地方政府融资平台贷款风险等级分类真实性的审计研究  

Authenticity Audit of Local Government Financing Platform Loan Risk Grade Classification

作　　者：刘骅[1] LIU Hua(School of Finance, Nanjing Audit University, Nanjing 211815, China)

机构地区：[1]南京审计大学金融学院,江苏南京211815

出　　处：《山东财经大学学报》2017年第2期1-8,共8页Journal of Shandong University of Finance and Economics

基　　金：国家社科基金后期资助项目"科技金融体系建设与效果评价"(16FGL013);江苏省人力资源和社会保障厅"六大人才高峰"项目"江苏省物联网产业项目运营绩效审计研究"(XYDXXJS-036);南京审计大学2016年首批政府审计研究课题"国家金融安全视角下政府债务审计治理研究"(GASA161016);江苏省高校优势学科建设工程项目应用经济学(苏政办发[2014]37号)

摘　　要：地方政府融资平台的出现具有明显中国式经济文化特色,对其市场化转型中贷款风险等级分类真实性审计正成为领域内关注焦点。结合融资平台信贷特征,构建其贷款风险评价指标体系;依据泛长三角地区融资平台信贷业务的真实数据,运用灰色关联分析法对指标体系进行约简,并采用数据挖掘技术建立基于支持向量机的融资平台贷款风险等级分类真实性审计二分类预测模型,进而对模型的效果进行检验。明确该模型在审计实践中的应用价值,从金融审计资源协同、技术创新和人才培养三个方面凝练提升融资平台信贷审计效率。Local government financing platforms have obvious Chinese economic and cultural characteristics, whose authenticity audit of loan risk grade classification in the process of market-oriented transformation becomes the focus of attention in the field. Combined with the characteristics of credit financing platform, this paper firstly con- structs a loan risk evaluation index system which is reduced based on grey correlation method and the real data from pan-Yangtze River Delta financing platform credit business, and then by adopting data mining technology, con- structs a SVM-based two-classification prediction model for financing platform loan risk grade classification authen- ticity audit. The effect of this model is further tested and its application value in audit practice is made clear. And this model is expected to improve the financing platform credit audit efficiency from the aspects of financial audit re- source coordination, technological innovation and personnel training.

关 键 词：融资平台 贷款风险 真实性审计 灰色关联 支持向量机 

分 类 号：F239.44[经济管理—会计学]