. Na segunda etapa, os fundos foram ordenados com base em seu desempenho no período anterior (mês, semestre e ano), com resultados observados pouco esclarecedores para a análise do tema, diferentemente do que é sugerido pela literatura do problema do principal-agente. Na última etapa, analisamos a prática de portfolio pumping nas ações negociadas na BM&F Bovespa, ordenando-as pela sua participação nos portfólios e pelo seu Market Cap. Os resultados obtidos indicaram que as ações com maior presença nos portfólios dos fundos de investimento têm retornos anormais mais elevados ao final dos períodos, reforçando a tese de que esse aumento nos preços de ações naqueles instantes pode ser uma consequência de uma ação deliberada por parte dos gestores desses fundos.

The core focus of this thesis is the principal-agent problem and its role in the outbreak of the 2007 financial crisis. Analysis of key elements of the problem like moral hazard and adverse selection lays foundations for identifying situations, in which conflict of interests led to negative impacts on the economic performance before and during the financial crisis. This work also studies the influence of factors that are often overlooked by theoretical economists but are still tightly connected to the principal-agent problem, like exogenous incentives based on the mechanism of trust. The author also evaluates mechanisms that have been put into place after the financial crisis and which could help lower agency costs. He also sketches out possible venues of future research in this area.

This diploma thesis focuses on the moral hazard aspects in the financial crisis on the USA market in the years 2007-2009, that later became known as a global crisis of real economic. The thesis aim is to prove a role of moral hazard in the financial crisis origins, role that can be detected in the activities and behavior of the market participants. There is a survey of the individual causes of the crisis and interpretation in the terms of moral hazard. First part of the thesis focuses on the theoretical basics of moral hazard that is being represented by the Principal-Agent Problem. This theory is further applied on the financial market theories. Second part of the thesis describes the origin and development of the financial crisis. It represents the introduction of the crisis without exploring the causes. The third part of the thesis is the core part, there is presented the role of moral hazard in the financial crisis due to the synthesis of the two previous parts. Item by item there are introduced the activities on the crisis market, where the principle of these activities is presented in terms of moral hazard. This part also includes moral hazard in the society that influences also the financial market. The last part focuses on the regulation of the financial markets in connection with the possibilities of moral hazard elimination.

Cette thèse a pour objectif la compréhension de plusieurs aspects du caractère rugueux de la volatilité observé de manière universelle sur les actifs financiers. Ceci est fait en six étapes. Dans une première partie, on explique cette propriété à partir des comportements typiques des agents sur le marché. Plus précisément, on construit un modèle de prix microscopique basé sur les processus de Hawkes reproduisant les faits stylisés importants de la microstructure des marchés. En étudiant le comportement du prix à long terme, on montre l’émergence d’une version rugueuse du modèle de Heston (appelé modèle rough Heston) avec effet de levier. En utilisant ce lien original entre les processus de Hawkes et les modèles de Heston, on calcule dans la deuxième partie de cette thèse la fonction caractéristique du log-prix du modèle rough Heston. Cette fonction caractéristique est donnée en terme d’une solution d’une équation de Riccati dans le cas du modèle de Heston classique. On montre la validité d’une formule similaire dans le cas du modèle rough Heston, où l’équation de Riccati est remplacée par sa version fractionnaire. Cette formule nous permet de surmonter les difficultés techniques dues au caractère non markovien du modèle afin de valoriser des produits dérivés. Dans la troisième partie, on aborde la question de la gestion des risques des produits dérivés dans le modèle rough Heston. On présente des stratégies de couverture utilisant comme instruments l’actif sous-jacent et la courbe variance forward. Ceci est fait en spécifiant la structure markovienne infini-dimensionnelle du modèle. Étant capable de valoriser et couvrir les produits dérivés dans le modèle rough Heston, nous confrontons ce modèle à la réalité des marchés financiers dans la quatrième partie. Plus précisément, on montre qu’il reproduit le comportement de la volatilité implicite et historique. On montre également qu’il génère l’effet Zumbach qui est une asymétrie par inversion du temps observée empiriquement sur les données financières. On étudie dans la cinquième partie le comportement limite de la volatilité implicite à la monnaie à faible maturité dans le cadre d’un modèle à volatilité stochastique général (incluant le modèle rough Bergomi), en appliquant un développement de la densité du prix de l’actif. Alors que l’approximation basée sur les processus de Hawkes a permis de traiter plusieurs questions relatives au modèle rough Heston, nous examinons dans la sixième partie une approximation markovienne s’appliquant sur une classe plus générale de modèles à volatilité rugueuse. En utilisant cette approximation dans le cas particulier du modèle rough Heston, on obtient une méthode numérique pour résoudre les équations de Riccati fractionnaires. Enfin, nous terminons cette thèse en étudiant un problème non lié à la littérature sur la volatilité rugueuse. Nous considérons le cas d’une plateforme cherchant le meilleur système de make-take fees pour attirer de la liquidité. En utilisant le cadre principal-agent, on décrit le meilleur contrat à proposer au market maker ainsi que les cotations optimales affichées par ce dernier. Nous montrons également que cette politique conduit à une meilleure liquidité et à une baisse des coûts de transaction pour les investisseurs. / The aim of this thesis is to study various aspects of the rough behavior of the volatility observed universally on financial assets. This is done in six steps. In the first part, we investigate how rough volatility can naturally emerge from typical behav- iors of market participants. To do so, we build a microscopic price model based on Hawkes processes in which we encode the main features of the market microstructure. By studying the asymptotic behavior of the price on the long run, we obtain a rough version of the Heston model exhibiting rough volatility and leverage effect. Using this original link between Hawkes processes and the Heston framework, we compute in the second part of the thesis the characteristic function of the log-price in the rough Heston model. In the classical Heston model, the characteristic function is expressed in terms of a solution of a Riccati equation. We show that rough Heston models enjoy a similar formula, the Riccati equation being replaced by its fractional version. This formula enables us to overcome the non-Markovian nature of the model in order to deal with derivatives pricing. In the third part, we tackle the issue of managing derivatives risks under the rough Heston model. We establish explicit hedging strategies using as instruments the underlying asset and the forward variance curve. This is done by specifying the infinite-dimensional Markovian structure of the rough Heston model. Being able to price and hedge derivatives in the rough Heston model, we challenge the model to practice in the fourth part. More precisely, we show the excellent fit of the model to historical and implied volatilities. We also show that the model reproduces the Zumbach’s effect, that is a time reversal asymmetry which is observed empirically on financial data. While the Hawkes approximation enabled us to solve the pricing and hedging issues under the rough Heston model, this approach cannot be extended to an arbitrary rough volatility model. We study in the fifth part the behavior of the at-the-money implied volatility for small maturity under general stochastic volatility models. In the same spirit as the Hawkes approximation, we look in the sixth part of this thesis for a tractable Markovian approximation that holds for a general class of rough volatility models. By applying this approximation on the specific case of the rough Heston model, we derive a numerical scheme for solving fractional Riccati equations. Finally, we end this thesis by studying a problem unrelated to rough volatility. We consider an exchange looking for the best make-take fees system to attract liquidity in its platform. Using a principal-agent framework, we describe the best contract that the exchange should propose to the market maker and provide the optimal quotes displayed by the latter. We also argue that this policy leads to higher quality of liquidity and lower trading costs for investors.

La planification temporelle constitue conceptuellement une partie du raisonnement temporelle et il appartient au domaine de recherche de l'intelligence artificielle. La planification temporelle peut être considérée comme une extension de la planification classique par les aspects temporels de l'action. La planification temporelle est généralement complété par des préférences ou des types différents decontraintes imposées à l'exécution des actiones. Il existe de nombreuses approches à ce problème. D'une part, il existe différents paradigmes pour la planification temporelle, par example: la planification par un recherche d'une solution optimale dans des graphes de planification (STRIPS), la planification via la satisfiabilité ou la planification pardes processus de Markov. Ces approches sont mutuellement incompatibles. D'autre part, la planification temporelle exige une sujet-spécification – comme il est défini d'une manière méthodologique. Selon cette situation, cette thèse vise à proposer une analyse approfondi de la planification avec des contraintes floues qui contient quelques remèdes à ces difficultés. À savoir, deux approches à la représentation et la modélisation de ces questions sont mises.Dans la première (chapitre 2, chapitre 3) - les relations floues d'Allen en tant que contraintes temporelles floues sont représentés par des normes de convolutions dans un espace de Banach des fonctions intégrables de Lebesgue. Cela nous permet de nous immergerles relations d'Allen dans les contextes computationnels de la planification temporelle (basée sur STRIPS et sur la procedure de Davis-Putnam) et d'élucider leur nature quantitative. Cette approche est développée dans un contexte des problèmes par systèmes multi-agents comme un sujet de cette approche. Dans les chapitres 4 et 5 les contraintes temporelles floues avec flou – introduit par préférences - sont représentées en termes logiques de la logique préférentielle de Halpern-Shoham. Cela nous permet d'adopter ces resultats dans une construction du contrôleur du plan. Cette approche est développée dans un contexte du problème du voyageur de commerce. Enfin, une tentative de réconcilier ces deux lignes de représentation des contraintes temporelles floues a été proposée dans le dernier chapitre. / Temporal planning forms conceptually a part of temporal reasoning and it belongs to research area of Artificial Intelligence and it may be seen as an extension of classical planning by temporal aspects of acting. Temporal planing is usually complemented by considering preferences or different types of temporal constraints imposed on execution of actions. There exist many approaches to this issue. One one hand, there are different paradigms to temporal planning, such as: planning via search in graphs (STRIPS), planning via satisfiability or planning in terms of Markov processes. These approaches are mutually incompatible. In addition, temporal planning requires a subject-specification as it is rather defined in a methodological way. On the other hand, temporal constraints are represented and modeled in different ways dependently on their quantitative or qualitative nature. In particular, Allen’s relations between temporal intervals – an important class of temporal constraints – do not have any quantitative aspects and cannot be considered in computational contexts. According to this situation, this PhD-thesis is aimed at the proposing a depth-analysis of temporal planning with fuzzy constraints which contains some remedies on these difficulties. Namely, two approaches to the representation and modeling of these issues are put forward. In the first one (chapter 2, chapter 3) – fuzzy Allen’s relations as fuzzy temporal constraints are represented by norms of convolutions in a Banach space of Lebesgue integrable functions. It allows us immerse Allen’s relations in the computational contexts of temporal planning (based on STRIPS and on DavisPutnam procedure) and to elucidate their quantitative nature. This approach is developed in a context of Multi-Agent Problem as a subject basis of this approach. In the second one (chapter 4, chapter 5) – fuzzy temporal constrains with fuzziness introduced by preferences are represented in a logical terms of Preferential Halpern-Shoham Logic. It allows us to adopt these result in a construction of the plan controller. This approach is developed in a context of Temporal Traveling Salesman Problem as a subject basis of this approach. Finally, an attempt to reconcile these two lines of representation of fuzzy temporal constraints was also proposed.

在資本主義盛行之今日，「公司治理」已成為舉世矚目之焦點。自從1997年亞洲爆發金融危機後，許多國際組織、非政府組織及亞洲經濟體投入大量的資源與時間研究此一議題，認為完善的公司治理是提昇企業策略能力及國際競爭力的最佳良方。在台灣，隨著資本主義盛行的浪潮，近二十年來產業轉型、高科技事業蓬勃發展的幕後功臣，除了政府主導科技研發計劃與租稅優惠及專案補助措施外，首推「創業投資業」；創業投資業之所以成為台灣高科技產業之推手，主要在於其「投資後管理」之積極性、策略性功能。根據經濟合作暨發展組織（OECD）於2002年所發表之研究報告顯示，創業投資業投資額佔GDP比率與該國公司治理績效之評比成正相關，由此可見創業投資業對提昇公司治理的貢獻。 本研究之目的在探討創業投資業於各投資階段，增進轉投資事業公司治理之最佳實務作法，期能提供我國創業投資業者參酌、運用，以發揮其對被投資公司積極、正面之監督控管功能，並符合公司治理的四個基本原則：公平性（Fairness）、透明性（Transparency）、課責性（Accountability）、責任性（Responsibility），建立二者間良性、穩定的溝通模式及信賴關係，讓創業投資業繼續為台灣產業競爭力加分。 「創業投資」（Venture Capital）又稱「風險投資」，而專為投資於高科技、高風險、具發展潛力且成長快速之事業所募集之資金，則為「創業投資基金」（Venture Capital Fund）。創業投資的投資模式是在高風險、高報酬的前提下，對具成長潛力的被投資公司提供中、長期資本及經營、資源、網絡的加值服務，協助投資標的快速成長，克服各項經營問題，申請上市櫃，以實現其高額之資本利得；因此，創業投資業與被投資公司間之投資關係，是建立在相互合作之互信基礎，互謀長期最大利益，而創業投資業最重視的則是，投資資金是否能順利回收及獲利，被投資公司的獲利是否有如預期、體制運作是否健全，進而進入資本市場，讓其所投入之資金有退出機制。 公司治理 (Corporate Governance) 這個議題是從美國發展出來的，當初國內學者對該名詞的翻譯不盡相同，有譯為「公司管理」、「公司治理」者；亦有譯為「公司監控」、「公司管控」者；在台灣統稱為「公司治理」，以兼顧管理、監控及自治、自律之功能。依據中華公司治理協會之定義：「公司治理是一種指導與管理的機制，以落實公司經營者責任為目的，在兼顧其他利害關係人利益下，藉由加強公司績效，保障股東權益」。 良好的公司治理必須符合四個原則，分別是公平性 (Fairness)、透明性 (Transparency)、課責性 (Accountability)以及責任性 (Responsibility)。公平性指的是對公司各投資人以及利益相關者予以公平合理的對待；透明性指公司財務以及其他相關資訊，必須適時適當地揭露；課責性指公司董事以及高階主管的角色與責任應該明確劃分；責任性則指公司應遵守法律以及社會期待的價值規範。 依據OECD公司治理原則之說明，良好的公司治理具有：提高經濟效率、降低資金成本、健全財務結構之基礎等功能。 在台灣，上市公司與大多數國家一樣，公司的所有權結構大都傾向於具有控制股東，而控制股東會透過金字塔結構、交叉持股與互為董事等方式而達到控制公司的目的，在此情況下，控制股東極可能透過利益輸送和掏空公司資產等方式，剝奪奪小股東的財富，並衍生道德危機與逆選擇的相關代理成本，而引發核心代理問題。 本研究發現創業投資業推動公司治理之誘因如下列： 一、期藉由灌輸經營團隊公司治理--「課責性」及「責任性」之基本原則，塑造正確之價值觀，建立企業長治久安之基礎。 二、強調公司治理--「公平性」之基本原則，保障少數股東之權益，維護創業投資業持有股權之股東權益及價值。 三、確信良好之公司治理會提昇轉投資事業之獲利能力及市場價值，同時增加創業投資業退出時所獲之資本利得。 本研究發現創業投資業評估轉投資事業公司治理之重點如下列： 一、轉投資事業之財會數據之正確性、可信賴度及資訊透明度、詳細度。 二、轉投資事業股東成員、股權結構及背景，及具實質影響力之股東特性。 三、轉投資事業董事會運作之情形，如：董事會成員、董事選任方式、董事會議題及會議進行模式。 四、轉投資事業經營團隊之商譽及誠信度，各項法令之遵循程度，公司章程之適法性及可執行程度。 五、董事會、董事長、總經理之權責劃分是否適當及課責性是否落實。 六、轉投資事業之會計制度是否建立且落實執行，以避免轉投資事業做假帳，並確保能按時提供財務報表。 本研究針對創業投資業推動創建期事業公司治理之方法建議如下： 一、建立企業主正規之「公司法制」觀念。 二、正視「公司章程」之法律地位及重要性。 三、導入「股份有限公司」之運作機制，建立「股東會」、「董事會」及「監察人」之基本功能及符合法令規定之決策模式。 四、建制符合「內部控制機制」之作業系統及程序。 本研究針對創業投資業推動擴充期事業公司治理之方法建議如下： 一、善用「公司章程」及「法令規章」解決「核心代理問題」。 二、強化「董事會」、「監察人」職能，推動「上市上櫃公司治理實務守則」。 本研究針對創業投資業推動成熟期事業公司治理之方法則建議由實踐股東行動主義著手；包括： 一、創投業於投資前，應確信公司治理之重要性，將投資標的推動公司治理之績效，納入投資組合考量。 二、執行投資後，創投業應經常與公司重要股東與經營階層充份溝通及於董事會、股東會中充份表達意見，以促使公司控制股東及經營階層重視少數股東之權益。 關鍵字：創業投資業、創建期、擴充期、成熟期、公司治理、核心代理問題、控制股東、公司章程、股東會、董事會、監察人 / According to the OECD research report released in 2002 that existed a positive relevant between the investment amount of venture capitalist to the GDP and the corporate governance performance in the most countries worldwide. The contribution of venture capitalist for the corporate governance is obviously numerous. The objectives of this thesis is intend to research venture capitalist how to enhance the investee corporate governance in the investees’ start-up, development, maturing stages practicably. Basing on this study, both of the venture capitalist and the investee can conduct a beneficial and stable communication manners and relationship. In Taiwan, most of the listing companies are same as the developed countries; the majority shareholders almost constitute the whole shareholders’ structure and possess the great majority voting rights. Under this circumstance, the majority shareholders instinctively intend to create the arms-in-length transactions to deprive the minority the shareholders’ benefits and wealth, and cause the core agent problems. The study was conducted by analysis, induction, and case study methodology covering three samples of venture capitalist. The results of this study are follows: 1.At the investees’ start-up stage, the measures of venture capitalist enhances corporate governance are listed as below: 1)Educating the business owner to establish the corporate legal concept. 2)Paying much attention about the article to follow the company law. 3)Introducing and following the company law’s operating structure, and establish the “shareholders”, “board of directors” and “supervisors” basic functions and decision model. 4)Establish the internal control system and procedures. 2. At the investees’ development stage, the measures of venture capitalist enhances corporate governance are listed as below: 1)Practicing the article and regulation to overcome the core agent problems. 2)Empowering the board of directors and supervisors function, and progressing the “Corporate Governance Best Practice Principles for TSEC/GTSM Listed Companies” in the investees business running. 3. At the investees’ maturing stage, the venture capitalist suggested to practice the shareholders activism; for example: taking the corporate governance performance into the investment evaluation. Key Words: venture capitalist, start-up stage, development stage, maturing stage, corporate governance, core agent problem, majority shareholders, article, shareholders’ meeting, the board of directors supervisor

財務三大決策：投資、融資與股利政策一直是學術界研究探討的主題，也是實務界關心的重要課題。過去有關融資與股利支付宣告的文章不少，但投資支出相關研究則一直被視為股權結構下的控制變數，鮮少學者探討有關投資宣告對公司價值的影響。 本研究嘗試從資訊不對稱的角度，探討首度上市上櫃公司管理當局為傳遞公司經營績效及未來成長機會給予外部投資人，透過自身的持股與投資規模進行初次投資宣告，藉以傳遞公司真正價值予投資人，以追求公司與自身財富的最大化。本研究在訊號放射理論的架構下，透過完美貝氏的分離均衡解，經過模式推導獲致三個研究命題，並以此研究命題、個案訪談與文獻整理推衍出可驗證的八個假說。 在假說驗證方面，以民國80年至89年共225家新上市上櫃公司的初次投資宣告為研究樣本，並依未來前景看佳與否、股權集中度高低與初次宣告投資規模大小來分類，將研究樣本區分為八個子樣本，利用事件研究法對宣告公司事件日的平均異常報酬率反應以及累積異常報酬率反應進行假說的檢驗。並且，以產業類別、高階主管持股、高階主管持股的二次方、投資規模、初次投資支出類型為自變數，公司規模、市場景氣與負債比率為控制變數，對事件窗口(-1,+1)的累積平均異常報酬率進行橫斷面迴歸分析，探討宣告事件對股票異常報酬率反應的影響。經由上述的實證過程，本研究獲得以下主要結論： 1.初次投資宣告具有資訊內涵，當股權集中度越高的高科技股公司，若進行大規模的初次投資支出宣告，則公司股票異常報酬率的正面反應訊號越強。 2.投資人對高科技股與非高科技股公司的初次投資宣告有不同的評價，屬於未來前景看佳的高科技股，進行初次投資宣告的股票異常報酬率反應顯著為正，且顯著高於未來前景不佳的非高科技股公司。 3.屬於資本支出類型的初次投資宣告有利於公司價值的提升。 目錄 第壹章 緒論…………………………………………... 1 第一節 研究動機與研究目的……………………………… 3 第二節 研究範圍與研究限制……………………………… 8 第三節 研究流程與論文架構……………………………… 11 第貳章 文獻探討與個案訪談整理…… ……………. 17 第一節 代理理論與資訊不對稱…………………………… 18 第二節 股權結構與公司價值之關係……………………… 23 第三節 資本支出與公司價值之關係……………………… 28 第四節 股權、資本支出與公司價值相關實證文獻……… 31 第五節 文獻評述與個案訪談整理………………………… 50 第參章 理論模式建立…………………… …………. 59 第一節 模式背景……,,…………………………………… 60 第二節 模式架構與模式條件設定………………………… 63 第三節 模式推導與命題建立……………………………… 72 第肆章 研究設計… ………………………………… 81 第一節 實證架構與研究假說……………………………… 82 第二節 資料來源與樣本選取……………………………… 93 第三節 變數操作性定義…………………………………… 96 第四節 實證分析方法……………………………………… 102 第伍章 實證結果分析…… ………………………… 113 第一節 樣本資料統計分析………………………………… 114 第二節 全體樣本初次投資支出宣告的效果分析………… 121 第三節 研究假說的驗證……………………………………. 132 第四節 投資支出宣告異常報酬率影響因素分析………… 150 第五節 實證結果與涵義……………………………………. 154 第陸章 結論與建議…… …………………………… 163 第一節 結論………………………………………………… 164 第二節 建議………………………………………………… 166 參考文獻 中文部分……………………………………………………… 171 英文部分……………………………………………………… 173 表目錄 表2-1 國外其他股權結構與公司價值關係實證文獻整理… 36 表2-2 國外其他資本支出與公司價值關係實證文獻整理… 43 表2-3 個案訪談發現與對本研究的涵義……………………. 56 表3-1 研究模式的前提設計…………………………………. 67 表3-2 研究模式符號說明……………………………………. 68 表4-1 研究範疇的分類情形…………………………………. 86 表4-2 研究範疇的分類情形與對應的假說…………………. 90 表4-3 市場景氣劃分……………………………………… 101 表5-1 整體上市上櫃公司的分配狀況……………………… 115 表5-2 研究樣本按初次投資支出宣告年度分類…………… 116 表5-3 研究樣本按產業類別分類…………………………… 117 表5-4 研究樣本按高科技類股與市場景氣分類統計……… 118 表5-5 各研究變數的樣本統計量…………………………… 119 表5-6 全體樣本初次投資支出宣告的平均異常報酬率…… 122 表5-7 全體樣本事件窗口累積異常報酬率………………… 123 表5-8 高科技股公司初次投資支出宣告的平均異常報酬…. 125 表5-9 非高科技股公司初次投資支出宣告的平均異常報酬率126 表5-10 高科技股公司事件窗口累積異常報酬率…………… 127 表5-11 非高科技股公司事件窗口累積異常報酬率………… 127 表5-12 投資規模大的公司初次投資支出宣告的平均異常報酬129 表5-13 投資規模小的公司初次投資支出宣告的平均異常報酬130 表5-14 投資規模大的公司事件窗口累積異常報酬率……… 131 表5-15 投資規模小的公司事件窗口累積異常報酬率……… 131 表5-16 研究樣本交叉分類分布情形(n=225) ……………… 132 表5-17 第一種情況公司初次投資支出宣告的平均異常報酬率134 表5-18 第一種情況公司事件窗口累積異常報酬率 135 表5-19 第二種情況公司初次投資支出宣告的平均異常報酬率136 表5-20 第二種情況公司事件窗口累積異常報酬率……………137 表5-21 第三種情況公司初次投資支出宣告的平均異常報酬率138 表5-22 第三種情況公司事件窗口累積異常報酬率……………139 表5-23 第四種情況公司初次投資支出宣告的平均異常報酬率140 表5-24 第四種情況公司事件窗口累積異常報酬率……………141 表5-25 第五種情況公司初次投資支出宣告的平均異常報酬率142 表5-26 第五種情況公司事件窗口累積異常報酬率……………143 表5-27 第六種情況公司初次投資支出宣告的平均異常報酬率144 表5-28 第六種情況公司事件窗口累積異常報酬率……………145 表5-29 第七種情況公司初次投資支出宣告的平均異常報酬率146 表5-30 第七種情況公司事件窗口累積異常報酬率……………147 表5-31 第八種情況公司初次投資支出宣告的平均異常報酬率148 表5-32 第八種情況公司事件窗口累積異常報酬率……………149 表5-33 累積異常報酬率的橫斷面多元迴歸分析………………153 表5-34 各研究假說實證結果……………………………………158 圖目錄 圖1-1 研究流程………………………………………………. 12 圖3-1 初次投資支出宣告的信號放射………………………. 63 圖3-2 研究模式與觀念架構………………………………... 65 圖3-3 初次投資出宣告的樹狀機率分析…………………… 71 圖4-1 實證架構…………………………………………………82 圖4-2 實證分析方法與流程……………………………………102 圖4-3 事件研究的各時間參數的關係……………………… 105 / Previous foreign studies on corporate investment impact on market prices have revealed that investors react positively to the announcement of increases in capital expenditure. But, the few papers study the topic in Taiwan. This study applies the agency theory, the signaling theory which reduce the asymmetric information and agent problems between the management and external investors and the perfect Baysian Equilibrium to construct the theoretic model to analyze the effect of the initial investment announcement. Following the theoretical model, case interview and literature review, the study derives the three propositions and eight hypotheses. We test these hypotheses for the publicly traded corporations in Taiwan considering a sample composed of the initial investment announcement. The period of the study was January 1991 to December 2000, and the firms selected were all companies listed on the Taiwan Security Exchange or the OTC. The number of the initial investment announcements initially collected was 365 for publicly traded corporations. Nevertheless, the final sample was reduced to 225 after excluding the confounding events and other screen. Finally, we separated the full sample to eight sub-samples by corporate type, ownership and investment scale and used the event study and regression analysis to test the information effect of the initial investment announcement. Our empirical results are not hold in the full sample since no significant market reaction, but are supported in the sub-samples grouped by high-tech. Moreover, the initial investment announcements of high-tech, high ownership, and high scale companies lead the significant positively abcdrmal rate of return at the announcement day. In addition, the investors significantly positively react to the initial investment announcements belong to the capital investment.