基于复杂网络的商业银行网点布局特征——以北京市中国银行为例

Characteristics of commercial bank branch networks based on complex networks theory: A case study on Bank of China in Beijing

以北京城区6个行政区内分布的中国银行网点为例，利用复杂网络理论，按照就近原则，分别构建服务半径400m和800m的网点分布复杂网络模型，并结合GIS空间可视化技术对其进行网络特征分析与解释。结果表明：①银行网点结构多呈现扁平化，联系稀疏，具有小世界网络特征。由于缩短了网络的特征路径长度，增强了聚类系数，在提升网点效率的同时，也提升了市场竞争的激烈程度。②随着服务半径的扩大，银行网点集聚程度增强，而稳定性降低。银行网点的最佳服务半径在400～800m，步行可达的网点网络比公交可达的网点更稳定。③银行网点分布存在区域差异性。一方面表现为高等级节点在局部区域出现高集聚现象，主要位于四环以内，并集中分布在金融街、三元桥和国贸CBD一带；另一方面表现为节点度之和在街道层面存在较大差异，内城经济发达的街道，其网点多且节点度高，而外围街道，特别是近郊外围街道网点分布少，或者没有网点，存在金融排斥现象。未来在银行网点规划中，可重点培育一些边缘节点，发挥其带动作用，扩大网点覆盖范围，完善整个区域网点服务职能。

Complex network theory, combined with social network technology, has been more and more widely used in the researches and applications related to realistically existing large irregular networks to find the rules of network operations and to improve operation efficiency. In human geography research, using complex net- work technology to analyze city economic activities can help improve the analysis and interpretation of the pat- tem, and the change of the pattern, of the distribution of geographic phenomena to a certain extent, and can be applied to the studies of the essential characteristics of geographic networks. In this paper, taking Bank of China as an example, using the theory of complex network and with the principle of proximity, based on data of bank branch distribution networks in the six administrative districts of Beijing, we constructed a complex network model of bank branch distribution networks with the service radius of 400 meters and 800 meters, and used GIS spatial visualization technology to analyze and interpret the network characteristics. The research shows that： （1） the bank branch networks showed a fiat structure with sparsity and small world network characteristics. The flat management pattern of the bank branch networks shortens the length of typical routes, which enhances the clus- tering coefficient, improves network efficiency, and at the same time increases intensifies the degree of the mar- ket competition. （2） The convergence of the layout of bank branch distribution increases with the expansion of the service radius, and stability decreases. For bank branch spots in the network, best service radius is 400 me- ters to 800 meters; the network spots accessible by walking are more stable than the network spots accessible by public transportations. Specifically, for the bank branches with service radius less than 400 meters, since the sharing of the customers is weak, and also faced with the challenges of increasing land rent and intensifying competition, they should community-focused bank branches in order to enhance the cohesion and competitive- ness of each community area. However, for the bank branches with service radius less than 800 meters, the in- crease of the length of typical routes causes the increase of the cost of the process and the decrease of operation efficiency. Therefore, the bank branches should diversify service features, improve the area＇s networking degree, enhance the survivability, and improve stability. （3） The bank branch distribution exhibit area differences. The high level nodes are showing high level of agglomeration in the areas within the 4th beltway, mainly concentrat- ed in Financial Street, Sanyuanqiao, China World Trade Center, and CBD area. On the other hand, the node de- grees show big differences at the street level. For example, on the streets of well-developed inner city areas, the network and node degree are high, while on the other streets, especially the streets close to suburban areas, there few bank branch networks, or even no nodes, indicative of financial exclusion. In the future the network plan- ning should focus on cultivating a number of edge nodes to play leading roles, expanding the network coverage, and improving the area network service functions.