在無母數的迴歸當中，因為原始的函數類型未知，所以常用已知特定類型的函數來近似未知的函數，而spline函數也可以用來近似未知的函數，但是要估計spline函數就需要設定節點(knots)，越多的節點越能準確近似原始函數的內容，可是如果節點太多有較多的參數要估計, 就會變得比較不準確，所以選擇適合節點個數就變得很重要。 在本研究中，用交叉驗證的方式來尋找適合的節點個數， 考慮了幾種不同切割資料方式來決定訓練資料和測試資料, 並比較不同切割資料的方式下選擇節點的結果與函數估計的效果。 / In this thesis, I consider the problem of estimating an unknown regression function using spline approximation. Splines are piecewise polynomials jointed at knots. When using splines to approximate unknown functions, it is crucial to determine the number of knots and the knot locations. In this thesis, I determine the knot locations using least squares for given a given number of knots, and use cross-validation to find appropriate number of knots. I consider three methods to split the data into training data and testing data, and compare the estimation results.

租買選擇對於家戶來說是一件十分重要的家庭決策。當家戶決定購屋時，由於購屋支出是一筆極大的支出，勢必會影響到家戶各方面的支出，如食衣住行育樂等。。另一方面家戶需要多儲蓄以因應未來房屋的支出。 本文根據中國家庭金融調查(CHFS)的資料，將中國的家戶分為三種租買選擇分別為租屋、買屋有貸款、買屋無貸款。首先，我們先估計中國家戶的租買選擇，接著我們利用Heckman二階段模型估計租買選擇對於中國家戶儲蓄的影響。我們發現inverse Mill’s ratio的結果為顯著說明了在估計儲蓄函數是該考慮家戶的租買選擇。另外其他家戶特性的變數也有所變化，這些結果說明了如果不將租買選擇考慮進去 儲蓄行為的估計結果會有偏誤。 / Tenure choice is a very important decision for households. When the household decide to buy a house, it will have a heavily effect on the household’s spending, including food, clothes, transportation, education, and leisure, and so on. On the other hand, household have to save more money to meet housing expenses. Applying a data set from “China Household finance Survey” (CHFS) and separating Chinese households with three types of tenure choice, namely, renters, owner-occupied with loan payment, and owner-occupied without loan payment. First, we estimate the tenure choice of Chinese households. And then we use Heckman two-stage model to estimate how tenure choice affect household saving behavior. We find that the coefficient of inverse Mill’s ratio are significant. And the coefficient of household characters have some different from traditional model. The result show that if we don’t consider the tenure choice of household, the household’s saving behavior we estimate will be biased.

第一章緒論 第一節研究之動機與目的 第二節研究之範圍與假設 第三節研究之方法 第二章有關文獻之評述 第一節對肯博斯所著" 客觀考試之應用 "一文之評述 第二節對褚端司爾及史密德合著之" 各種改良式多項選擇測驗 "一文之評述 第三節對貝肯所著" 一種改銀式多項選擇測驗題在各種情境中之使用 " 一文之評述 第三章本研究測驗題之編製與實驗 第一節現行選擇測驗題之種類 第二節本測驗之編製 第三節本測驗之計分方法 第四節本測驗之實驗 第四章實驗所得資料之統計 第一節統計配對 第二節求各組測驗二與其標準測驗之相關 第三節變異數分析 第五章總結 第一節實驗結果之分析與比較 第二節對本研究之評述與討論 參考書目 附錄

我國勞工退休金新制自2005年7月1日開始實施，由過去的確定給付制改為具有確定提撥特色的「個人帳戶制」。對於勞工而言，確定提撥制和過去確定給付制不同之處在於承擔退休金投資風險的責任將由雇主轉由個人承擔。如何透過退休機制的設計以降低退休金的投資風險是近年來的重要議題，因此本文主要從個人偏好與風險的觀點探討保證投資收益率制度與制度轉換選擇權等兩個降低確定提撥制投資風險的重要配套措施。 在本論文的第一篇研究中發現，資產配置與國際投資對保證成本的影響頗大，在個人可選擇資產配置的情況下，高投資風險的資產選擇將造成政府未來龐大的或有負債。為了解決政府保證成本過高造成代內與代間的財富移轉，本文從使用者付費與個人效用的觀點探討保證投資收益率制度的設計，發現藉由設立保證投資收益上限可提升風險趨避者、損失趨避者與後悔趨避者等偏好下的預期效用，且能降低個人管理下方風險所需的提撥成本與退休計畫參加者所需繳交的保證費用，故建議政府可將投資收益率上限納入保證投資收益率制度，供退休計畫參加者選擇合於本身偏好的保證收益率上限。 在近來許多國家的公、民營退休體系由過去以確定給付制改為確定提撥制，為了降低在確定提撥制下的退休金投資風險，在美國的佛羅里達州之公務人員退休體系中，存在著可供個人選擇是否轉換到確定給付制的機制。在我國勞退新制中除了從過去的確定給付制改為確定提撥制外，亦輔以「年金保險制」供勞工選擇與轉換，若年金保險制具有確定給付制的特徵，則勞工等於是擁有一個從確定提撥轉換轉到確定給付制的選擇權，因此制度選擇權的探討對我國而言亦是相當地重要。在本論文的第二篇研究中發現，當風險趨避程度越高則轉換至確定給付制的機率越高，轉換到確定給付制的高峰期會出現在開始工作的初期與屆臨退休之際等兩段期間。隨著工作期間的延長，個人轉換到確定給付制的機率越低，但仍可有效地提升退休金的所得替代率與達到降低退休金下方風險的效果，在加入退休制度初期不得轉換的限制之後，會降低轉換到確定給付制的機率。 / The Labor Retirement Pension Act enacted in 2005 introduced defined contribution (DC) pension plan for substituting the traditional defined benefit (DB) pension plan. In the defined contribution pension plan, the investment risk is transferred to the participants. However, the design of rate of return guarantee makes the investment risk less severe for participants. In the first essay, we find that the asset allocation and foreign investment have large impact on the guarantee cost: the high risky investment may result in large potential liability of the government in the future if participants have the investment portfolio choice. This study develops a framework to analyze design of rate of return guarantee from the financial engineering and user paid principle view. We find that the cap of investment return guarantee not only increases the expected utility of risk aversion, loss aversion and regret aversion, but also decreases the contribution cost to participant associated with managing the downside risk. Around the world, the defined contribution (DC) plans have been the primary trend of pension reform in the both public and private sector. In an attempt to decrease the investment risk associated with DC plan, the public employees are provided with an option to buy back DB plan in the Florida State of U.S.A. In the second essay, we find that the higher level of risk aversion is, the higher probability to buy back DB plan is. During the employee’s early years of service and as the employees near retirement, the probability to exercise the option is the highest. The probability to exercise the option is decreasing with the years of service being increasing; the option also increase the pension replacement rate as well as decrease the downside risk of pension. The probability to exercise the option is lower, when the option to buy back the DB plan is prohibited during the employee’s early years of service.

我國金融市場積極發展多元化與國際化，將由台灣期貨交易所推出臺指選擇權，建立選擇權市場以完備金融市場。由於選擇權市場掛牌交易契約眾多與選擇權的交易特性，因此有賴造市者發揮提供流動性、風險移轉、價格效率性及價格資訊揭露等經濟功能。本研究以造市者之功能性逐項比較我國造市者制度與EUREX、SEHK、LIFFE等各大交易所造市者制度之差異，用以發現我國首度引進的造市者制度對於選擇權市場發展的影響。 本研究為瞭解選擇權造市者營運的知識與技術，整理造市者作業循環的流程，針對造市者業務進行分工，建立起造市風險分析架構。推導得出造市風險包含造市交易風險與部位管理風險，呈現出市場波性風險、報價及交易策略風險、造市交易之市場機制風險、市場行情變動風險、遇險交易策略風險、部位管理之市場機制風險及財務與作業風險等七大類的造市風險。 本研究亦提出對於我國選擇權市場未來發展方向的關切，以及建議重視造市者利基與市場運作之關係，並提出未來與造市者相關之研究建議。 / Taiwan's financial market has been developing constructively into a globalization and multiplicity market within these years. To create a more completed financial market, Taiwan Futures Exchange (TAIFEX) will issue "the Taiwan stock index option" recently. As the complex and various features of the options trading, the option market will be successful and efficient through the market maker's economic functions, which are adding liquidity, increasing price efficiency, transferring risk and proving price information. This survey tried to compare the different functions of market maker system with TAIFEX, EUREX, SEHK and LEFFE, and to disclose the influence of the new market maker system on Taiwan's option market. This exposure takes great interest in market maker's knowledge and technology. By summarizing and classifying the operation process cycle flow of market maker, this research further organizes an analytic structure in market maker's risks. Including the market making risks and the positions management risks, the discussion demonstrates the market change risk, pricing strategy risk, limitary quoting risk, market volatility risk, hedging risk, mechanism risk, and finance and operation risk. The consequent also furnishes the concerns about the future development of Taiwan option market, emphasizing the importance of the relationship between market makers and market' benefit, and the suggestions to the further research.

金融海嘯發生後，全球經濟在衍生性商品包裝下應聲受挫，反映過去主要國家中央銀行過度強調物價穩定，放任金融創新，對房貸疏於管制；因此當房價崩跌時，銀行的資產品質急遽惡化，銀行業因而嚴重受創，導致金融體系無法正常運作，衝擊實體經濟；失業率攀高，民眾付不出房貸、房屋被法拍，最後陷入金融與經濟不穩定相互影響的惡性循環。 影響房地產的因素眾多，可歸納為：（一）經濟因素，如所得、利率等；（二）社會文化因素，如人口成長率；（三）住宅條件，如交通的便捷性；（四）政治因素政局是否穩定，將影響社會大眾購置不動產的意願；（五）房屋本身特質，如坪數、建材、樓層等；及（六）政府政策因素，如貨幣政策、都市計畫管制，選擇性信用管制等。 由於經濟快速成長與國民所得激增，國內游資充裕物價高漲，帶動台灣房地於民國62至63年、68至69年、76至78年三次景氣循環達到高峰。也因此，民國78年2月28日，央行宣布實施選擇性信用管制，強制金融機構限定土地貸款成數、期限，首開房地產市場受金融管制之先例。 時至98年第2季金融風暴後，由於各國寬鬆貨幣政策陸續發酵，我國遺贈稅率調降至10%，及簽署金融監理合作備忘錄（MOU）、海峽兩岸經濟架構協議（ECFA）等兩岸經貿利多，吸引外資回流，陸資來台投資意願，再加上股市大漲創造的財富效果，推動資金行情加溫，帶動房市價量俱增。 民國99年，全球景氣持續復甦，我國出口及廠商大幅擴張，及全球需求增溫，推升原油等國際原物料行情，房價持續飆漲。央行遂於民國99年6月24日除調升重 貼現率、擔保放款融通利率及短期融通利率，並訂定「中央對金融機構辦理特定地區購屋貸款業務規定」，自6月25日實施，適用地區為新增放款過度集中在台北市及新北市10個縣轄市。 本研究針對本次特定地區信用管制措施，財政部一連串「打炒房」政策效果逐漸彰顯之際，擬藉由央行、金管會銀行局統計報表，分析銀行授信策略如放款科目、對象別及其他授信結構的差異、專家學者論述所集結論證研究，進行Difference in Difference方法的應用，探討房貸緊縮下公、民營銀行授信差異，提出適切論證以提供政府政策執行參考。 / The financial tsunami caused by US subprime mortgage crisis devastated the global economy and revealed the overemphasis of major central banks on price stability, overrated financial innovation, and the willful regulation of the mortgage. Hence, the quality of banks’ assets deteriorated rapidly and resulted in the breakdown of the financial system, leaving a long-lasting impact on the real economy. With unemployment rising, house owners lost their ability to sustain the over-priced mortgage. The end results are way too many foreclosed houses initiated a vicious cycle of financial and economic instability. Real estate markets are often affected by many factors which we summarize as follows: (1) Economic factors, such as income, interest rates, etc; (2) Social and cultural factors, such as population growth rate; (3) Housing conditions, such as the convenience of transportation; (4) The political factors, whether the political situation is stable would affect the willingness of communities in purchasing real property; (5) The characteristics of the properties, such as floor numbers, building materials and so on; and (6) The policies of government, such as monetary policy, urban planning control, discretional credit control, and etc. The thesis aims to study Taiwan’s real estate market with the above mentioned factors, in particular with focus on the effects of credit controls. The central bank declared the first credit control policy on February 28th, 1989 to regulate financial institutions in the forms of capped land loans and strict due dates. Until the second quarter of financial crisis in 2009 and owing to the quantity easing of many countries worldwide, Taiwan’s central bank again resorted to the tightening credit control policy tools in setting "central to financial institutions to handle specific areas housing loan business requirement" effective from June 25, 2010. Areas included Taipei City, New Taipei City, and other 10 cities in Taiwan were deemed as the housing bubble zones. This study uses data from the Statistical Reports of the central bank and FSC Banking Bureau to analyze the banks’ counter-credit-policy responses, such as lending subjects, objects, and other differences in credit structure. Difference-in-Difference approach is used to explore the differences of credit rationing between state-owned (or state-controlled) banks and private banks. Policy recommendation is provided in Chapter V in reminding the regulators to pay special attention to the non-universal effect of a universal credit control measure.

This business plan is written for a to-be-founded trade company which focuses on the trade between Taiwan and Switzerland. The environmental analysis (Chapter 4) shows that there are countless opportunities for trade between the two countries. The goal of the company is to find products and solutions with considerable potential. Therefore, this business plan does not focus on just one product; instead a process defines how the company selects and develops the products in its portfolio (Chapter 3). By making these products available in the other country, the company gives Taiwanese and Swiss manufacturers new business opportunities and gives the potential customers better and different solutions to their problems. The company’s goal is to build long-term partnerships with its suppliers in order to ensure a high level of quality (Chapter 5). The company will set up two logistic centers, one in Taiwan and one in Switzerland, which will be the hubs for standardized and efficient distribution processes (Chapter 7). In order to keep the costs low during the first years, marketing and sales (Chapters 8 and 9) will focus most of its activities on B2C products which are mainly sold through ecommerce activities. Finally, the financial forecasts (Chapter 10) show that there is a very high level of uncertainties for the financial success of the company. Therefore the cost structure of the company must be highly flexible to be adjustable to the company’s growth.

政治知識之於民主社會有其重要性，在政治學界中與政治知識相關的研究產出相當豐富，研究者利用政治知識此一變數進行相關研究前，對於題目如何選定、選項如何提供、題型的差異都是研究者需要去關注的重點，而本文主要的研究目的即是聚焦於討論何種題型更適合用來測量民眾的政治知識。 目前國內測量政治知識的問卷題型較為常見的為開放題與選擇題題型，在這兩類題型的討論上，前者被認為會低估受訪者政治知識程度，後者的測量結果則被質疑提供猜題空間導致高估了受訪者的政治知識程度，然而目前國內外卻缺乏足夠的實證研究來證明這兩個題型的適用性。 本文採用具有實驗設計性質的二手資料，利用前後測的方式讓受測者填答相同題目不同題型的問卷，藉此檢視各種知識程度的受測者在面對不同題型時是否會產生回應模式上的差異。本研究發現，開放題會使得較高政治知識程度的受訪者被低估，選擇題反而能準確測量出此類受訪者的知識程度。為了進一步證實受訪者在偏難的題目上所增加的猜題比例並非是來自於盲猜，本文採用多項機率單元模型來檢視受訪者於於選擇題選擇各個答項的機率。研究發現，儘管選擇題無法避免受訪者猜題，但受訪者並非是盲猜，反而會根據其具有的知識依據來答題，故政治知識程度高的受訪者能採用猜題方式答對題目，政治知識低的受訪者無法利用猜題方式猜中答案。整體而言，選擇題比起開放題更適合用來測量民眾的政治知識。 / Political knowledge plays an important role in the democratic society, and therefore there has been much research on political knowledge in the discipline of political science. To study political knowledge, political scientists have to understand the way of questions and options presented, and also the differences between a variety of question formats. This paper aims to analyze which question format is better for measuring the political knowledge of the public. The open-ended and multiple-choice items are both common formats for measuring political knowledge in Taiwan. The open-ended question is always considered to underestimate the respondents’ level of political knowledge, while the multiple-choice format is thought of overestimating the levels of political knowledge for providing the respondents with opportunity to guess. However, a strong evidence to decide the most suitable format for the measurement of political knowledge is still lacking. This paper uses the secondary data which is collected by a pretest-posttest questionnaire to examine whether guessing behavior will emerge or not when the respondents facing the same question with different formats. This research finds that open-ended questions underestimate the respondents’ knowledge levels who has higher level of political knowledge originally, but the multiple-choice questions can estimate the levels more accurately. To further confirm that the higher guess proportions in the more difficult questions are not resulted from the blind guessing, the study examines the probabilities of options selected by Multinomial Probit Model. The research finds that though the respondents may have guess more in multiple-choice question, however, they tend to answer the questions based on their knowledge instead of blind guessing. Therefore, the respondents who have higher levels of political knowledge can guess correctly, while those who have lower levels of political knowledge cannot. In summary, the multiple-choice questions are more suitable to measure people’s political knowledge.

為研究地主國租稅環境對台灣跨國企業海外子公司區位選擇的影響，本文以巢式羅吉特模型研究 2001 年至 2013 年間台灣上市櫃公司海外關係企業（排除金融保險業和 F 股）的資料。相對於過去探討海外子公司區位選擇的文獻僅著重於地主國稅率和地主國是否有反避稅制度（移轉訂價制度和資本弱化條款），本文首度加入租稅協定相關的變數：「地主國是否與台灣簽訂租稅協定」和「地主國租稅協定網絡」，探討兩者與海外子公司區位選擇的關係。 本研究實證結果發現地主國與台灣簽訂租稅協定並不能增加海外子公司在該地設立的機率，反而可能因租稅協定中的資訊交換條款而降低在該地設立子公司的機率；地主國租稅協定網絡愈豐富則是會增加跨國企業在該地設立海外子公司的機率，表示跨國企業可能藉由設立在租稅協定網絡豐富國家的海外子公司，適用該國的租稅協定利益。其他與地主國租稅環境相關變數的實證結果為潛在地主國稅率愈低及無反避稅制度的國家會較吸引海外子公司的設立。 / Using nested logit model, the thesis examined the impact of international taxation on location choice of Taiwanese multinational firms from 2001 to 2013. Compared to previous studies that only focused on the impact of tax rate and anti-avoidance legislations (transfer pricing rule and thin capitalization prvision), the thesis focused on the impact of tax treaty, examining the relationship between location choice and bilateral or multilateral tax treaties. The results showed that countries which had tax treaty with Taiwan did not attract Taiwanese multinational firms, and the possible reason was the exchange of information article in tax treaties. The study also found that the more tax treaties one potential host country had, the higher probability of Taiwanese multinational enterprises setting up subsidiaries in that host country, which suggested that Taiwanese multinational enterprises may have the behavior of treaty shopping. Other results related to tax variables showed that potential host countries with lower tax rate or with no anti-avoidance legislations would attract Taiwanese multinational firms.

由於一般投資人在買進或賣出選擇權時，並不會同時買進多個履約價的選擇權，故會造成選擇權隱含波動度的微笑曲線出現有不連續的現象。本文嘗試運用台指選擇權建構一個日內的隱含波動度微笑曲線交易策略，利用曲線配適的方法來捕捉瞬間時點下隱含波動度曲線發生不連續的現象，雖然最後出來的損益並不如預期但還是驗證了台指選擇權市場有多次這種不連續的機會且價格失衡的狀態會回歸正常。 / Option’s implied volatility smile curves discontinuous phenomenon exists when general investors buy or sell options, they won’t buy in every strike’s options. This paper attempts to use Taiwan Index Options (trading code: TXO) to construct a trading strategy based on the implied volatility. We use curve fitting method to capture volatility smile curve’s instant discontinuous. Although we find out that the strategy won’t make a profit, there were several times when TXO market’s implied volatility smile curves were discontinuous, and the market option price will eventually go back to the theoretical price.