我国商业银行对公客户内部评级系统建设与分析

发布时间：2018-06-27 08:41

本文选题：巴塞尔协议 + 内部评级　；参考：《山东大学》2016年硕士论文

【摘要】：近年来,我国商业银行受经济增速回落、转型升级、国内需求乏力和产能过剩等多方面不利因素影响,银行的信贷资产和风险管理面临诸多挑战,急需持续改进风险管理和推动业务创新,以增强风险监控水平,提升资本监管的风险敏感性。《巴塞尔协议II》和《巴塞尔协议III》的出台,代表了现代银行业风险管理的发展方向,使银行能够更加全面、客观地管控各项风险,通过内部评级法的实施,能够有效提高国内银行的信用风险管理水平。实施内部评级法是银监会的监管要求,更是银行用来提高自身风险管控和资本管理质量的内在驱动力。内部评级法作为新资本协议的重点,是一个二维的评级体系,包括客户评级和债项评级,这是巴塞尔新资本协议对于内部评级法的要求,为商业银行提供了更为科学、全面的信用风控计量方法。对于内部评级体系,商业银行需要确定信用风险资产进而计算资本充足率,最终的落脚点是违约概率(PD)、违约损失率(LGD)、违约风险暴露(EAD)和期限(M)四个参数的确定。本文的研究主题是针对现有商业银行对公法人客户的内部评级系统存在的问题,根据内部评级体系和内部评级系统建设的要求,提出系统建设与完善的相关对策。通过对内部评级二维评价体系方法论和评级打分模型的研究,提出了内部评级模型库设计和关键计算的工程化实现方案。文中通过灵活的模型配置和因子、参数设置,能够对模型计算的优化调整作出快速响应,以求做到内部评级系统更高的可维护性、可扩展性。文中首先采用聚类方法进行客户评级模型的分类和设计,将包括评级模型识别、指标计算、评级等级计算、违约概率计算等在内的一系列客户评级模型环节进行模块化分拆、组装；接着基于债项评级的业务特点,对债项评级关键计量指标和债项评级因子进行集中参数化管理设计,更快地适应系统管理和外围环境变化的需要；最后,对内部评级系统建设使用情况进行了分析总结,并对系统以后的研究重点和方向提出了建议。
[Abstract]:In recent years, China's commercial banks have been affected by many adverse factors, such as falling economic growth, transforming and upgrading, weak domestic demand and overcapacity, and so on. The banks' credit assets and risk management are facing many challenges. It is urgent to continuously improve risk management and promote business innovation in order to enhance the risk monitoring level and enhance the risk sensitivity of capital supervision. The introduction of Basel II and Basel III represents the development direction of modern banking risk management. Through the implementation of internal rating law, the credit risk management level of domestic banks can be effectively improved. Implementing internal rating law is the regulatory requirement of CBRC, and it is also the internal driving force for banks to improve their risk control and capital management quality. As the focus of the new capital agreement, the internal rating method is a two-dimensional rating system, including customer rating and debt rating. This is the requirement of the new Basel Capital Accord for the internal rating method, which provides more scientific for commercial banks. Comprehensive credit wind control measurement method. For the internal rating system, the commercial banks need to determine the credit risk assets and then calculate the capital adequacy ratio. The final foothold is the probability of default (PD), the default loss rate (LGD), the default risk exposure (EAD) and the maturity (M). The research theme of this paper is to aim at the problems existing in the internal rating system of the commercial banks to the customers of public legal persons, according to the requirements of the internal rating system and the construction of the internal rating system, and put forward the relevant countermeasures for the construction and perfection of the system. Based on the research on the methodology of internal rating two-dimensional evaluation system and the rating scoring model, the engineering implementation scheme of internal rating model base design and key calculation is put forward. Through flexible model configuration, factor setting and parameter setting, this paper can make a quick response to the optimization adjustment of the model calculation, in order to achieve higher maintainability and expansibility of the internal rating system. In this paper, a series of customer rating models, including rating model identification, index calculation, rating grade calculation, default probability calculation and so on, are divided into modules by using clustering method to classify and design the customer rating model, which includes the identification of the rating model, the calculation of the index, the calculation of the rating grade and the calculation of the default probability. Then, based on the business characteristics of debt rating, the key measurement indicators and debt rating factors of debt rating are designed for centralized parameterized management to meet the needs of system management and external environment change more quickly. This paper analyzes and summarizes the construction and use of the internal rating system, and puts forward some suggestions for the future research of the system.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F832.33

【相似文献】