個人抵押貸款風險之研究-以某銀行某分行為例

The Study on Risk Analysis of Personal Mortgage Loan-Evidences from N Branch of N Bank

盧彥碩(Yan-Shuo Lu)

指導教授 ： 陳美菁

朝陽科技大學/管理學院/保險金融管理系/碩士(2011年)

摘要

銀行收益來源之一為貸款利息收入，為確保貸款本金及利息能如期扣繳，信用風險管理品質之良窳對以吸收風險賺取報酬為本質之銀行業務而言影響至為重大。為能建立完善的個人信用風險管理制度及加強個人信用風險管理方法，同時落實新巴塞爾資本協定風險管理之精神，藉以達到提升個人授信資產之獲利性與安全性的雙重目標，是目前銀行業重要的議題之一。否則一旦發生逾期放款，對銀行造成的損失將是相當顯著且深遠，故建立安全及明確之授信審核程序並配合模型之運用來覈實審查，當可提升授信之安全性及有效性。
本研究將依前述之觀念，採用某銀行某分行至民國99年9月30日止已完成貸放程序之個人授信資料進行分析，對於申請中或已核准尚未貸放之授信案件予以剔除。利用邏輯斯迴歸分析(Logistic Regression)，探討個人抵押貸款可能產生逾期風險及逾期之關鍵因素。其研究結果發現，以違約率0.074作為模型之臨界值可得到最小誤判成本，最後實證結果得知保證人人數、核定利率、財團法人金融聯合徵信中心資訊及支出所得比為影響個人抵押貸款風險之重要變數。

關鍵字

邏輯斯迴歸 ； 個人抵押貸款 ； 誤判成本 ； 臨界值

並列摘要

One of the bank incomes is personal mortgage-loan interests. In order to make sure that the loan capital and its interests can seize and hand in for cancellation as scheduled, it is important to take balance between risks and benefits using appropriate personal risk management method for bank. For established & strengthen individual credit risk management and its control to realize the spirit of New Basel Capital Accord, so it is an important issue for the bank to achieve the goal and to promote the profits and security of personal mortgage loan. Otherwise it once occurs NPL(Non-performing Loan), the losses to the bank will be quite remarkable and profound.It can increase the asset security and effective when the bank establish safety and clearity about review procedure with utilization of the model.
Following the aforemention, this paper adopts and analyze the N Branch of N Bank totally personal lending mortgage loan, the period is to September 30,2010, but it is not including the case that on application or unlending. By using Statistics of Logistic Regression, we can discuss the risks and key factors of what NPL(Non-performing Loan) happened.The results revealed that the model with 0.074 cut-off value obtained relatively minimum misclassification cost. Finally, the important variables that affected the loan default were the number of guarantor,interest rate,information from JCIC(Joint Credit Information Center)and the ratio of income and expenditure.

並列關鍵字

Logistic Regression ； Personal Mortgage-Loan ； Cut-off Value ； Misclassification Cost

參考文獻

1.王思評（2006）。房屋抵押貸款授信風險評估研究-以X銀行為例。未出版碩士論文，私立大同大學事業經營研究所，台北。

10.蔡士斌（2006）。房屋貸款戶信用評量之研究-以國內某一銀行為例。未出版碩士論文，國立中興大學應用經濟學系所，台中。

15.Orgler, J.（1970）Scoring Model for Commercial Loans. Journal of Money,Credit and Banking,2（4）, 435-445.

16.Ohlson,J.（1980）Financial Ratios and the Probabilistic Prediction of Bankruptcy. Journal of Accounting Research,18, 109-131.

Steenackers,A.and M.J.Goovaerts（1989）,A Credit Scoring Model for Personal Loan,Insurance Mathematics Economic. Updegrave（1987）,How Lender Size You Up,Monday.