基于BP神经网络的绿色信贷信用风险评价研究

发布时间：2018-03-16 03:06

本文选题：绿色信贷　切入点：信用风险评价　出处：《中国海洋大学》2013年硕士论文　论文类型：学位论文

【摘要】：进入21世纪,人类社会面临的资源环境问题日益凸显,调整经济结构、转变经济增长方式、实现低碳可持续发展成为一种共识。而资金作为经济发展的血液,银行作为资金配置的枢纽,加之银行履行社会责任的要求,使得银行在其中扮演至关重要的角色,调整银行信贷理念和结构、关注环境风险、发展低碳金融已成为国际银行业共有之举。目前,国际大的金融机构大多已加入“赤道原则”,碳金融的理论和实践得到快速发展,我国则在内部环境和经济转型的压力及国际碳金融发展的趋势下于2007年提出绿色信贷,通过开展绿色信贷调节社会资金流向,从源头上切断“两高一剩”企业发展资金,支持低碳环保型行业企业发展。但与此同时,绿色信贷也使银行面临复杂多变的环境风险,对其风险评价和管理的理论和实践上的不足阻碍了绿色信贷的开展。因此,必须建立绿色信贷信用风险评价管理机制,探索完善绿色信贷信用风险评价指标体系及评价模型和方法,为绿色信贷开展顺利开展提供理论依据。本文正是从信用风险评价的角度出发,旨在通过对绿色信贷及其信用风险评价研究现状及相关理论进行梳理的基础上,构建绿色信贷信用风险评价指标体系,并运用BP神经网络以低碳环保类上市公司为样本进行案例分析。具体看来,文章共分如下六部分： (1)对可持续金融理论、银行环境风险管理理论、企业社会责任理论及银行信用风险管理理论进行梳理,为后文奠定理论基础。(2)对绿色信贷信用风险评价相关内容进行定性分析研究。在对绿色信贷内涵及其环境风险进行介绍的基础上,对绿色信贷信用风险的表现形式、特点及成因予以分析,之后对绿色信贷信用风险评价的内涵、现状及问题进行阐述。(3)绿色信贷信用风险评价指标体系及模型构建。在分析总结已有的绿色信贷风险评价指标体系的基础上,针对绿色信贷信用风险的特点构建出包含传统财务指标和环境风险指标在内的指标体系,其中环境风险指标体系包含每股社会贡献值等四项的企业绩效及“三废”排放量三项的环境质量两大类；之后,选择BP神经网络进行风险评价,介绍BP神经网络相关原理,构建绿色信贷信用风险评价模型。(4)案例分析。选取110家低碳环保类上市公司为样本,运用该指标体系和模型,进行网络训练、检验和使用,并依据所得结果对绿色信贷信用风险状况进行评价。(5)从外部制度环境和内控两方面提出绿色信贷信用风险管理的对策建议。(6)对文章进行总结,并从研究方法、思路等方面对绿色信贷信用风险评价研究进行展望。文章的创新点在于：(1)以绿色信贷的企业信用风险为研究视角,选用低碳环保类上市公司为研究样本,对银行绿色信贷信用风险进行评价,为绿色信贷信用风险评价及管理提供新的思路。(2)实证方面,基于绿色信贷信用风险的特性及数据的可得性,引入每股社会贡献值、单位产值能耗、“三废”排放量等七个环境风险指标,构建出包含财务指标和环境风险指标的绿色信贷信用风险评价指标体系,并运用BP神经网络对其进行定量化的案例分析,为绿色信贷信用风险评价模型及方法做出探索性研究。
[Abstract]:In twenty-first Century, resource and environmental problems of human society has become increasingly prominent, the adjustment of economic structure, change the mode of economic growth, to achieve low-carbon sustainable development has become a consensus. As capital is the economic development of the blood bank, as the hub of the allocation of funds, coupled with the banks to fulfill the social responsibility requirements that banks play a vital role in the the adjustment of bank credit, the concept and structure, focus on environmental risk, the development of low carbon finance has become the international banking industry has to move. At present, most large international financial institutions have joined the "Equator Principles", the theory and practice of carbon finance has been the rapid development of China in the internal environment and the pressure of economic restructuring and international carbon the trend of financial development in 2007 put forward green credit, regulating the flow of social capital through the development of green credit, cut off from the source of the "two high and one left" enterprise development capital Kim, support the development of low-carbon industry. But at the same time, the green credit also makes the bank face environmental risk complex, lack of the risk assessment and risk management theory and practice hinders the green credit. Therefore, we must set up green credit risk assessment management system, explore and improve the model and method green credit risk evaluation index system and evaluation, to carry out green credit smoothly and provide a theoretical basis.
This article is from the perspective of credit risk evaluation perspective, aims to sort out the basis of green credit and credit risk assessment research status and related theories, constructs the evaluation index system of green credit risk, and the use of BP neural network with a low carbon environmental protection of listed companies as a sample for case analysis. Specifically speaking, the article is divided the six part is as follows:
(1) the financial sustainable development theory, the theory of banking environment risk management, credit risk management theory and the theory of corporate social responsibility and the bank to sort out the theoretical basis for the later study. (2). Qualitative analysis of relevant content of green credit risk evaluation based on the connotation of green credit and environmental risk on the introduction form of green credit, credit risk, characteristics and causes are analyzed, then the connotation of green credit risk evaluation, the present situation and problems in this paper. (3) the green credit risk evaluation index system and model construction. Based on the green credit risk evaluation index system of the existing, according to the characteristics of the green credit risk index system is constructed including traditional financial indicators and environmental risk index, the environmental risk index system includes social contributions per share The value of four of the performance of enterprises and the "three wastes" emissions of environmental quality of three items in two categories; then, risk evaluation and selection of BP neural network, introduces the principle of BP neural network, the construction of green credit risk evaluation model. (4) case analysis. Selected 110 low carbon Environmental Protection Corporation the sample, using the index system and model, network training, inspection and use, according to the results of the credit risk of the green credit evaluation. (5) put forward countermeasures and suggestions of green credit risk management from the two aspects of the external institutional environment and internal control. (6) the article summarized, and the research methods. Ideas and other aspects of the green credit risk evaluation research was discussed.
The innovation of the paper is: (1) to the credit risk of green credit business as the research perspective, the use of low carbon environmental protection listed companies as research samples, to evaluate the green bank credit risk, to provide new ideas for the green credit risk evaluation and management. (2) the empirical data, and characteristics of green credit the availability of credit risk based on the introduction of social contribution value per share, the unit value of energy consumption, "three wastes" emissions seven environmental risk indicators, including financial indicators and constructs the environmental risk index of green credit risk evaluation index system, and use BP neural network to quantify the case analysis, make study on green credit risk evaluation model and method.

【学位授予单位】：中国海洋大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.4;TP18

【引证文献】