关键词: 商业银行, 经营效率, 复杂DN-SBM-DEA模型, Kruskal-Wallis秩和检验

Abstract: Reasonably measuring the internal operational efficiency of commercial banks and exploring its internal reasons for operational inefficiency are of great practical significance for promoting the sustainable development of China’s banking industry. This article divides the internal business process of commercial banks into three stages, namely deposit absorption, debt allocation, and asset profitability from the perspective of sustainability, liquidity, safety, and profitability in business activities. Then, we construct a complex network structure for operating process of commercial banks, and a complex DN-SBM-DEA (Dynamic Network SBM-DEA) model. Taking 44 domestic and foreign commercial banks in China mainland from 2013 to 2017 as research samples, this article mainly draws the following conclusions: (1)Compared with the SBM-DEA model and NSBM-DEA model, the DN-SBM-DEA model considers both the internal processes of bank operations and the lagged effects of carryover variables, so the measurement results are more reasonable. (2)With the continuous expansion of openness in China’s banking industry, the overall operational efficiency of 44 domestic and foreign commercial banks in China mainland in 2017 slightly decreased compared to 2013. This is mainly due to the decrease in efficiency during the asset profit stage, which is related to the rise in non-performing loan ratios in the banking industry in recent years, due to the downward pressure on China’s economic environment. (3)Chinese banks have a relatively stronger ability to absorb deposits and allocate assets, which is mainly due to insufficient asset profitability. However, foreign banks have relatively stronger an asset profitability, which is mainly due to insufficient deposit absorption and asset allocation capabilities.
Compared with the previous studies, the network structure constructed in this article is more closely related to the internal processes of bank operations, and in the third stage, it considers the impact of non-performing loans, an unexpected output, on asset profitability, making it more reasonable. In addition,previous studies have mostly shared the investment of the first stage in the second or third stage, but the investment of the first stage in the second or third stage is not fully utilized, and the true utilization rate is difficult to know. If treated directly in a fully utilized manner, it may lead to incorrect evaluation results. This article does not directly share the inputs of the first stage in the second or third stages, but rather selects the corresponding input and output variables reasonably based on the characteristics of the stage. However, this article adopts the mainstream approach of setting stage weights and period weights, which means that the importance of each stage (i.e., weight) is considered the same, and the importance of each period (i.e., weight) is also considered the same. This approach has certain limitations in practice, for example, in the business process of commercial banks, the importance of the three stages of deposit absorption, debt allocation, and asset profitability may not necessarily be the same. Therefore, the future expansion direction is how to determine reasonable stage weights and period weights.

Key words: commercial banks, operational efficiency, DN-SBM-DEA model, Kruskal-Wallis rank sum test