自2007年爆發次級房貸風暴後，造成許多原本信用良好的企業出現財務危機，甚至倒閉，引發一連串公司間違約傳染的連鎖效應。因此，公司間或是產業間之榮枯是相互關聯的，且均會受到違約傳染和總體經濟因素的影響。另外，近年來新型態的信用衍生性商品逐漸發展起來，例如：遠期生效擔保債權憑證、擔保債權憑證選擇權等，這些商品與以往不同的地方，是與時間有著強烈的相關性，而以往評價常用的單因子相關模型，因其無法描述損失分配的期間結構，造成無法評價與時間高度相關的信用衍生性商品。是以，許多學者開始研究動態的信用違約模型，藉此描述損失分配的期間結構。因此，本研究結合違約傳染效果與動態信用違約模型，假設個別公司存活機率之對數轉換服從跳躍過程，並將資產池內的資產分為傳染公司與被傳染公司，發展出容易執行與從市場資料校準參數的動態違約傳染模型。之後利用市場上的資料校準模型參數，並說明次級房貸風暴發生前後，參數改變的結果。 結果反應出發生次級房貸風暴後，傳染效果與總體經濟因素的變數明顯的變大，且損失分配的重心明顯的向右方移動，與事實一致。除此之外，利用校準出的參數評價遠期生效擔保債權憑證與擔保債權憑證選擇權，並分析參數改變對於擔保債權憑證選擇權的變化，發現傳染效果和跳躍高度與擔保債權憑證選擇權是反向關係，而回復率與擔保債權憑證選擇權是正向關係。最後計算擔保債權憑證選擇權之避險參數，提供避險者避險決策的依據。

放款的利息收入雖是商業銀行主要之獲利來源, 但借貸行為卻同時使得銀行承受著違約風險。銀行應透過風險管理方法, 計算經濟資本以提列足夠準備來防範預期以及未預期損失。 另外, 若銀行忽略違約行為之間的相關性, 將有可能低估損失的嚴重性。因此, 為了在考量違約相關性下提列經濟資本, 本文由 Merton (1974) 模型出發, 以信用風險因子模型判定放款對象是否違約, 進而決定銀行面對的整體損失為何。 為簡化分析, 本文假設違約損失率 (loss given default) 為 100%。 再者, 為加強相關性, 本文亦將違約傳染性加入因子模型並比較有無傳染性效果時, 模型所計算出的損失孰輕孰重。 而在決定違約與否時, 須利用來自移轉矩陣上的無條件違約機率, 然信評機構所發布之移轉矩陣概遺漏諸多訊息, 依此, 本文以多期連續的移轉矩陣修正之並得到另一不同的無條件違約機率。 最後, 以臺灣的 537 家上市櫃公司作為資產組合, 經由蒙地卡羅模擬得到兩個因子模型的損失分配, 我們發現具有傳染性效果存在時, 預期損失和非預期損失較大且損失分配也較為右偏。 / Despite interest income from loans is a major profit contributor for commercial banks, lending inevitably makes banks bear default risks. For the sake of avoiding expected and unexpected losses, risk management methods ough to be employed by banks to meet the ecomical capital requirement. Besides, loan loss severity may very well be underestimated if the correlation between default events is disregarded. Therefore, in order to calculate economical capital when taking default correlation into account, we start from Merton (1974) model, and identify if loans will be in default via facor models for portfolio credit risk and portfolio losses can then be detemined. To simplify our analysis in this paper, loss given default is assumed to be 100%. To intensify correlation, default contagion is, moreover, introduced to our factor model and we investigate which model results in larger losses as well. When determining default, we have to utilize rating transition matrices to obtain unconditional probability of default. Transition matrices published by credit rating agencies, however, have embedded drawback of insufficient information. We correct this flaw by means of another transition matrix based on continuous-time observations and produce different unconditional probability of default. Through Monte Carlo simulation, loss distributions are calibrated respectively from the two factor models under portfolio of 537 Taiwan listed and OTC companies. We find that expected and unexpected losses are larger and loss distribution is more right-skewed when infectious effects exsit.

本研究探討雷曼兄弟倒閉事件對美國金融機構股價報酬率所帶來的傳染效果。我們的研究結果顯示，持有雷曼兄弟股份的金融機構受到較顯著的影響，而且在股東當中以商業銀行及投資銀行所受到之衝擊最為顯著，而投資銀行受影響之程度又高於商業銀行。我們也發現，金融機構對雷曼兄弟的持股比例愈高，其股價受到雷曼倒閉危機影響之程度亦愈高。此外，一些規模較小的金融機構在一些事件中亦呈現了顯著的反應，顯示在某種程度上，雷曼兄弟危機事件在金融產業當中引發了傳染效果（Contagion effect）。我們的實證結果也顯示，美國政府此次未介入救援雷曼兄弟的做法，被市場解讀為金融機構不再是「太大而不能倒（too-big-to-fail）」。 / We examine the contagion effect of Lehman Brother’s bankruptcy on US financial institutions’ stock returns. Our results show that financial institutions which held Lehman’s shares were affected more significantly. Furthermore, within Lehman’s shareholders, commercial banks and investment banks were affected most significantly, and impacts on investment banks were more significantly than commercial banks. We also find that the higher financial institution’s ownership percentage of Lehman was, the more its stock price was affected. Besides, some smaller financial institutions were also influenced significantly in some events, to some extent, implying a contagion effect in the financial sector. Our empirical results also indicate that the way that US government not to rescue Lehman Brother was perceived by the market that no financial institutions are too-big-to-fail.

本研究旨在說明資本適足率對於銀行業資訊傳遞效果之影響，利用Allen and Gale (2000)模型討論在不完整市場結構下，銀行間因為持有銀行同業存款而形成相連的傳染途徑，進而影響整個系統，本研究擴展Allen and Gale (2000)的模型，加入資本適足率的考量，從而進一步探討透過資本適足要求能否有效提高銀行整體穩定性。 模型假設因為不同區域對於早、晚期消費需求不同，可藉由區域間的資源移轉，來達到最適分配情況。隨著資本適足率的納入，將改變最適分配解，同時分析緩衝(buffer)、擴散效果(spillover effect)及傳染(contagion)的變化。文中傳染定義為擴散效果扣除緩衝力道的淨結果，並說明若有超額流動性消費需求衝擊時，一家銀行的倒閉將如何傳染至整個銀行體系。 此研究發現，在資本適足規定下，若長期資產報酬率越大，會更有機會取得較大的緩衝能力，但將面對較大的擴散效果。關於傳染現象，則是發覺當銀行同業存款越小，在資本適足規定下的傳染機會越低；若長期資產的早期報酬率越大，同樣可降低發生傳染現象機率，即驗證資本適足率對於銀行穩定性的貢獻。 / The objective of this study is to testify the effect of capital requirement with regard to information transmission among banks. We develop a model based on Allen and Gale (2000) to discuss that under incomplete market structure, contagion channel is built because of interbank deposits market. We also expand Allen and Gale’s model by putting new parameter, capital requirement, into this model to analyze the impact of capital requirement with respect to stability in banking system. Due to different liquidity demands at each date in different regions, banks can exchange resources in the system to reach the first-best allocation. With capital requirement, the first-best allocation varies and so does buffer, spillover effect and contagion. In this article, contagion is defined as the net result of spillover effect minus buffer. Besides, we explain how the bankruptcy in one region evolves into the bankruptcy in the whole system under excess demand for liquidity. We find out that with capital requirement, if return of long-term asset at final date is higher, there will be more chances to have more buffers but larger spillover effect. As for contagion, it shows that with lower interbank deposits or higher return of long-term asset at early date, the possibility of contagion will be reduced. As a result, we can conclude that capital requirement really improves the stability in banking system.

無論在電腦動畫、電玩或電影產業，利用電腦自動產生虛擬人群已逐漸成為不可或缺的要素之一。這些虛擬人群，往往是系統先賦與每個虛擬代理人(agent)基礎智能，然後藉由個體之間的互動法則所自動產生。然而，過去因為普遍未考量真實群體情境中的傳播與互動模式，使得虛擬人群所表現的群體行為與現實情況仍有些差距。因此，我們引用社會心理學文獻，建立一個具有溝通機制的人群模擬平台(IMCrowd)，以期自動產生與現實群眾動態更相似的模擬人群。IMCrowd是多代理人(Multi-agent)基礎的系統，其中每個虛擬代理人都具有區域的感知範圍與自主能力，因此他們能夠自動地與環境中的其它物件互動與反應。由於我們為IMCrowd所建立的溝通模型考量了社會心理學的理論，因此虛擬人群能浮現真實群體動態中的社會互動模式，如情緒傳染與從眾效應。本研究以IMCrowd執行了多種情境下群眾暴動與群眾控制的模擬，藉此展現本系統的應用將不僅可提升群體模擬的真實度，亦可做為社會心理學家研究群體行為的工具。 / Using computer to automatically generate simulated crowd has become a trend in animation, computer game, and film productions. Many of these works were produced by modeling the intelligence of the agents in a crowd and their interactions with other nearby agents and the environment. However, the perceived facts or elicited emotions usually do not propagate in the crowd as they should in the real life. In this work we attempt to build up a communication model to simulate a large variety of crowd behaviors including the course of crowd formation. The proposed crowd simulation system, IMCrowd, has been implemented with a multi-agent system in which each agent has a local perception and autonomous abilities to improvise their actions. The algorithms used in our communication model in IMCrowd are based heavily on sociology research. Therefore, the collective behaviors will emerge out of the social process such as emotion contagion and conformity effect among individual agents. Several elaborate riot simulations and riot control simulations are demonstrated and reported in this thesis as the application examples of IMCrowd. Thus, we claim that IMCrowd may not only benefit on enhancing realism of crowd animation but also be useful in studying crowd behaviors such as panic, gathering, and riots.

本研究使用美國、德國、英國股票指數期貨之日內交易資料，從報酬率中分離出連續波動度與跳躍項，再以MLE法估計Hawkes相互激勵過程之參數，衡量跨市場跳躍傳染現象。擴展文獻中僅兩市場的分析至三市場模型，更能從整體的角度解釋市場間的關係及跳躍傳染途徑。實證結果顯示，美國能直接影響其他市場，而其他市場反過來不易干涉美國，呈現非對稱影響效果。歐洲兩國能互相傳染，英國對德國的影響較大，也更有能力影響美國，稱英國為歐洲的影響輸出國，德國為歐洲的影響輸入國。

【研究緣起與問題】2003年春季，SARS在全球造成一陣轟動，同時也重創全球的經濟，如今SARS會不會再來？政府如何因應？政府如何監測疫情？如何防治疫情？如果SARS等新興及再浮現感染症 （Emerging and Re-emerging Infectious Diseases）出現時，負責全國傳染病防治任務的行政院衛生署疾病管制局是如何作好全國傳染病防治工作呢？中央及地方主管機關之溝通，如何共同消除指責與誤會，發揮團隊精神，如何共同作好全國傳染病防疫任務，爭取公眾對政府之瞭解並加以支持，同時防止任何可能之攻擊，才是全民之福祉。故以「全國傳染病通報系統」的績效，其所仰賴之溝通通路之問題研究，是本研究的動機。了解「傳染病通報管理系統(WEB版)」最大的問題及新興傳染病（如SARS）防治通報的問題及「傳染病通報管理系統」最需要改進之的地方，是本研究的目的。 【研究方法與資料】本研究採取量化與質化並重的方法，以網路問卷方式的進行問卷調查，問卷對象為全國使用傳染病通報管理系統之相關人員，問卷問題內容的設計主要的型式含有開放式問卷（Open –Ended Question）、封閉式問卷（Close-Ended Question）及混合型問卷（Mixed Question）三種。同時以九十二年針對新興傳染病SARS期間為例，深度訪談當時任職於中央與地方負責傳染病防治權責具代表性者五位主管進行訪談，瞭解受訪者對新興傳染病如SARS防治的問題及「傳染病通報管理系統」最需要改進之的地方，以彌補問卷調查法無法深入瞭解到問題之缺點。 【研究結論與建議】 一、根據筆者初步研究經問卷的實證調查之後，從問卷分析資料顯示，使用者使用電腦的時間愈久，對以新版為溝通工具滿意度愈低；而愈瞭解通報系統的使用者，其以新版為溝通工具滿意度愈高，愈滿意新版改進之電腦網路溝通效果。顯示新版的通報系統在設計的操作介面與程序上，沒有人性化，作業內容繁多、需輸入的項目、資料太繁瑣、版面太繁瑣等，反而徒增使用者的困擾；而愈瞭解通報系統的使用者，其對新版的滿意度越高，表示對於使用者在操作上的教育訓練仍須加強，才能發揮新版通報系統的成效。 二、從敘述統計分析中發現： 受訪者對「訊息異動通知作業流程滿意」的使用滿意度的平均數最高，平均數為2.8095表示：較為滿意。「通報作業流程滿意嗎」的使用滿意度的平均數最低，平均數為2.3934，受訪者表示：較為滿意。以新版為溝通工具整體溝通滿意情形，平均數為2.558滿意。 三、開放式問卷分析資料顯示，花太多時間、速度太慢、網路的頻寬過窄又不穩定或無法登錄、作業內容繁多、作業流程問題、資訊管理問題及功能增添與建議等部分。 四、根據筆者初步研究深度訪談之後，彙整中央主管機關及地方主管機關的訪談核心二部分的問題與建議。 五、研究建議： 綜合問卷調查及深度訪談結果，提出政策層面、組織溝通層面、管理層面、實務操作層面等之建議。 【關鍵字】傳染病通報系統、嚴重急性呼吸道症候群(severe acute respiratory syndrome, SARS)、組織溝通、溝通通路、行政院衛生署疾病管制局 / 【Background and Issues】 In the spring of 2003, SARS outbreaks devastated the whole world; they also heavily damaged the world economy. Will SARS come back? How can governments manage it? How can government monitor the epidemics? How can we control the infection? When emerging and re-emerging infectious diseases such as SARS appear, how can the Center for Disease Control of the Department of Health, an organization in charge of national disease control, plan and execute measures for the control of communicable diseases? Effective communication between the central and the local competent authorities to remove together blames and misunderstanding, to develop team spirit, to attain jointly the goal of disease control, to solicit understanding and support of the public to the government, and to prevent any likely attack are some of the considerations in improving the welfare of the public. The reason of the present study was to understand, through the achievements thus far of the National Communicable Disease Reporting System, issues involved in the communication channels. The purposes of the study were to understand the major problems of the Communicable Disease Reporting and Management System (the Web version), issues involved in the reporting of emerging infectious diseases, and areas of the Communicable Disease Reporting and Management System where improvement was most urgently needed. 【Method and Materials】 Both the quantitative and qualitative methods were used. Questionnaire interview was conducted through the web to persons using the Communicable Disease Reporting and Management System throughout the country. The questionnaire contained open-end questions, closed-end questions and questions of mixed types. At the same time, to understand in depth areas not properly covered by the questionnaire interview, five key persons responsible for disease control at the time of the SARS outbreaks in 2003 at the central and the local levels were interviewed to understand their attitudes toward the control of emerging infectious diseases such as SARS, and the improvement most urgently needed for the Communicable Disease Reporting and Management System. 【Results and Recommendations】 1. Preliminary findings from analysis of the questionnaire interview showed that users became less satisfied with the new version communication means the longer they were in use of computers. Users who knew more about the use of the reporting system were more satisfied with the new communication means, were at the same time, more satisfied with the communication effects of the improved version. These facts suggested that the new reporting system was less user-friendly in the designing of interface and procedures. The system required more work, more detailed inputs, and thus added additional burdens on the part of the users. The more the users understood the reporting system, their satisfaction was higher, indicating that, for the new version reporting system to function in full, training of the users in the use of the system should be strengthened. 2. From the descriptive analysis, it was found that the average score of user satisfaction on the “information change reporting procedures” was as high as 2.8095, suggesting that the users were fairly satisfied. The satisfaction score on “reporting procedures” was the lowest at 2.3934. The overall satisfaction score on the use of the new version as a communication means was 2.558. 3. Findings from the open-end questions gave the following recommendations: taking too much time, too slow, frequency of the web too narrow, unstable, and unable to load, too detailed, issues related to operational procedures, issues related to information management and additional functions, etc. 4. By analysis of the in-depth interview, some issues were raised and recommendations made to the central and the local competent authorities. 5. Through questionnaire survey and in-depth interview, some recommendations concerning policies, organizational communication, management and practical operation were made. 【Key Words】 Communicable Disease Reporting System, SARS (Severe Acute Respiratory Syndrome), organizational communication, communication channels, Center for Disease Control, the Department of Health

本論文運用史學、統計與GIS繪圖法，自表至裏，自應用至體制，再至緣由，論述1895～1945年50年間臺灣預防接種史之開展、歷程與引發原因。論證臺灣人用疫苗發展歷史超過百年；日治時期臺灣住民已廣泛接種多量疫苗；臺灣諸免疫界變革不可忽視日本人事體制。 全文首先以法定之傳染病為例，論證疫苗在臺灣民間應用之普遍。首先論述最早在臺灣實施，作為全民接種開始，且採皮上切種之天花種痘，其官方制度演變、於民間推展方法、技術變革與臺日差距（少），以及從統計面所見之臺灣高接種率（社會高免疫力）、低天花感染率和死亡率，和1900年代一遇疫情風聲，即立刻實行臨時接種或擴大定期接種之防疫定制。其次論述血清型疾病預防注射。如1900年鼠疫接種，是臺灣最早由官方執行較大規模針式＼侵入性皮下預防注射之始。但1904年後因花費金額高、人體副作用大而少在臺灣實施。再如霍亂，1902年臺灣已應用霍亂血清，1916年較大規模實施＼實驗，1919～1920年國際霍亂大流行更極力推廣，兩年間，每年各有百餘萬人接種，是各地臺灣人普遍認識和接種預防注射針之始。1918～1920年流行性感冒，確切菌種和傳染原因等等均未知；但當第二波疫情較集中且嚴重時，官方已鼓勵配合副致死病因如肺炎球菌等，實施預防注射，與今日流感疫苗類似。其後，對霍亂或流感使用預防接種之防疫法，終日治結束均為常例。又如流行性腦脊髓膜炎，自1917年臺灣出現購買疫苗案例後，該疫苗自1920年代起即大量應用，1930年代後之接種人數更屢以十萬或百萬為單位。而臺灣1935年前所謂流行性腦炎或腦炎，常是指流行性腦脊髓膜炎；流行性腦炎或腦炎預防接種，實際指流行性腦脊髓膜炎接種。以上各種疫苗大量應用，至約1920年代已形成官方一遇疫情即立刻接種地積極、強力防疫定制，亦影響臺灣該等傳染病感染率和患者死亡率，以及官民對上述傳染病恐懼態度轉變。 繼應用面，本文次論人事、體制之制度面發展。首先，臺灣在1902～1905年，先後出現在臺灣設置血清藥院聲音、鼓勵細菌血清學研究、設置牛疫血清作業所、起議在臺實施全新生兒種痘和創設中央科學研究所。1916年開始製販血清疫苗，擇製本島常見菌種、強調「賣捌」非「販賣」，此後製品種類陸續增加。臺製疫苗亦供應沖繩和中國等等臺灣以外地區。1922年起，臺灣不再製造牛疫血清，改由朝鮮提供防疫之常備用品；1939年士林廠房完工，臺灣開始製造乾燥和精緻血清疫苗。另就中央衛生部人事和業績比較，細菌血清類研究和專任人力常多居各研究單位之首，1920年代達高峰；1930年代研究比重雖看似降低，然其負責人員數量、製劑內容和販賣數量，卻更深刻且快速進展，故臺灣中央一直重視免疫醫學或細菌學研究。而使臺灣於各階段發生細菌學興起、開始製販血清疫苗、研製機構體制和製品製程改變原因，首先，使細菌學、免疫學和血清研製機構在臺展開，首須歸因後藤新平和高木友枝等臺灣首長個人意識，以及其背後穩固之內務省衛生局（行政）、傳染病研究所（技術）和眾議會（審議）等人事資源和總體意向，並因人際網絡使臺灣得參酌美國作法。1916年臺灣開始製販血清疫苗，源自1914年傳研移管；擇製某類型製劑且限定專賣，則為前期人事派系延續和檢定制度便宜之計。傳承人事脈絡與社會局勢，再配合新發地政治社會需求，使1920年代後期起中央研究所人事漸變化，來自地方者再回歸地方，中央日益學術、大學化。最極致表現是1936～1939年中央研究所成為臺北帝大附屬和東大傳研化。此外，臺灣與海外各血清疫苗單位有連絡互通；臺灣是日本諸外地中最早大量製販各種血清疫苗者；也是日本國境內，極少數可以帝大附屬研究所之姿製販血清疫苗者。 / The dissertation aims to use historical data a long with GIS methods to explore the vaccination prevention to certain infectious diseases in colonial Taiwan (1895 – 1945). The smallpox vaccination was the earliest vaccination has been practiced in Taiwan. The change of polices and influstructure in cowpox manufacture revealed how the vaccination adminstration transformed and shaped local society. With the cowpox vaccine popularized, the immunity was gained as well as the low infection and mortality rate after 1910s implied its effectiveness. Secondly, the plague vaccination in 1900~1904 was the first big-scale preventive injection by the Taiwan official, but rarely injected after 1904 due to the high cost and side effects. following that, the cholera serum was often used in scale in 1916 which 3 million people received injecttion during 1919~1920, a period of cholera pandemic. One effective that cholera imunizationleft was that Taiwanese strated to accept needle injection a way of vaccinated prevention. The pandemic influenza invaded Taiwan in 1918~1921but the etiology was unknown currently. As the second wave of the epidemic occurred, the authorities began to pay attention and encouraged people to take injection from indirect causes of death such as pneumococcal. This injection strategy was applied in the whole Japanese ruled era. The epidemic cerebrospinal meningitis, since its vaccination reported in 1917, was larger implemented in the 1920s, then greatly practiced after the 1930s. Furthermore, the epidemic encephalitis or encephalitis named after Taiwan before 1935 usually referred to the epidemic cerebrospinal meningitis. Therefore, the epidemic encephalitis or encephalitis vaccination actually meant the meningococcal meningitis vaccination. In general, most implementation of vaccination appeared around 1916 progressed in 1920s and prevalently practiced after the 1930s. The police of vaccination brought down the infection and mortality rate in Taiwan a long withthe attitude of acceptance to immunization injection among people. Furthermore, Taiwan set up a professional animal serum factory and established the Central Science Institute of Taiwan Governor in 1903. In 1916, Taiwan started to make and sell serums and vaccines to the public which indicated the vaccination were already prevalent among Taiwanese. The govenemnt monopolized the sale of vaccines and the products could however offer to China and other places. In 1939, Taiwan went a step further to manufacture the dried vaccine, toxoid, and anatoxin. The phenomenon above implied various meanings. First, the promotion of bacteriology, immunology and serum in Taiwan were based on personal commitment by Gotou Shinpei (1857-1929) and Takagi Tomoe (1858-1943). With their connection to the Bureau of Sanitation of the Ministry of Interior, the Institute of Infectious Diseases (Kitazato Shibasaburou, 1853-1931), and House of Representatives, their works might link to the Rockefeller Institute which brought strong trust to the developing of bacteriology. In 1916 the attempt to produce and sell vaccine was a side-effect deriving from the domentic friction of the Institute of Infectious Diseases in 1914. It caused by the struggle of the administrative and resources, the dispute of practical and academic, and maturates of the bacteriology; lead to the essentially open to the serums and vaccines market in Japan and its foreign lands, included Taiwan. Taiwan closed to the faction of Kitazato. Soon after, the great spread of infectious diseases in 1918-1921 bright the emergency to apply and manufacture of biologics in 1920s in Japan and all of its territories. Taken the impact, the Central Research Institute of Taiwan Governor changed its infrastructure and products in the 1930s. It is worth noting that the Central Research Institute of Taiwan Governor and the Taiwan Medical Specialized School became branches of the (Taipei and Tokyo) Imperial University in 1936-1939, for the necessity of dried vaccine, toxoid, and anatoxin.