資料串流探勘是一個重要的研究領域，因為在現實中有許多重要的資料以串流的形式產生或被收集，金融市場的資料常常是一種資料串流，而通常這類型資料的本質是變動性大的。在這篇論文中我們運應了資料串流探勘的技術去預測台灣加權指數期貨的漲跌。對機器而言，預測期貨這種資料串流並不容易，而困難度跟概念飄移的種類與程度或頻率有關。概念飄移表示資料的潛在分布改變，這造成預測的準確率會急遽下降，因此我們專注在如何處理概念飄移。首先我們根據實驗的結果推測台灣加權指數期貨可能存在高頻率的概念飄移。另外實驗結果指出，使用偵測概念飄移的演算法可以大幅改善預測的準確率，甚至對於原本表現不好的演算法都能有顯著的改善。在這篇論文中我們亦整理出專門處理各類概念飄移的演算法。此外，我們提出了一個多分類器演算法，有助於偵測「重複發生」類別的概念飄移。該演算法相比改進之前，其最大的特色在於不需要使用者設定每個子分類器的樣本數，而該樣本數是影響演算法的關鍵之一。 / Data stream mining is an important research field, because data is usually generated and collected in a form of a stream in many cases in the real world. Financial market data is such an example. It is intrinsically dynamic and usually generated in a sequential manner. In this thesis, we apply data stream mining techniques to the prediction of Taiwan Stock Exchange Capitalization Weighted Stock Index Futures or TAIEX Futures. Our goal is to predict the rising or falling of the futures. The prediction is difficult and the difficulty is associated with concept drift, which indicates changes in the underlying data distribution. Therefore, we focus on concept drift handling. We first show that concept drift occurs frequently in the TAIEX Futures data by referring to the results from an empirical study. In addition, the results indicate that a concept drift detection method can improve the accuracy of the prediction even when it is used with a data stream mining algorithm that does not perform well. Next, we explore methods that can help us identify the types of concept drift. The experimental results indicate that sudden and reoccurring concept drift exist in the TAIEX Futures data. Moreover, we propose an ensemble based algorithm for reoccurring concept drift. The most characteristic feature of the proposed algorithm is that it can adaptively determine the chunk size, which is an important parameter for other concept drift handling algorithms.

2005年爆發台灣首宗財經記者勾結金管會高層的股市禿鷹案，記者涉嫌非法以職務之便報導股市消息，從中牟利千萬元，不但震撼記者圈，也讓媒體機構驚覺從此必須要加以記者規範利益迴避；隔年又發生財經記者涉嫌勒索台灣首富事件，突顯專職報導上市櫃企業動態與企業主的財經記者，所面臨的利益衝突問題。 本研究深度訪談14名不同媒體與年資的記者，發現記者的利益衝突可區分為餽贈、兼職、買賣股票、轉職壓力、及媒體機構與大股東的商業與政治利益等五種。不同的利益衝突帶來不同對新聞報導的影響，但普遍而言，利益衝突影響新聞方向的程度，依序為媒體利益（包含置入性行銷）、買賣股票、轉職壓力、餽贈、兼職。 研究發現，台灣的新聞採訪「送禮」的陋習由來已久，利益衝突是財經記者在實務上不能避免的問題，恪遵利益迴避原則可能影響記者的新聞表現。本研究也發現，媒體機構訂定具體明確的規範，記者越不能刻意利用職務去謀取私利。 訪談發現，「企業出錢、記者請假、媒體刊登」成為企業招待記者出訪，以圖撰寫報導的固定模式；不少受訪者被迫以記者之名撰寫置入行銷稿，而感到無奈與憤怒。記者待遇不佳、成為貪圖業外收益的主因。 證交法2005年修法之後，把記者納入內線交易的規範範圍，最重可處三年以上十年以下有期徒刑的重罪，一半以上受訪者仍然不清楚其採訪工作可能帶來觸法的風險，勾勒出台灣財經記者缺乏法制觀念的現象。

本論文在分析 SP500 指數其中交易最為頻繁的 345 家公司在 1996年各月份的股票數據,市場模式的空間特性已經被證實出來[3][4][5],而我們利用卡忽南 -拉維展開來分解股價對數報酬的時間序列, 利用傅立葉分析,並考慮股價對數報酬是否有時間序列重疊,與比較高頻移動平均對時間序列的影響,藉由參考各股市系統的特徵參數來尋找相似或不同之處。 / We present the results of our analysis of time series for a collection of 345 stocks listed in S&P 500, to show that integrated information on collective fluctuations in financial data can be revealed quantitatively by combined analysis, focusing separately on either the deterministic or the stochastic contents of the system. In comparing the fluctuations of high frequency one-day moving averages (HF1MA) of the original prices of individual stocks with those inherited in the trajectories of Brownian particles [1], also comparing the log return with overlapping time interval with the log return without overlapping time interval, we can quantify the time characteristic properties of the whole system which would direct the motions of tracer particles. In this study, we decompose the fluctuations in Karhunan-Loeve expansions and reveal the system-specific collective properties by analyzing those collective modes in their time-wise as well as the stock-wise bases, obtained for either the original prices or those of HF1MA, and for the log return with or without overlapping time interval.

本研究主旨有三：(一)整理現有股市財經節目的營運方式、節目內容並瞭解財經節目主持人及其口語傳播技巧。(二)從股市財經節目觀眾收視行為來觀察，以節目收視率及收視觀眾輪廓(audience profile)為基礎，觀察股市財經觀眾其日常生活中收看股市財經節目的相關行為。(三)結合對財經節目與觀眾收視行為的觀察與研究，找出股市財經節目的關鍵要素並提出建議。 本研究以文本分析、深度訪談、二手資料整理等研究方法進行，分析整理後，發現股市財經節目觀眾的收視動機非常明確，為主動閱聽人，收看節目係屬工具性行為，主要目的係希望投資獲利。因此，節目設計應先考量以下兩點：(1)股市財經節目的觀眾特性：男女比例相當、年齡層較高、學歷較高、居住北部地區較多、家庭所得較高、對資訊需求差異大、對節目是否應教育功能意見分歧。(2)股市特性：資訊瞬變、影響股市漲跌因素甚多、需專家解說。 而股市財經節目雖與其他節目有明確區隔，唯同類型的電視投顧節目或股市貼盤節目本身並無差異化，造成電視投顧節目以消耗投顧老師的方式進行經營，四個股市貼盤節目僅有二個尚稱成功，本研究建議節目產製單位應以更深層、更能吸引觀眾的相關投資資訊來提升閱聽眾對節目的忠誠度，而同類型節目之間則應以觀眾對資訊偏好的需求、對節目教育功能的不同的認知來進行差異化。 / This analysis will 1) examine the current format for Stock Market Analysis Television Programs, the program content, the program hosts types, and the host speaking styles. 2) Examine program viewer behavior, based on viewer ratings, audience profile, and the role that these programs play in the daily lives of its viewers. 3) Ascertain the key elements of these programs and provide recommendations based on conclusions reached through cumulative analysis of such television programs and the behavior of their viewers. Based on text analysis, in-depth interviews, and second-hand data, viewers appear to have clear utilitarian motivations for watching these television programs, they are active listeners, and their primary objective is to profit on their stock market investments. Program design should take into consideration the following elements: 1) Viewer Demographics: even proportion of male and female viewers, higher age-bracket, highly educated, living primarily in Northern Taiwan, high household incomes, strong demand for new information, split on whether such programs provide educational value. 2) The Nature of the Domestic Stock Market: information changes quickly, factors affecting the rise and fall of stock prices are many, professional market insight and analysis is needed. Stock Market Analysis Television Programs, which are quite different from other television programs, fall into one of two main types: Investment Consulting Company-produced Analysis Programs, Real-time Market Data and Analysis Programs, with very little variation between competing programs in each category. The former generally relies on a string of new program hosts in an attempt to maintain viewer interest. There are currently only four of the latter type of program currently running with only two of them showing even mild success. This report will recommend that producers of such programs should offer more in-depth content so as to better attract and maintain viewers. Competing programs should also distinguish themselves from one another by offering different types of data and analysis content, and different levels of educational content in their programs.

不同股市的報酬關聯隨時間動態改變，本研究欲了解近期美國、台灣與亞太地區的中國大陸、香港、日本及韓國的報酬連動關係，並進一步觀察次級房貸風暴期間美股對這些地區的關聯改變趨勢。本論文採用灰色理論與時間序列兩種方法，實證發現次級房貸風暴發生期間，台股及亞太地區主要指數不論在報酬率或是報酬率波動性受美股影響的程度大多增強。 實證結果顯示，在風暴期間的報酬率傳導關係，亞太以韓國影響台股最顯著，美股則全面影響亞太指數；在報酬率波動性溢傳上，亞太以日本、美股以道瓊工業影響台股最強，台股則是電子類股被美股影響最重，但營建類股在與美股或是亞太指數的關聯趨勢變化卻最明顯。另外，灰關聯分析對時間序列檢定的關聯組合可以提供互補的關聯強弱關係說明，且具有相當的正確性。 / Connections between stock markets are dynamically changing, and it affects investor's transnational investment portfolio. We focus on the relationships of stock markets among the United States, Taiwan, Japan, Korea, China and Hong Kong, and eager to understand the connection tendency between Untied States and Asian-Pacific area during the subprime mortgage crisis period. The identified research methods are time series and grey theory, including Granger causality test, GARCH model and grey relational analysis. We find out the returns and volatility in Asian-Pacific stock markets were all affected increasly by U.S. market during the subprime mortgage crisis. The main empirical results are as follows: In the relationships of returns, Korea affects Taiwan mostly in the Asian-Pacific area, and U.S. market affects all the others entirely during the subprime crisis. In the relationships of volatility, Japan and Dow Jones index affects Taiwan deeply during the period; within all the Taiwan indexs, Electronic Sector Index was affected by the U.S. market mostly than others during the same period, but the connection tendency in the Construction Sector Index with other markets changes more obviously. Otherwise, grey relational analysis can provide complementary explainations as compared to time sereies in the strength of relationships, and the explainations are with plenty credibility.

本文旨在研究地下經濟規模與金融發展對實質人均GDP成長率之影響。我們使用二種不同金融發展資料，分別為私部門信貸規模及股市成交值規模。使用的資料型式為追蹤資料，時間由西元1981年至2008年，國家共22個，其中有14個已開發國家，8個開發中國家。模型部分共採用三種型式，首先為一般線性模型，其次為納入地下經濟規模與金融發展交互作用之非線性模型，最後則是門檻迴歸模型。 結果顯示，在一般線性模型中，地下經濟規模及私部門信貸規模對實質人均GDP成長率的影響並不顯著，而股市成交值規模對實質人均GDP成長率有顯著正面影響。 而在納入地下經濟規模與金融發展之交互作用項後，私部門信貸規模與地下經濟規模的交互作用項對實質人均GDP成長率有顯著負面影響，而股市成交值規模與地下經濟規模之交互作用項對實質人均GDP成長率有負面影響，但不顯著。由上面的結果，我們推斷地下經濟規模與實質人均GDP成長率可能存在著非線性的關係。 在門檻迴歸中，若把顯著水準設在10%，我們發現納入私部門信貸規模與股市成交值規模的模型存在顯著的門檻效果。在門檻迴歸中，若把私部門信貸規模或是股市成交值規模當作門檻變數，當金融發展程度低時，地下經濟規模對實質人均GDP成長率有顯著正面影響，而當金融發展程度高時，地下經濟規模對實質人均GDP成長率的影響並不顯著。 總結來說，我們建議低度金融發展國家適度的允許地下部門發展，因為政府無法干預地下部門，因此地下部門的營運較有效率，適度允許反而有助於提高資本投資的效率，進而促進實質人均GDP成長率。相反地，我們建議高度金融發展國家的政府應限制地下部門的發展。 / This paper focuses on linear effects and nonlinear effects of underground economy and financial development on the growth rate of real GDP per capita. Utilizing two alternative measures of financial development, including the size of private credit and stock trade. The analysis relies on a sample of 22 countries for the period 1981-2008, including 14 developed countries and 8 developing countries. We use three different models, including linear model, nonlinear regression with a cross-term and panel threshold model. The results show that in the linear model, underground economy and private credit have no significant impact on the growth rate of real GDP per capita, but stock trade has a significant positive impact on the growth rate of real GDP per capita. Moreover, the interaction between private credit and underground economy has a significant negative impact on the growth rate of real GDP per capita, on the other hand, the interaction between stock trade and underground economy has an insignificant negative impact on the growth rate of real GDP per capita. Hence, we speculate there have a nonlinear effect between underground economy and the growth rate of real GDP per capita. In the panel threshold model, if the level of significance is set in 10%, we find that the model with private credit and stock trade have threshold effect, it implies that the sample can be split into two regimes: High degree of financial development and Low degree of financial development. Underground economy has a significant positive impact on the growth rate of real GDP per capita when private credit is low, so does stock trade. In conclusion, we suggest that it may be optimal for countries with a less developed formal financial sector to accept more tax evasion, because it accelerates the efficiency in capital investment and then facilates the growth rate of real GDP per capita. On the contrary, countries with a more developed formal financial sector should impose more tax compliance.

No description available.

異質信念(Heterogeneous beliefs)修正傳統資產定價理論中同質信念(Homogeneous beliefs)的基本假設，探討投資者間所持有的不同資訊，以及對於資訊的不同參考程度，如何影響資產定價。 本文試圖以Banerjee(2008)的模型，估計出臺灣投資者對於台灣各家公司股票的價格參考密度；並且透過外部研究者的預測作為市場不同信念的代理變數，探討異質信念對於臺灣股市交易的價量影響。 經由實證結果發現，在台灣的股市交易市場上，當市場的資訊流通速度愈快，投資者對公開資訊的參考密度愈低，投資者愈易依賴自己所持有的私人資訊。當投資者的行為決策將愈顯紛歧時，對交易量和報酬率的影響皆為正。 / Heterogeneous beliefs, which revise the basic assumption of traditional asset pricing theory- Homogeneous beliefs, study the impact on asset pricing by different information owned and referred by investors. I use the model derived from Banerjee(2008)to estimate the degree how Taiwan investors will take into account stock prices when they make investment decisions. Also, I study how heterogeneous beliefs of investors influence stock prices and trading volume in Taiwan stock market, using predictions of external researchers as a proxy variable of dispersion in beliefs. The empirical results show that the degree which investors take into account prices will be lower when a faster information flowing speed exists in Taiwan stock market. When investors rely more on their private information, their investment decision will become much diversified. Dispersion in beliefs has a positive influence on stock trading volume and return.

一、金融發展對經濟成長的影響：動態門檻效果的分析 本研究旨在探討於全球的架構下，各國金融發展對於經濟成長之關係為何？並且進一步探討銀行發展及股市發展是否對經濟成長有不對稱效果。對於過去相關文獻無法獲得金融發展與經濟成長間一致的關係，我們懷疑應與金融發展與經濟成長之間為非線性關係有關。延伸 Shen and Lee (2006) 我們將探討是否於金融發展與經濟成長間存在著銀行的門檻效果，並提出兩個假說，第一是「blessing-in-low-regime」，即在低度銀行發展區域，金融發展對於經濟成長有正面影響。第二是「curse-in-high-regime」即在高度銀行發展區域，金融發展對於經濟成長有負面影響。本文發展一個新的模型：dynamic panel threshold model (DPTM)是延伸Hansen (1999)的panel threshold model，認為經濟成長會受到自己上一期變數所影響，結果指出就銀行發展對於經濟成長的影響而言，在低度銀行發展區域支持「blessing-in-low-regime」；在高度銀行發展區域支持「curse-in-high-regime」。反之，就股市發展對於經濟成長的影響，在低度銀行發展區域並不支持「blessing-in-low-regime」;在高度銀行發展區域亦不支持「curse-in-high-regime」。 二、金融發展與經濟成長的雙向因果關係 本研究探討是否金融發展與經濟成長之間存在一項非線性的雙向因果關係，且此項因果關係是否受到金融發展以及經濟成長程度的影響。在全球的架構下，利用42個國家1976年到2005年的資料，使用一項新發展的計量模型：dynamic panel threshold model (DPTM)來探討此項因果關係。實證結果顯示，當使用銀行發展做為門檻變數時，在低度銀行發展區域，銀行發展對於經濟成長有正面影響，然而股市發展則對經濟成長有負面影響；而在高度銀行發展區域，銀行發展對於經濟成長的影響性則轉向負面影響，而股市發展則轉向正向支持經濟成長。相對地，無論在低度銀行發展區域或是高度銀行發展區域，經濟成長對於銀行發展皆有正面影響。最後，當使用經濟成長做為門檻變數時，本文發現金融發展與經濟成長的因果關係並未改變。 三、金融發展如何影響所得分配？倒U型分配假說與線性假說 本研究使用1976年到2005年42個發展中與已發展國家的資料，分析金融發展與所得分配的關係，並且進一步探討金融發展是否對所得分配有不對稱效果，隨著銀行發展程度的不同，其對所得分配的影響性將隨之改變。過去金融發展與所得分配的相關文獻提出兩項相對的理論假說，分別為Greenwood and Jovanovic (1990)的「倒U型分配假說」與Galor and Zeira (1993)的「線性假說」。本研究發展一項計量模型dynamic panel threshold model (DPTM)來檢驗這兩項假說。分析結果顯示，在低度銀行發展區域，股市發展將提升所得不均；而在高度銀行發展區域，股市發展則轉為減緩所得不均，支持「倒U型分配假說」。相對地，無論在低度銀行發展區域或是高度銀行發展區域，銀行發展對所得不均的影響性皆為負向的減緩效果，不支持「倒U型分配假說」，然而其對所得分配的影響性仍存在不對稱的門檻效果。 / Essay 1: Blessing or Curse? The Role of Financial Development to Economic Growth This study aims to investigate the asymmetric effect between financial development and economic growth by considering the threshold effect. Based on Shen and Lee's (2006) findings, we examine whether the effect of financial development on economic growth depends on the threshold variable of bank development. Our hypothesis is that bank development is a blessing to economic growth at the low bank development regime, but it is a curse at the high bank development regime. To examine the “blessing-in low-regime” and “curse-in high-regime” hypothesis, we develop a dynamic panel threshold model (DPTM) to test this hypothesis. The DPTM is a direct extension of the non-dynamic panel threshold model of Hansen (1999). We conclude that the effect of bank development on economic growth supports the hypothesis. Nevertheless, the effect of stock market development on economic growth does not support the hypothesis. Essay 2: A Bivariate Causality between Financial Development and Economic Growth This study hypothesizes that causal relationship between financial development and economic growth is not linear; however, it may be influenced by the level of financial development or economic growth. A new econometric method, dynamic panel threshold model (DPTM) is proposed to investigate conditional causality. Herein, the thresholds of “bank development” and “economic growth” are applied. When bank development is used as threshold in the low bank-developed regime, bank development is beneficial for economic growth. However, it poses adverse effects on the stock market. In contrast, in the high bank-developed regime, bank development exhibits an adverse effect on economic growth whereas the stock market manifests the opposite effect. Nevertheless, economic growth is beneficial for bank development in both regimes, though the influence is stronger in the low bank-developed regime. Results are robust when the income level of a country is utilized as a threshold. Essay 3: How does Financial Development Affect the Distribution of Income? Inverted U-shaped Hypothesis or Linear Hypothesis This study analyzes the relationship between financial development and income distribution using panel data from both developing and developed countries between 1976 and 2005. Specifically, we analyze whether financial development has an impact on income inequality and whether this impact depends on the threshold variable of bank development. We define the low and high bank development regimes when a country's bank development is below and above the threshold and test two alternative hypotheses the “inverted U-shaped hypothesis” and the “linear hypothesis” with a dynamic panel threshold model (DPTM). The DPTM is a direct extension of the non-dynamic panel threshold model of Hansen (1999). We find little evidence to support the inverted U-shaped relationship between inequality and finance, the effect of stock market development on inequality supports the inverted U-shaped hypothesis. Nevertheless, the effect of bank development on inequality does not support the inverted U-shaped hypothesis. However, the relationship between financial development and income distribution is nonlinear.

本研究的目的在比較研究中國大陸與台灣股市，其上市公司的財務資訊特性，以及這些財務資訊在這兩個股市中的有用性。大陸股市迄今年齡尚輕，僅七歲。唯大陸股市的法規，散戶多，同文同種等與台灣股市相近；不同之處在於，例如國家對上市公司所持有的股份仍佔一定比例、法人極少（中國大陸自五月份始方開放共同基金的發行）、股市分為 A、B 及 H 股等。因此在不同的股市特性及股市環境下，到底公開資訊被使用的情形如何？這種比較分析可令我們深入瞭解，在不同股市特性及股市環境（包括證劵管理環境）下，投資者使用公開財資訊的深度與廣度。因此本研究以公開財務報表為例，探討並比較台灣股市與中國大陸間資訊有用性及有用程度。 本研究採用傳統盈餘／股價關係的研究設計來探討其間的關係，根據　Easton and Harris(1991)及 Lev(1989)的建議，在傳統的盈餘反應係數模式中加入二個與盈餘相關變數，一為「當期盈餘水準除以期初股價」的盈餘水準變數；另一為「當期盈餘變動除以期初股價」的未預期盈餘變數，以此二盈餘變數實證比較兩岸股市，其相對盈餘資訊內涵。其次，依據 Lipe(1986)的建議及其所建立的實證模式，本研究測試盈餘組成分子是否具備增額資訊內涵，並比較兩岸盈餘組成分子增額資訊內涵是否顯著不同。最後，本研究借用 ERC 模式，並依據 Collins and kothari(1989)加入 ERC 決定因素如公司系統風險、盈餘持續度、公司成長機會等，來增進盈餘／股會間關係的解釋能力，以期在最具解釋力的模式下，比較兩盈餘資訊內涵。本研究台灣方面以台灣證劵交所股票上市公司為對象，研究期間自民國 77 年至 85 年止，共計 9 年為樣本期間，大陸則以在上海、深圳交易所上市的股票上市公司樣本，研究期間則為 1994 年至 1996 年研究結果顯示： 1.兩岸之資訊環境存在顯著差異性。 2.兩岸之盈餘資訊皆具有用性。 3.盈餘組成項目均具增額資訊內涵。 4.考慮 ERC 模式後可增加模式之解釋力。 / This study mainly applies ERC (Earnings Response Coefficient) model to compare the information contents of earnings in Taiwan and China stock markets. In addition, this study investigates the disclosure rules of these two markets in order to develop related hypotheses. This study includes four hypotheses: (1) information environment hypothesis, (2) earning level hypothesis, (3) earnings components hypothesis, and (4) ERC determinants hypothesis. The samples strategically select from. listed companies in Taiwan and China to facilitate our comparisons. The findings can be summarized as follows. 1. In the Taiwan stock market, the firm size is not a significant factor for determining optimal return window; but in the China stock market, the response of larger firms is earlier than smaller firms to the earnings announcement (information environment hypothesis). 2. The Easton-Harris earnings level variable can significantly increase the explanatory power of ERC in the Taiwan stock market; but it cannot significantly increase the explanatory power of ERC in the China stock market (earnings level hypothesis). 3. The earnings components can increase the explanatory power of earnings/return relationship in both Taiwan and China stock markets. In addition, the non-operating components have better explanatory power than operating components in earnings (earnings components hypothesis). 4. The determinants of ERC can increase the explanatory power of earnings/return relationship; but the determinants of ERC have negative effects of the explanatory power of earnings/return relationship.