本研究的目的，在於檢視購併案產生後，企業的報酬是否會受到平均數復歸（Mean reversion）的影響，而逐漸回復到產業的平均報酬。並且分析主併公司與被併公司佔雙方總資產比例的高低差異，所受到的不同影響程度。 本次研究的樣本採用1987～2006年中的美國電子電機產業共77件購併案，並依照主併公司佔雙方總資產的比例不同，分成為兩個群集。將兩群集做比較分析。實證結果顯示，兩個群集均有顯著的平均數復歸現象。而主併佔雙方總資產比例較高的群集，在購併第一年即有顯著的優於產業平均報酬，而主併與被併資產相近的群集，則是於第四年開始才有顯著的績效改善。根據實證結果，我們做出以下的結論： 1.企業的報酬具有平均數復歸的特性，表現優良的公司會被他人學習，而使整體產業平均報酬逐漸提升，漸漸蠶食原先與產業平均之間的差距；相反的，表現不佳的公司亦透過學習，逐漸改善自己的績效，也使的整體產業平均報酬上升，最後雙方都將趨向產業的平均水準。 2.當研究者要分析企業遭受偶發性事件（如購併）的影響時，應當考慮平均數復歸的所帶來的效果。如果忽略了平均數復歸所帶來的效果，將會對事件產生的影響給予錯誤的解讀。 3.面對購併績效的研究時，必須移除平均數復歸的負面效果，如此才得以正確估計購併後的績效表現。且此時主併公司資產佔雙方資產比例較高者，於購併第一年就有明顯的績效改善的效果，而雙方資產相近者，則在購併成立後第四年開始才有較佳的績效表現。 / The purpose of this research is to look over the return of enterprise can receive influence from mean reversion on post-merger performance. And it analyzes the degree of influence on acquirer companies with different proportion in total assets of both acquirer and target companies. The sample of this research adopts 77 cases of electronic and electrical equipment industry of U.S.A. in 1987- 200, take according to the proportion of total assets of acquirer company in both total assets of acquirer and target company, divide into two clusters. Make comparative analysis of these two clusters. According to the empirical results, we make the following conclusions: 1.The return of enterprises has characteristic of mean reversion. Both well performed companies and worse performed companies will move toward the whole industry's average return. 2.As the researcher wants to analyze the influence of enterprises suffered from the sporadic incident (such as M&A), they should consider the phenomenon of mean reversion. Take off the negative result of the mean reversion in order to estimate correctly for the exactly performance after M&A. 3.The acquirer companies which total assets have a higher proportion in both acquirer and target companies, will have superior to industry average return apparently in first year, and the acquirer companies which total assets close to the target companies, will have apparent improvement in performances until the fourth year.

該篇文章中提出一個新的交易方式：條件機率交易模型conditional probability trading model。 這個模型應用了三個主要的基本假設： （1）總體經濟因子和股價指數間有相關性。因此可以透過總經指標來衡量股市應有的合理價位。 （2）股價具有回歸均數的特質。亦即股價一旦過度偏離基本價值，理論上會傾向回復到基本價值之上。 （3）股價指數相對於基本價值線的距離，將會影響偏態係數的大小。 根據以上三個性質，試圖建構出一個能夠捕捉股價指數變動的模型，並用以進行交易模擬，觀察其是否能獲取正報酬。 / The trading strategy, conditional probability trading model(CPTM), is presented in this article. We’ve tried to develop a new trading strategy which is built up by the combination of the properties which includes 1)the relationship between macroeconomic factors and stock market. 2) mean reversion and 3) conditional skewness. The conclusion implies that we may successfully find out a method to combine fundamental and technical analysis, if this method is proved effective. The former hypothesis is assumed that the different level of stock market index may stand for a specific condition of macroeconomic environment. Meanwhile, a better fundamental economic condition could reasonably create a higher stock market index, vice versa. By observing the fundamental value, we can figure out the market ,currently, is over-priced or under-priced. Next, we construct a trading model which is graphed like Bollinger bands. According to specific rules, it provides buying or selling signals. In some special situations, it can also forecast the turning points of the stock market precisely. 3) Skewness also plays a very important role in CPTM, because one of the hypothesis assumes that overpriced /underpriced stock market probably accompanies with left-skewed / right-skewed distribution of daily stock return. The hypothesis of dynamically adjusted skewness implies the concept that over-priced/under-priced stock market has higher propensity to decline/rise. To judge the trading timing is the core value in this model.

由於經濟環境的演進和企業規模的日益擴大，有時效性的工作項目因而與日俱增。所以內部稽核須與外部審計人員搭配協調充分合作，避免工作重覆而造成人力，物力有財力的浪費，以期達到雙方查核效率的提高。而在相互合作之前，外部審計人員應先評估內稽核功能發揮的程度，以作為合作程度的依據，然而外部審計人員如何評估內部稽核的功能呢？此一問題應當會反映在外部審計人員評估內部稽核時所考慮的各種因素之內。因此，本論文將探討外部審計人員評估內部稽核所考慮因素之重要性等級及各因素之間的相關性，以期能此一過程加以評估。 本研究系以問卷調查方式，以外部審計人員為研究對象，根據國外所做相關研究之理論，試圖將外部審計人員評估內部稽核所考慮之因素及其重要性作一確認與評估，在進行分析時，將利用平均數及標準差來排列各影響變數之重要性，利用因素分析將各影響變數歸類為影響因素，利用復相關分析以測定影響因素與內部稽核的能力，客觀性及工作品質間是否具有顯著相關，以及對影響因素和外部審計人員的基本資料特質從事變異數分析。

本論文所探討之主題乃是針對一般所利用之集群方法，試著以整數規劃的方法來探討 集群分析的問題。 整數規劃之特性在於其所得之分組結果為真正的最佳解，而一般集群方法（如連鎖法 ，k 一平均數法）所得之結果僅是局部最佳解。 本文共分五章，第一章為緒論；第二章簡介一般集群方法；第三章建立四個整數規劃 的模型，俾用以解決不同需求之集群分析的問題；第四章實例探討，以某國中學生之 學科成績做為集群分析之變數，將每個學生依其成績高低而予以分組，並就一般集群 方法及整數規劃方法各作分析，並予比較；第五章為結論。全文共計一冊，約一萬五 仟字。

本論文主要針對無心F 分配 (Noneentral F Distribution)之基本理論、特性詳加探 討, 並對其應用作進一步之研究。 第一章為緒論。 第二章, 無心F 分配之基本理論, 由均數不為零的常態分配導出無心卡方分配, 進而 藉由無心卡方分配導出無心F 分配, 並探討無心F 分配的各種性質。 第三章, 無心F 分配之推廣, 導出雙重無心F 分配 (Doubly Noncentral F Distribu -tion), 進而討論雙重無心F 分配之各種性質及其在交互作用檢定上檢力之應用。 第四章, 無心F 分配在檢力上之應用, 首先由均數檢定導出Hotelling's T2統計量 , 並證其在對立假設為真下, 服從無心F 分配, 其次則討論由均數差之檢定及變異數分 析時, 導出Hotelling's T2統計量, 最後根據無心F 分配之性質, 即檢力為自由度及 離心參數( Noncentrality Paramenter )之函數, 導出檢定變量增加、檢力是否增加 之檢定統計量為一無心F 分配。 第五章, 無心F 分配在求樣本數上之應用, 討論在各種統計假設下, 為達到某一程度 之檢力, 所需之最小樣本數為何。 第六章為結論。

整體而言，會計盈餘提供財務報表使用者有關於企業獲利能力之相關資訊。然而，由於會計盈餘同時包含了雜訊 (noises) 與偏差 (biases)，因而影響到會計盈餘對公司獲利能力評價的正確性。因此，過去的會計文獻發展出不同的盈餘管理估計模型 (如: Healy 1985; DeAngelo 1986; Friedlan 1994; DeFond and Jiambalvo 1994 以及Dechow, Sloan and Sweeney 1995)，以嘗試去除這些雜訊與偏差。然而，究竟哪一個估計模型能夠提供最為純淨之非裁量性淨利 (nondiscretionary income) 的衡量指標，則並無定論。在效率市場 (market efficiency) 的假說下，本研究透過本益比 (P/E ratio) 的平均數復歸 (mean-reverting) 現象來評比五種盈餘管理估計模型。由於過去的文獻同時發現盈餘成長率與風險係數均會影響本益比的高低，因此，本研究同時將這兩個變數納入考量。 　　實證結果發現，依照上述五種盈餘管理估計模型所估計之本益比皆有平均數復歸的現象。其中Friedlan (1994) 模型在全體樣本與控制盈餘成長率之後，其本益比平均數復歸現象均較其他模型為快；次佳之盈餘管理估計模型為DeFond and Jiambalvo (1994) 與Dechow et al. (1995) 兩模型；最差的則為 Healy (1985) 模型。 / Overall speaking, accounting earnings provide financial statement users with useful information about a firm's profitability. However, because of the biases and noises included in the accounting earnings, the accuracy and reliability of accounting earnings to the evaluation of a firm's profitability may be adversely influenced. In light of this, prior earnings management studies have developed various estimation models of nondiscretionary income (e.g., Healy 1985; DeAngelo 1986; Friedlan 1994; DeFond and Jiambalvo 1994; Dechow, Sloan & Sweeney 1995) with an attempt to remove the biases and noises embedded in the accounting earnings. Nonetheless, there is no consistent empirical evidence about the relative performance of these estimation models. Assuming market efficiency, the main purpose of this study is to utilize the mean-reverting phenomenon of P/E ratios to evaluate the relative performance of these models. Since prior studies have found that earnings growth rate and risk coefficient may affect the magnitude of P/E ratios, we also control for these two variables in our analyses. 　　The empirical results reveal several findings. First, P/E ratios calculated using different earnings management estimation models exhibit the mean-reverting phenomenon. Second, the Friedlan (1994) model has the best performance among all models when we use the overall sample and three subsamples grouped based on the earnings growth rate. In addition, the DeFond and Jiambalvo (1994) and Dechow, Sloan & Sweeney (1995) models perform moderately. Finally, the Healy (1985) model shows the worst performance.

近年來，能源商品的價格隨著國際政治情勢、國際金融環境以及景氣循環的影響產生劇烈波動，基於避險的需求，衍生性商品交易量也逐漸增加。然而，在評價能源衍生性商品的過程中，即期價格動態模型的選擇對於訂價與避險的結果有著顯著的影響，如何選擇一個適當的動態模型以評價能源商品便成為本文研究的目標。在指數與股價選擇權的評價模型中，大多以Black and Scholes (1973)提出的選擇權評價模型作為基礎，但Black-Scholes模型是否適用於評價能源市場的選擇權價格卻是有待商榷。Schwartz (1997)提出以均數回歸模型 (Mean Reversion Model)描述能源即期價格，發現比Black-Scholes模型中所假設的即期價格動態模型更能描述能源市場即期價格的波動。本研究也考慮能源市場遇到重大事件而造成即期價格產生劇烈波動的情況，因此在模型中加入跳躍項以捕捉價格跳躍的現象。另外，能源商品的需求與季節變化有高度相關性，因此本文亦考量即期價格的變動會受到季節性的變動影響，在模型中加入季節性函數，以補捉季節性的價格變化。基於前述模型考量，本研究在各種描述能源商品即期價格特性的動態模型之下，推導各個模型的期貨選擇權定價公式，進一步測試各模型在金融風暴與非金融風暴期間的訂價誤差與避險誤差，以提供投資人或避險需求者於原油期貨選擇權模型選用上之參考。 / In recent years, the price of energy commodities has fluctuated with the international political situation and the international financial environment. For the sake of hedging demands, the trading volume of derivatives has been gradually increasing. In the process of valuation of energy derivatives, choices of the spot price dynamics model have a significant impact on pricing and hedging. Therefore, how to choose an appropriate dynamic model to evaluate the energy commodities has been main purpose of this study. Two main models are tested in this paper. One is the option pricing model supposed by Black and Scholes (1973), and another is the mean reversion model supposed by Schwartz (1997). This study also considered the volatility of the spot price in the energy market in case of major events, so the researcher adds the jump to explore the mean reversion model. In addition, the demand for energy commodities is highly correlated with seasonal variations. The vibration of spot price often affected by the seasonal variations is considered in the research. Therefore, the researchers also take the seasonal function into the research to capture the seasonal price changes. Based on considerations described above, the pricing formula for each model of futures option is evaluated in the research. The researcher further tests the pricing errors and hedging errors of each model during the financial crises and non-financial crises in order to provide the investors and hedging demanders with some suggestions about selecting oil futures option models.

本文主要是為了瞭解統計套利中的配對交易策略是否可運用於台灣股票市場，並驗證此策略在台灣股市之獲利性及可行性，統計套利的優點很多，可以降低投資風險及提高投資交易獲利的機會，是一個能獲取超額絕對報酬率的交易策略。本研究並不考慮兩檔配對股票的任何基本因素變化或事件因素導向所產生的套利機會，只使用統計模型且考慮如股票間投資報酬率之相關性係數等研究方法，並扣除交易成本。 在實證部份，台灣五十成分股之金融產業表現都能達到不錯的結果。電子產業部分，其長期來看是會回到均衡的，所以績效表現也都能如預期般有良好的表現。電信業三雄不管任兩組配對也都能獲得穩定之報酬。最後，傳產業中部分表現差強人意，未來或許可利用其他模型來找尋其適合配對之模型。

Schwarz (1997)提出均數回復過程（Mean-Reverting Process, MR）捕捉能源價格的動態過程，而Lucia and Schwarz (2002)將此模型結合確定季節性函數，並推導出期貨價格封閉解。然而，能源價格常會因為未預期事件的發生而產生大幅度的變動，為了描述價格跳躍的現象，Clewlow and Strickland (2000)延伸Schwarz的模型提出均數回復跳躍擴散模型（Mean-reverting jump diffusion process, MRJD），此模型除了保留均數回復模型對能源價格會回復至長期水準的描述外，再加上跳躍項來描述價格的異常變動。而Cartea and Figueroa (2005)則同時考慮季節性和跳躍因子，並推導出期貨價格封閉解。另外，雖然台灣目前並非京都議定書所規範的國家，但環境保護是未來的趨勢，故在衡量能源產業財務風險時，除了考慮相關原料和產品，應考慮碳權交易之影響。為了探討財務風險管理在能源產業之應用，本文以煉油廠為例，將其表示成特定期貨部位的投資組合，並透過計算投資組合風險值來衡量煉油廠的財務風險。文中使用結合季節性的均數回復過程、均數回復跳躍擴散過程進行模型配適。實證結果顯示，均數回復跳躍擴散模型在回溯測試下表現最佳；另外，考慮碳權交易後會使得煉油廠的財務風險上升。 / Schwarz (1997) proposes the mean-reverting process (MR) to model energy spot price dynamics, and Lucia and Schwarz (2002) extend this model by including mean reversion and a deterministic seasonality. This model can capture the mean-reversion of energy price, but fail to account for the huge and non-negligible price movement in the market. Clewlow and Strickland (2000) extend Schwarz’s model to mean-reverting jump diffusion process (MRJD). Cartea and Figueroa (2005) present a model which captures the most importance characteristics of energy spot prices such as mean reversion, jumps and seasonality, and provide a closed-form solution for the forward. Although Taiwan is not the member of Kyoto Protocol, but Environmental Protection is a trend in the future. In order to measure the financial risk induced by energy industries, we should consider the effect of emission trading. In this paper, we discuss the implication of financial risk management in energy industries by analyzing the exposure of refinery which represented certain energy futures portfolios. We use MR and MRJD process with seasonality to model energy spot price dynamics, and calibrate the parameters to historical data. And, we consider the interaction of all of positions and calculate the Value-at-Risk of portfolios. The results show that among various approaches the MRJD presents more efficient results in back-testing, and emission trading poses additional risk factors which will increase the financial risk for refineries.

本文將模糊理論的觀念，應用在估計、檢定及時間數列分析上。研究重點包括離散型及連續型模糊樣本的定義與度量，模糊參數的最佳估計，模糊排序方法應用於無母數檢定，模糊相似度的定義、性質，以及如何將其應用於辨識不同時間數列間的落差l期相似程度等。我們首先將常見的模糊資料分為離散型及連續型，並針對不同類型的資料，給定對應的模糊平均數、模糊變異數等模糊參數的概念與一些重要性質。接著我們提出幾種估計方法，針對不同的模糊參數進行最佳估計並提出可行的評判準則。進一步地，我們將模糊排序方法應用於無母數檢定推論。最後我們提出模糊相似度的定義與度量。經由系統性的模擬與分析，我們建立兩時間數列間模糊相似度演算法則。實證分析方面，我們利用提出的方法對台灣的股價加權指數、個股股價進行估計及檢定；同時，針對台灣歷年GDP、民間消費、毛投資間的相似性進行偵測，以驗證我們提出的模糊參數估計、模糊無母數檢定及模糊相似度演算法的效率性與實用性。 / In this paper, we apply fuzzy theory in estimation, nonparametric test, and time series analysis. Our focus is on: How to define and measure the discrete type fuzzy data and continuous one? How to find the optimal estimators for fuzzy parameters? How to apply fuzzy ranking methods in nonparametric test when the data is vague? How to define and find the degree of fuzzy similarity between two time series? First, fuzzy data is classified according to its type, discrete or continuous. Then we give some definitions and properties on fuzzy mean, fuzzy variance for different type of fuzzy data. Next, we proposed some estimating methods and evaluation rules. Moreover we apply fuzzy ranking methods in nonparametric test, such as Sign test, Wilcoxon signed rank test, Wilcoxon rank sum test, and so on. Finally, we suggest the definitions as well as the algorithm for computing the degree of fuzzy similarity between two time series. We also give some simulate and empirical examples to illustrate the techniques and to analyze fuzzy data. Results show that fuzzy statistics with soft computing are more realistic and reasonable for the social science research.