亞洲金融風暴爆發後，確實重創國內的經濟環境，使體質較差的企業無法按時償還貸款，導致逾放比節節升高，影響層面甚廣。國內銀行在控管信用風險時，如未能考量總體環境對放款價值的衝擊程度，將無法計提適當的經濟資本，甚至有可能誤導銀行制訂授信政策的方向。基於此，本研究參照 Kim（1999）建立總體經濟模型，即利用4個經濟指標建構一項信用循環指標Z，預測未來景氣的盛衰，並依景氣狀況調整信用轉換機率矩陣，求得考慮總體條件下的機率矩陣，並以回顧測試來驗證該模型的正確性。接著，將歷史平均、考慮總體條件下，以及實際發生的信用轉換機率矩陣代入CreditMetricsTM模型，計算銀行在三種矩陣下放款部位的信用風險值，比較其中差異，進行相關分析。
主要實證結果發現：1. 一銀、華銀、中信銀和土銀四家銀行，利用總體迴歸模型調整後求得條件機率矩陣算出的信用風險值，與實際情況有一致且接近的走勢，顯示該模型能提升銀行計提準備資本的適切性。若以兩者的誤差百分比來看，其中以中信銀的誤差幅度最小；誤差最大的是華銀。2. 由考量總體條件之信用風險值對放款總額的比率，分析各銀行貸放品質的良窳。結果顯示，四家銀行以中信銀的放款風險最低，原因可能來自於，中信銀的風險控管制度良好，能有效監測授信企業的信用品質。3. 利用K-W檢定檢驗四家銀行不同的授信產業，承擔的信用風險是否存在顯著差異。結果顯示，僅有土銀與另三家銀行存在顯著差異。原因可能是營建業之授信客戶信用品質不良，且多以投機級公司居多。

Credit risk management is one of the most critical issues of the commercial banking business. The contribution of the research is to provide the technique of conditional transition matrix which improves the accuracy of credit loss simulation. This research was divided into two sections. One part based on Kim (1999) is to describe a model for estimating the conditional matrix. The main idea is to adopt an established framework with four financial parameters, which we could incorporate credit cycle dynamics into the transition matrix. To implement the technique, we first build a credit cycle index, which indicates the credit state of the financial market as a whole. Furthermore, we adjust the transition matrix in terms of the credit cycle index. Besides, we use backtesting to check out the validity of the model. The other part is that we employ CreditMetricsTM model to count Credit VaR on the loan position of commercial banks in accordance with the unconditional, conditional, and realized transition matrix.
The major empirical results are as follows: First, the Credit VaR of these four banks (First Bank, FB; Hua Nan Commercial Bank, HNCB; Chinatrust Commercial Bank, CTCB; Land Bank, LB) are consistent and close to the trend of the reality. It indicates that the model could really improve the appropriateness of the required capital. Second, we use the ratio (the Credit VaR to the total loans) to analyze the quality of extending credit for banks. The result shows that CTCB has the smallest risk due to well credit risk management that could supervise the credit quality of a borrowing enterprise efficiently. Third, we use K-W test to verify that four banks may take the difference degree risk owing to the different industries. The result indicates that only the LB is different from other banks. It maybe caused by the poor credit quality of construction industry and the most companies are speculators.

書籍部分
1. 周大慶、沈大白等人（民91），風險管理新標竿—風險值理論與應用，智勝文化事業有限公司。
2. 孫丕垣（民89），企業信用風險與國家貨幣風險之關係，東吳大學經濟學系89年碩士論文。
3. 郭素綾（民91），本國銀行信用評等實證模型之研究，國立中正大學企業管理研究所碩士論文。
4. 郭敏華（民89），債信評等，智勝文化事業有限公司。
5. 陳順宇（民86），迴歸分析，華泰書局。
6. 陳木在、陳錦村（民90），商業銀行風險管理，新陸書局。
7. 陳浩誠（民91），選擇權架構下公司債信用風險溢酬之探討，輔仁大學金融研究所碩士論文。
期刊部分
1. 白鎮維、趙令斌（民90），「信用風險之評估-介紹KMV之EDF及若干國內實證」，貨幣觀測與信用評等，頁131-146。
2. 沈大白、蔡嘉倩（民91），「TCRI與累積倒帳機率之關係」，貨幣觀測與信用評等，頁127-135。
3. 葉金江（民87），「退票率與經濟景氣之關係-影響退票率因素之實證分析」，貨幣觀測與信用評等，第十一期，頁78-85。
4. 賴柏志（民91），「台灣地區信用循環指標的建立」，貨幣觀測與信用評等，頁123-127。
5. 賴柏志、白鎮維、張嘉娥（民91），「信用循環指標應用於信用風險修正之研究」，貨幣觀測與信用評等，頁123-127。
網路部分
1. 中央銀行網站：http://www.cbc.gov.tw/
2. 中國民國證券櫃臺買賣中心網站：http://www.otc.org.tw/c_index.htm
3. 台灣經濟新報網站：http:/ www.tej.com.tw/
4. 央行：國銀應加強風險模型研究，中時電子報，91/08/25，http://news.sina.com.tw/sinaNews/chinatimes/twHeadline/2002/0825/10624853.html.
5. 行政院主計處網站：http://www.dgbasey.gov.tw/
6. 行政院經濟建設委員會網站：http://www.cepd.gov.tw/index.jsp
7. 財政部金融局網站：http://www.boma.gov.tw/
8. 網址：http://www.tytrust.com.tw/report9011_2.html.
9. 蘇金祥，「利率模型參數估計：利率期間結構之估計」，http://www.bondnet.com.tw/doc/parestimate.htm.
英文部分
1. Allen, L.（2002）, “A survey of cyclical effects in credit risk measurement models, ” Working Paper.
2. Altman, Edward. I. (1968), “ Financial ratios, discriminant analysis and the prediction of corporate bankruptcy, ” Journal of Finance, 23(4), pp. 589-609.
3. Belkin, B., Stephan J. Suchower, and Lawrence R. Forest (1998), “The effect of systematic credit risk on loan portfolio value-at-risk and loan pricing, ” CreditMetrics Monitor.
4. Belkin, B., Stephan J. Suchower, and Lawrence R. Forest (1998), “A one-parameter representation of credit risk and transition matrices, ” CreditMetricsTM Monitor, Third Quarter, JP Morgan, New York, pp.46-56.
5. Carty, L. and D. Lieberman (1996) “Defaulted bank loan recoveries,” Moody’s Global Credit Research, Special Report.
6. Credit Suisse(1997), CreditRisk+：A credit risk management framework, Credit Suisse Financial Products.
7. Crosbie, Peter J. and Jeffrey R. Bohn（2002）, “Modeling default risk, ” Moody’s KMV.
8. Crouhy, M., Galai, D., and Robert Mark（2000）, “A comparative analysis of current credit risk models, ” Journal of Banking & Finance, pp. 59-117.
9. Finger, Christopher C. (1999), “Conditional approaches for credit metrics portfolio distributions, ” CreditMetrics Monitor, 2(1), pp.14-33.
10. Gupton, Greg M., Finger, Christopher C. and Mickey Bhatia（1997）, CreditMetrics™ － Technical document, New York: Morgan Guaranty Trust Company.
11. Kim, J.（1999）, “A way to condition the transition matrix on wind, ” Credit Risk Supplement, pp.1-12.
12. Marrison, C.（2002）, The fundamentals of risk measurement, The McGraw-Hill Companies, Inc.
13. Merton, Robert C. (1974), “On the pricing of corporate debt: The risk structure of interest rates, ” Journal of Finance, 28, pp. 449-470.
14. Wilson, Thomas C.（1997）, “Credit portfolio risk Ⅰ, ” Risk Magazine October .
15. Wilson, Thomas C.（1997）, “Credit portfolio risk Ⅱ, ” Risk Magazine November.