| Summary: | 碩士 === 國立屏東大學 === 國際貿易學系碩士班 === 107 === With the increasing number of domestic universities, university education has become increasingly popular, and everyone is pursuing higher education. However, from an economic point of view, reading is also a high-cost effort, regardless of the employment outcome after graduation. In the process of studying, the tuition and living expenses paid are a very large expenditure, not everyones burden. To this end, the government started to open a school loan in September of the Republic of China, providing most students with financial difficulties to help them to avoid school dropout.
In this study, the number of students who have been enrolled in the school has remained high in recent years, and the rate of over-release has been high. The sample mother of this study used the information of a financial institution that hosted a school loan in Pingtung, and began to repay the loan on August 1, 105. The students are the research subjects, and the original loan applicants application form is reviewed in the field and the current data status of the borrower in the computer file is observed. As of the end of June, the extreme value and the settled are deleted, and the final selection is as shown. 4-1 2246 valid households, 754 households with default, and 3,000 households as research samples.
The empirical results found that 16 explanatory variables, using the Logistic Regression model, found that in the 16 variables that affect the credit risk of the school loan, the total loan amount (million yuan), loan interest rate, school attribute, legal complaint, The eight explanatory variables, such as authorized deduction, tuition and fee reduction, low-income households and application extension, reached a significant level of 0.01, and the number of explanatory variables such as the number of persons was 0.05, and the remaining explanatory variables were not significant. Among them, the loan interest rate was significantly positively correlated with the two variables such as the lawsuit, and the total amount of loans (10,000 yuan), guaranteed number, school attributes, authorized deductions, tuition and fees reduction, low-income households and application extensions were significant. Negative correlation. The credit risk assessment model established by the Institute has a good predictive ability. It should have an early warning effect on the case credit granting case for the case bank, which will objectively and quickly detect the credit default risk of the student loan applicants, and can reduce the risk. The overdue ratio of the loan for the study is to improve the quality of the
credit, and to reduce the banks bad debt loss to improve the operation of the bank, and to provide reference for relevant government policy decisions.
|
|---|
