商业银行不良贷款预警模型与规则优化研究

Research on Optimization of Non-performing Loan Early Warning Rules of Commercial Banks

不良贷款率一直是商业银行高度关注的重点.商业银行在进行不良贷款分析时,可通过企业财务指标对贷款质量进行客观和科学的判断,从而实现对不良贷款的及时预警.基于49家存在失信记录的批发零售业上市企业年报数据,首先运用Entropy信息熵和大数据python技术对显著影响不良贷款的关键财务指标进行选取;其次,运用SPSS软件和Python的相关语句分析了关键指标的样本数据分布情况,以及不同指标之间的相关关系;最后,基于计算结果提出了完善不良贷款预警模型及规则的相关建议.研究成果以期在不良贷款预警中起到重要的参考作用.

Non-performing loan ratio has always been the focus of commercial banks. When analyzing non-performing loans, commercial banks can make objective and scientific judgment on loan quality through corporate financial indicators, thus realizing timely warning of non-performing loans. In this paper, based on the annual report data of 49 listed wholesale and retail enterprises with dishonest records, the Entropy and big data Python technology are used to select the key financial indicators that significantly affect non-performing loans. Secondly, SPSS software and Python related statements are used to analyze the sample data distribution of key indicators and the correlation between different indicators. Finally, based on the calculation results, some suggestions are put forward to improve the non-performing loan early warning model and rules. The research results are expected to play an important reference role in non-performing loan early warning.