网络借贷（P2P）平台的量化监管研究

发布时间：2018-03-24 08:42

本文选题：网络借贷(P2P)　切入点：量化监管　出处：《华南理工大学》2015年硕士论文

【摘要】：网络借贷(P2P)作为一种网络金融时代下的金融创新模式,对于完善我国金融体系、弥补中小企业融资缺口以及缓解民间资本投资需求具有重要意义。随着网络借贷(P2P)平台数量的剧增,网络借贷(P2P)行业的竞争也日渐激烈,相关风险也在不断积聚,2013年以来网络借贷(P2P)平台不断出现诈骗、跑路和体现困难等风险事件,大大阻碍行业发展。因此,网络借贷(P2P)监管政策的出台迫在眉睫,通过监管改变现状,促使行业健康发展。但是,网络借贷(P2P)作为一个新兴行业,监管部门应给予其更大的发展空间,避免出现“一管就死”。为此,本文尝试基于一系列量化方法建立一套适合我国网络借贷(P2P)运营特征的量化监管体系,为银监会监管网络借贷(P2P)平台提供参考。本文在定性分析方面,首先通过一些指标数据对我国网络借贷(P2P)平台的发展现状进行归纳总结,包括平台的运营情况和运营模式,并指出我国网络借贷(P2P)行业正面临着法律风险、商业经营风险、信用违约风险等问题。接着,分析国内网络借贷(P2P)监管的发展,确定我国监管层对待网络借贷(P2P)这一新兴行业持有鼓励和支持的态度,为之后的量化分析奠定基础。在量化监管研究方面,针对我国网络借贷(P2P)平台的风险特征,借鉴传统线下信用评级方法,本文首先运用相关性分析和主成分分析方法构建监管评价指标体系,并基于我国网络借贷(P2P)平台发展的特点选取基于熵值修正G1组合赋权的评价模型进行监管评价,以此体现监管层的适度监管和分类、分级监管的原则。熵值修正G1组合赋权能同时反映客观数据信息和专家意见,由此反映平台运营特征及监管政策,这种组合赋权方法更适合并能更有效地对我国网络借贷(P2P)平台进行监管评价。其次,在对平台评价结果的基础上,建立基于Logistic回归模型的风险预警模型,通过实证研究证明该模型能有效地对我国网络借贷(P2P)平台的风险进行预测,为监管程序提供决策支持。最后,详细分析了监管评价指标体系、监管评价模型和风险预警模型在实际监管过程中的应用,以量化手段落实网络借贷(P2P)的分类监管、适度监管、科学监管。
[Abstract]:As a kind of financial innovation mode in the era of network finance, network lending and lending (P2P) can improve the financial system of our country. It is of great significance to make up for the financing gap of small and medium-sized enterprises and to ease the demand for private capital investment. With the rapid increase in the number of online lending platforms, the competition in the online lending and lending industry is becoming increasingly fierce. The risks are also accumulating. Since 2013, there have been a lot of risks such as fraud, running the road and difficulties in implementing the P2P platform, which has greatly hindered the development of the industry. Therefore, the introduction of a regulatory policy on online lending and lending is imminent. Changing the status quo through regulation to promote the healthy development of the industry. However, as a new industry, regulators should give them more room for development and avoid the "death of one tube". Based on a series of quantitative methods, this paper attempts to establish a set of quantitative regulatory system suitable for the characteristics of China's network lending and lending, which provides a reference for the CBRC to supervise the network lending and lending platform. First of all, through some index data, this paper summarizes the current situation of the development of China's network lending platform, including the platform's operation and operation mode, and points out that our country's network lending and lending industry is facing legal risks and business risks. Credit default risk and other issues. Then, by analyzing the development of domestic online lending and lending (P2P) regulation, it is determined that China's regulators have an encouraging and supportive attitude towards the emerging industry of online lending and lending (P2P). In the research of quantitative supervision, aiming at the risk characteristics of our country's network lending platform, we draw lessons from the traditional offline credit rating method. In this paper, we first use the methods of correlation analysis and principal component analysis to construct the regulatory evaluation index system, and based on the characteristics of the development of China's network lending and lending platform, we select the evaluation model based on entropy modified G1 combination weight to conduct regulatory evaluation. According to the principle of appropriate supervision and classification and hierarchical supervision, entropy correction G1 combination weight can reflect both objective data information and expert opinion, and thus reflect platform operation characteristics and regulatory policies. This combination weighting method is more suitable and effective for monitoring and evaluation of China's network lending and lending platform. Secondly, based on the evaluation results of the platform, a risk early warning model based on Logistic regression model is established. It is proved by empirical research that the model can effectively predict the risk of China's network lending and lending platform and provide decision support for the regulatory process. Finally, the monitoring evaluation index system is analyzed in detail. The application of supervision evaluation model and risk warning model in the process of actual supervision, the classification supervision, moderate supervision and scientific supervision of network lending and lending are carried out by quantitative means.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F724.6;F832.4

【参考文献】