汽车消费贷款信用风险管理研究——以z银行为例-research on credit risk management of automobile consumer loan - taking z bank as an example.docxVIP

万方数据 万方数据 摘 要 我国汽车消费信贷业务开展时间不长，规模相对较小，产业链存在较多不成熟的地方。 相比欧美等发达国家高达 70%的汽车消费信贷渗透率，我国这一比率仅为 20%，未来将 迎来更为快速的发展。与此同时我国汽车消费信贷的风险管理机制，特别是其信用风险评 价体系，也需要同步进行完善。 本文以汽车消费贷款行业中具有领先地位的 Z 银行为案例，应用商业银行风险管理 理论分析 Z 银行的汽车消费信贷风险评价体系存在的问题，并提出相应的改善建议，为 Z 银行和其他金融机构在完善汽车消费贷款业务风险评价体系方面提供有益的启示。通过分 析发现，Z 银行的汽车信贷业务不良贷款率总体水平较高；在风险控制上，对汽车消费贷 款业务的利率风险防范较为成功，但对信用风险、操作风险的评估防范仍存在不足。本文 重点针对其信用风险评估指标不统一、评估方法主观性较强、评估结论没有量化等问题提 出改进方法，利用 Logistic 模型构建适用于 Z 银行的信用风险评价体系。 研究结果表明：在 Z 银行的汽车消费贷款风险评估模型中，婚姻情况、学历情况、 单位性质、月收入、担保情况（自有房产和共申/担保人情况）4 个主要指标对个人信用 评分结果有重要影响；其中，未婚客户的违约风险较已婚客户高。最后，针对存在的问题， 提出如下建议：一是进一步完善信用风险评价体系建设，有效控制不良贷款率的上升；二 是加强经销商管理，过滤渠道所带来的信用风险。 关键词： 汽车消费贷款；信用风险； Logistic 模型 I Abstract The automobile consumption credit business in China was developed not for a long time. Its scale is relatively small and the industrial chain is immature. Compared to Europe, the United States and other developed countries, which are 70% permeability of automobile consumption credit, in China, the ratio is only 20%. There will be rapid growth in the future. At the same time, the automobile consumption credit risk management mechanism of our country, especially its credit risk evaluation system, also needs a perfect synchronization. This paper takes Z bank as an example, which has a leading position in the auto loan industry, and use the commercial bank risk management theory to analyze problems of automobile consumption credit risk evaluation system in Z bank. Based on that, propose corresponding improvement suggestions and to provide beneficial enlightenment. According to the analysis, we found that the non-performing loan ratio of auto credit business in Z bank is higher. On risk control, interest rate risk prevention of auto loan business is success. But the assessment and precaution in both credit risk and operational risk, are still insufficient. This paper focuses on problems, such as the heterogeneity of credit risk evaluation index, the strong subjective evaluation method and the non-quantization evaluation c