信贷业务初期小微企业信用评估指标确权策略与方法

快速构建各种与异质小微企业群体金融需求相适应的信用评估模型，是当前信贷机构亟需解决的关键问题，其中评估指标的确权比较棘手。鉴此，基于三标度、多专家、独立判断、多数为主兼顾少数等策略思想的3-SEIJ的确权方法，采用标度映射、判断矩阵一致性调整，以及融合计算三个关键技术，以解决实践中专家判断意见难以清晰表达，需要反复判断，以及意见分歧难以综合的问题。结果表明：在没有牺牲评估准确性能的前提下，本文方法简化了确权操作而提升了效率，适合如信贷业务初期等数据欠缺情境下的信用评估模型的开发。

Strategies and Methods for Confirming the Credit Evaluation Index of Small and Micro Enterprises in the Initial Stage of Credit Business

To meet the numerous and various financial requirements of all kinds of small and micro enterprises (SME), credit institutions need a system to help them quickly build suitable credit evaluation models. However, in practice, determining the index weight of the credit scoring index system is one of the most difficult problems. Aiming at this problem, this paper proposes a new method called Weighting based on 3-Scales and Expert Independent Judgment (3-SEIJ). This method is mainly based on the ideas of 3-scale, multi-expert, independent judgment, majority-first and minority-second, which are implemented by using three key technologies: scale mapping, judgment matrix consistency adjustment, and fusion computing. The proposed method mainly solves following two major problems in practice. It is difficult for experts to express their opinions clearly in discussion so it needs to be judged repeatedly; It is hard to synthesize the different opinions of experts to meet an agreement. The paper describes the operation points of the 3-SEIJ method, the implementation algorithms of each technology and the calculation examples. Comparing the 3-SEIJ method with the normal AHP method through an enterprise application case, we find that the 3-SEIJ method can simplify the operation and improve the efficiency of index weight determining under the premise of not sacrificing the performance of credit model. The 3-SEIJ method can meet the rapid development needs of the credit model for the initial stage of credit business.