中国房地产信贷网络拓扑特征及网络鲁棒性研究

发布时间：2018-04-17 08:28

本文选题：房地产信贷网络 + 复杂网络　；参考：《天津大学》2014年硕士论文

【摘要】：中国房地产金融是以房地产为载体,以房地产企业和金融机构为直接主体、政府为间接主体,涉及到生产、流通、消费各环节,影响到经济、社会、生态环境各方面的复杂巨系统。宏观层面,改善中国房地产信贷结构和优化信贷资源配置,微观层面,形成房地产企业与金融机构相互促进、共生共赢的状态,对于中国房地产业发展有着至关重要的影响。房地产信贷结构本质上反映的是房地产金融资源配置,而宏观结构是微观个体行为的涌现,所以研究房地产金融这个复杂巨系统中的房地产企业与金融机构之间的信贷关系,是优化房地产金融市场信贷资源配置、提高房地产信贷资源利用效率的内在要求。本文利用复杂网络分析方法,以2007年至2013年沪深A股上市房地产公司为样本,基于房地产长期借款信息构建房地产金融机构二分网络。对该二分网络进行加权投影得到以金融机构为网络节点的无向含权的单顶点网络,即中国房地产信贷网络。从信贷网络的拓扑统计特征和网络鲁棒性两个方面,研究中国房地产信贷网络的演化规律和抗冲击能力。在信贷网络的拓扑统计特征方面,主要分析了小世界效应、无标度特性、度的相关性和网络中心性。实证研究发现:2007-2013年间房地产信贷网络是小世界网络,但不具有无标度特性。随着网络规模的扩大,新增网络节点的连接具有异配性。同时国有五大行及部分股份制银行在房地产信贷网络中占据枢纽地位,对于网络稳定性有重要作用。在信贷网络鲁棒性方面,研究随机攻击和选择性攻击两种冲击下,中国房地产信贷网络的最大成分W、聚集系数C和平均路径L变化情况。实证研究发现:中国房地产信贷网络整体鲁棒性较差,选择性攻击对中国房地产信贷网络的破坏明显大于随机攻击。网络视角分析房地产信贷网络结构拓扑特性和网络鲁棒性,有助于在房地产金融市场搭建微观和宏观之间沟通的桥梁,同时开拓了房地产信贷领域新的研究范式。
[Abstract]:China's real estate finance is based on real estate as the carrier, real estate enterprises and financial institutions as the direct main body, the government as the indirect main body, involving production, circulation and consumption, affecting the economy and society.Complex giant system in all aspects of ecological environment.At the macro level, improving China's real estate credit structure and optimizing the allocation of credit resources, micro level, the formation of real estate enterprises and financial institutions to promote each other, symbiosis and win-win state, for the development of China's real estate industry has a vital impact.The real estate credit structure essentially reflects the allocation of real estate financial resources, while the macro structure is the emergence of micro individual behavior. Therefore, we study the credit relationship between real estate enterprises and financial institutions in the complex huge system of real estate finance.It is the inherent requirement of optimizing the allocation of credit resources in real estate financial market and improving the utilization efficiency of real estate credit resources.This paper uses the complex network analysis method, taking Shanghai and Shenzhen A-share listed real estate companies as samples from 2007 to 2013, based on the long-term loan information of real estate to construct a two-part network of real estate financial institutions.Based on the weighted projection of the binary network, an undirected weighted single vertex network with financial institutions as the network node is obtained, that is, the real estate credit network in China.In this paper, the evolution and impact resistance of Chinese real estate credit network are studied from two aspects: topological statistical characteristics and network robustness.In terms of the topological statistical characteristics of credit networks, the small-world effect, scale-free property, correlation of degree and network centrality are analyzed.Empirical research shows that the real estate credit network between 2007-2013 is a small world network, but not scale-free.With the expansion of network scale, the connection of new network nodes has heterogeneity.At the same time, the five state-owned banks and some joint-stock banks occupy a pivotal position in the real estate credit network, which plays an important role in the stability of the network.In terms of the robustness of the credit network, the changes of the largest component of the real estate credit network in China under random attack and selective attack are studied, and the aggregation coefficient C and the average path L are also studied.The empirical study shows that the overall robustness of China's real estate credit network is poor, and the damage to China's real estate credit network caused by selective attacks is obviously greater than that of random attacks.The network perspective analyzes the topological characteristics and network robustness of real estate credit network structure, which is helpful to build a bridge between micro and macro in the real estate financial market, and opens up a new research paradigm in the field of real estate credit.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.4;F299.23

【相似文献】