各項信用衍生性商品中，最廣為人知的商品即為違約風險交換(credit default swap; CDS)，但由於金融市場與商品的擴張，標的資產不再侷限單一資產而是增加至數家或數百家，而多個標的資產的違約風險交換稱為一籃子違約風險交換(basket default swap; BDS)。 根據Chiang et al. ((2007), Journal of Derivatives, 8-19.)，在單因子模型中應用importance sampling (IS) 來估計違約給付金額，不僅可以確保違約事件的發生，還可以提高估計的效率，因此本文延伸此一概念，將此方法拓展至多因子模型。本文分為三種方法：一為將多個獨立因子合併為一邊際因子，並針對此邊際因子做importance sampling；二為找出其最具影響性的因子應用importance sampling；最後，我們針對portfolio C 於Glasserman ((2004), Journal of Derivatives, 24-42.) 將標的資產分為獨立兩群，我們將分段利用exponential twist及Chiang et al. (2007)所提出的單因子方法，提升違約事件發生的機率。 借由數值模擬結果，發現將多個獨立因子合併為一邊際因子的方法應用於標的資產為同質模型(homogeneous model)，會有較佳的結果；對具影響性的因子應用importance sampling的方法於各種模型之下的估計結果都頗為優秀，但其variance reduction較差且流程較不符合現實財務狀況，方法三則為特殊模型的應用，其只適用於能將標的資產獨立分群的模型，並且估計準確與否和選取exponential twist的位置有重要關係，第四節我們將同時呈現兩個不同位置的估計值與variance reduction. / Credit default swap (CDS) is the most popular in many kinds of credit derivatives, but number of obligor couldn’t be one always because of the expansions of financial market and contracts. CDS which has been contained more than one obligor is called basket default swap (BDS). According to Chiang et al. ((2007), Journal of Derivatives, 8-19.), applying importance sampling to estimate the default payment in one factor model could not only guarantee the default event occurs but also improve the efficiency of estimation. So this paper extends this concept for expanding this method to multiple factors model. There are three methods for expanding: First, merge multiple factors into a marginal factor and apply importance sampling to this marginal factor; second, apply importance sampling to the factor which has higher factor loading and third, we consider portfolio C in Glasserman ((2004), Journal of Derivatives, 24-42.) and divide total obligors into two independent groups. We would use the ways of exponential twist and the method in one factor model of Chiang et al. (2007) considered in two parts to raise the probability of default event occur. Borrow by the result of numerical simulation, method 1 has better results when obligors are homogeneous model; the results of method 2 are outstanding in each model, but its efficiency is worse and the procedure doesn’t fit with the realistic financial situation; the third method is the application of the special model, it could only apply to the model which could separate obligors independently, and the accuracy of estimates is strongly correlated to the position of exponential twist. In section 4, we would display the estimator and variance reduction in two different positions.

本研究的目的為建立一套適用於台灣股市的主動式量化投資策略。本研究利用多因子 Alpha 模型為分析架構，試圖掌握多維度的股價影響因子，以資訊係數（Information Coefficient）、T-test of ICs、成功率（Success rate）以及 Quintile 累積報酬做因子有效性的檢定，篩選出穩定且有效解釋股價報酬的月頻率因子，再組合因子形成Alpha 股票評分，Alpha 可拆解成三部分，包括市場波動度、因子預測下一期報酬的能力以及因子的獲利能力。本論文以此評分做為股票投資權重的依據，建構一個以台灣中型 100 指數為標竿指數的投資組合。實證結果發現，此主動式量化投資策略能夠有效擊敗標竿指數，獲得平均每個月 3.7%的超額報酬。 　　本研究並嘗試以設定原始權重保留率的方法，控制追蹤誤差以降低週轉率與交易成本，實證結果發現，此方法可有效降低追蹤誤差，但隨著保留率提升，資訊比率（Information Ratio）與投資組合的超額報酬將降低。 / The objective of this study is to build an investment process of active quantitative stock selection model. In this study, we use the Alpha Multi-factor model to find a multitude of factors which are significantly relative to the stock return. The tests we conduct to select the factors that end up in the final multi-factor model are monthly Information Coefficient, T-test of ICs, success rate and quintile cumulative return. Then we examine how to optimally combine correlated factors and calculate the Alpha score for each stock for each period. Alpha is Volatility times IC times Score. Volatility is the cross-sectional volatility of the residual return. IC is the predictive power of the model. And Score are the cross-sectional scores for each stock. We utilize a simple method to construct the portfolio that uses the Alpha score to adjust the weight of component stocks in the benchmark. The empirical result reveals that this investment process successfully outperform the Taiwan Mid-Cap 100 Index benchmark. Moreover, this study tries to decrease the turnover rate and transaction costs by controlling the tracking error. We set the original weight retention rate of the benchmark to control the tracking error. The empirical result reveals that the method works. But as the retention rate rises, the Information ratio and the excess return drops.

商業銀行是以借貸為主的金融機構，銀行獲利的主要來源，是從存款大眾手中取得短期資金，再將資金貸放給政府或企業進行長期投資。銀行「借短貸長」的業務，常使得其資產與負債產生存續期間不一致的問題，當利率非預期變動時，會改變資產與負債的真實價值，進而影響到銀行的淨值及股票報酬率。此外，匯率變動的風險也是銀行常常面臨的問題，尤其是當銀行涉足國際業務時，匯率的變動常常會使銀行所持有的外幣部位價值改變，進而影響到銀行的真實價值。另外一個會影響到銀行資產與負債價值的因素，就是信用風險的問題，總體經濟環境的信用品質變動，常常會影響銀行放款的還款機率，進而改變銀行放款的實質價值。 本文採用過去學者們所研究過的銀行股價三因子模型，即市場因子、債券因子、匯率因子，並加入代表總體信用風險的第四個因子，以及代表抵押品價值變動的第五個因子，成為銀行股價五因子模型。以日本銀行業的股價報酬為研究對象，實證結果顯示:新加入的總體信用風險因子，對於銀行股價報酬率的確產生顯著的負向影響，也就是當借貸市場信用品質愈差（信用風險越高）時，整體銀行股價的報酬率下降。且在四種類型的銀行中，地方銀行所估計出的信用風險顯著的比例最高，代表資產規模較小、放款業務較集中的地方銀行，其信用風險確實較其他類型的銀行為高。另外，在日本泡沫經濟破滅以後的銀行危機時期，以股價多因子模型來衡量的銀行信用風險也有上升的現象。