查詢結果分析
來源資料
頁籤選單縮合
題名 | 銀行業房貸授信風險評估因素之選擇=The Choice of Risk Valuation Factors of Mortgage Loan within the Banking Industry |
---|---|
作者 | 周建新; 于鴻福; 陳進財; | 書刊名 | 中華管理評論 |
卷期 | 7:2 2004.04[民93.04] |
頁次 | 頁77-103 |
分類號 | 563.1 |
關鍵詞 | Logistic迴歸模型; 房貸戶信評模式; 不動產押值; 不動產估值; Logistic regression model; The mortgage loan accessment model; The pledged value of the estate; The estimated value of the estate; |
語文 | 中文(Chinese) |
中文摘要 | 房屋放款業務為銀行業資金運用最主要之方式,為保證授信之安全性,如何建立一套客觀具體之房貸戶信評模式,實為刻不容緩之課題。本研究以Logistic迴歸模型作為信評模式,分析銀行業在房屋貸款授信模式中,必須考慮之風險評估因素,除了納入其他研究廣為採用的10項變數外,並增加與借款戶信用條件差異性非常密切之五項變數。經由實證資料發現,職業、年齡、年所得、年所得佔借款金額之比率、各行庫額度、各行庫餘額、借款(申貸)金額、不動產押值、估值等九個變數,為房貸戶信用評估模式之重要鑑別因子,以此信用評估模型分析,實際歸類為正常貸款戶之正確預測能力為94.5%,而實際歸類為逾期放款借款戶之預測能力為93.2%,整體而言,其總體正確率亦高達94.1%。因此,本研究中所包含之變數,將可做為銀行業審核房貸戶綜合信用及作為決定授信是否准駁之合理依據。 |
英文摘要 | The mortgage loan is one of the major use of cash deposits in banking industry. In order to guarantee the safety of mortgage loans, it is an important issue to establish an objective and realistic model in accessing the mortgage loans. This paper uses the Logistic regression model as the mortgage loan accessment model. This model attempts to find the key factors that can best explain the mortgage loan market. Besides the ten variables adopted by tradional research, we also include another five varibales related to credit condition of the borrowers. In this study, we find the occupation, age, annual income, the ratio of annual income to the loan, the account quota, the account suuplus, the borrowing amout, the pledged value of the estate and the estimated value of the estate are the significant discriminative variables for the acessment model. Besides, in this model, the correct prediction ratio is 94.5% for the normal borrowers, and 94.1% for the default borrowers. And the overall correct rate is as high as 94.1%. So, the variables considered in our paper can be used as a reasonable and decisive criteria for judging the cmprehensive credit conditions of the mortgage loan market. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。