| Summary: | 博士 === 國立中山大學 === 企業管理學系 === 88 === Consumer credit is an extremely significant factor in the economic expansion of every country. Financial institutions and credit industry with increasing frequency have applied credit scoring and related models that assign different weights to various items in the application form in order to apprise the creditworthiness of loan applicants. The sum of all the entire multiplied by their appropriate weights formulate the total score for each applicant. Credit granters evaluate this score with predetermined cutoff score to determine whether accept the application or not.
Screening procedures are primarily applied in quality control. It is a way of selecting items whose performance variable is within specifications by observing correlated screening variables instead of the performance variable directly. The problem feature about credit granting decision that is a process of applying measurable information of applicants to predict one’s future defaulting risk is similar to a screening procedure. In this study, screening models are used to assist the credit granting decision. There have been numerous judgement and quantitative methods applied in the field about credit scoring. It is not like those methods to score the applicant directly, but coordinate with other methods to utilize the measured score to decide acceptance or rejection based on statistical or economic objectives.
Firstly, a two-stage screening procedure associated with two performance variables is used to construct two mail credit models that are with/without credit audit and economic models. They are considered to determine the optimal solution of cutoff points. Secondly, the multiple screening variables are used to improve the accuracy of credit assessment. Two models are proposed. One is a separate scorecard approach. The other is a combined scorecard approach. The latter approach can easily be extended to the case with multivariate screening variables. In addition, an economic model is provided so that the optimal cutoff points are determined in economic view. Finally the combined scorecard approach is applied to the mail credit assessment process. Based on economic objective, four improvement models of screening are proposed to upgrade the effect of credit granting decision in advance.
|
|---|
