摘要
在国内外大量前人研究基础上,将农村商业银行风险主要分为信用风险、流动性风险、市场风险、操作风险和政策风险。应用专家评分法筛选各风险的关键影响因素后,共采用17个影响因素构建风险指标体系。以向福建省农村各地区,包括镇、乡和村的各商业银行农村分支行分散投放后回收的有效调查数据为依据,应用包含17个输入节点、30个隐藏节点和一个输出节点的五层BP神经网络构建评估模型,并进行样本训练和验证,通过对实际输出和标准输出的一致性对比得出结果并表明,农村商业银行风险评价指标体系和模型较为理想,有较好的有效性和可行性,可为农村金融机构准确评估风险提供理论依据。
Based on predecessors＇ research the major risks facing rural commercial banks are divided into five types：credit risk, liquidity risk, market risk, operation risk and policy risk. The method of expert scouring is taken to screen the key influential factors of each risk. 17 influential factors which are identified as key ones by screening are taken to construct a risk index system. And then based on the effective survey data which were obtained from an questionnaire survey conducted in the rural sub-branches of commercial banks scattered in the Fujian rural areas,including towns and villages, a five-layer BP neural network which contains 17 input nodes, 30 hidden nodes and an output node is used to construct an assessment model, which is used for sample training and verification. The comparison between actual output and standard output in consistency shows that the model of the risk assessment index system of rural commercial banks is ideal in effectiveness and feasibility and can provide theoretical basis for the exact risk assessment of rural financial agencies.
出处
《沈阳农业大学学报（社会科学版）》
2017年第5期530-536,共7页
Journal of Shenyang Agricultural University(Social Sciences Edition)
基金
福建省中青年教师教育科研项目社科类(JBS14434)
