基于BP神经网络的小额信贷信用风险评估研究

发布时间：2018-07-20 16:35

【摘要】：小额信贷最初的目的是扶贫,后来服务的对象变得更加广泛并逐渐转向商业化。从20世纪70年代开始,小额信贷从无到有,逐渐发展壮大。在经济发展中,小额信贷起到了重要作用。与此同时,在小额信贷快速发展的背后也蕴含了多种风险,信用风险便是其中之一。信用风险是指借款人在到期日无法偿还贷款或者不愿偿还贷款而给放款人造成损失的风险。对小额贷款信用风险评估的准确性高低关乎小额借贷行业发展的好坏。本文对小额信贷的概念进行了详细阐述,并对相关理论进行了梳理。在对相关信用风险评估模型进行比较后,最终选择BP神经网络模型作为本文评估小额信贷信用风险的模型。BP神经网络模型有强大的学习和推理能力,能够处理非线性关系,仿真能力强,这些优点正是本文评估小额信贷信用风险所需要的。在参考已有文献的基础上,本研究在信用风险评估指标体系建设和BP神经网络结构设计上进行了创新。之后,利用本文获取的数据建立BP神经网络模型,得到了较为理想的结果。并且,本研究所得到的BP神经网络模型得到了业界人士的认可,可以为小额信贷信用风险评估提供参考。本文得出的结论是:(1)小额信贷最初是出于扶贫的目的而诞生,在后来的发展中逐渐由扶贫性转向商业化;(2)信用风险是小额信贷的主要风险之一,降低信用风险可从减少信息不对称、建立违约惩罚机制和增强借款人风险控制能力入手;(3)在小额信贷信用风险评估中,BP神经网络模型有一些独特的优势;(4)本文通过对小额信贷信用风险评估做实证分析,发现本研究中所建立的模型对违约预测的正确率要高于对不违约所做的预测正确率。针对小额信贷信用风险评估问题,本文提出的可行对策有:(1)完善征信体系建设,降低信息不对称;(2)建立小额信贷违约惩罚机制;(3)增强借款人信用意识,提高其风控能力;(4)完善小额信贷信用风险评估体系;(5)改进BP神经网络模型。
[Abstract]:Microfinance was originally intended to help the poor, but later became more widespread and gradually commercialized. Since the 1970 s, microfinance has grown from scratch. In economic development, microfinance played an important role. At the same time, there are many risks behind the rapid development of microfinance, and credit risk is one of them. Credit risk refers to the risk that the borrower is unable to repay the loan on the maturity date or is unwilling to repay the loan, thus causing losses to the lender. The accuracy of credit risk assessment of small loans is related to the development of microfinance industry. In this paper, the concept of microfinance is described in detail, and related theories are combed. After comparing the relevant credit risk assessment models, the BP neural network model is chosen as the model to evaluate the microcredit credit risk. The BP neural network model has strong learning and reasoning ability, and can deal with the nonlinear relationship. The simulation ability is strong, these advantages are exactly what this article needs to evaluate the microcredit credit risk. On the basis of reference to the existing literatures, this study innovates in the construction of credit risk assessment index system and the design of BP neural network structure. Then, the BP neural network model is established by using the data obtained in this paper, and the ideal results are obtained. In addition, the BP neural network model obtained in this paper has been recognized by the industry, which can provide a reference for the credit risk assessment of micro-credit. The conclusions of this paper are as follows: (1) Microcredit was first born for the purpose of poverty alleviation and gradually changed from poverty alleviation to commercialization in later development; (2) Credit risk is one of the main risks of microcredit. Reducing credit risk can begin by reducing information asymmetry, Establishing default penalty mechanism and enhancing borrower's risk control ability; (3) BP neural network model has some unique advantages in microcredit credit risk assessment; (4) this paper makes empirical analysis on microfinance credit risk assessment. It is found that the prediction accuracy of the model is higher than that of non-default prediction. In view of the problem of credit risk assessment of micro-credit, the feasible countermeasures proposed in this paper are as follows: (1) to improve the construction of credit system and reduce the information asymmetry; (2) to establish a penalty mechanism for microcredit default; (3) to enhance the borrower's credit consciousness. Improve its ability of wind control; (4) improve the credit risk assessment system of microcredit; (5) improve the BP neural network model.
【学位授予单位】：云南财经大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;F832.4

【参考文献】