商业银行贷款损失准备前瞻性研究——基于预期损失模型的新证据

Research on Forward-Looking Loan Loss Provisions of Commercial Bank——New Evidence from the Expected Loss Model in China

本文基于中国160家商业银行2013—2020年数据,研究预期损失模型的应用对银行业计提前瞻性的影响。2018年预期损失模型应用以后,运用模型的非上市银行和城商行相对于未应用组的计提前瞻性显著提升,且应用效果在抗风险能力较弱的银行中表现更加突出。前瞻性提升有助于缓解银行的顺周期性风险,但在部分子样本中不显著。机制分析表明,使用预期损失模型的银行管理预期的视角从未来盈利转向未来损失,计提的信号传递动机因此下降。运用模型以前计提前瞻性较弱、信号传递动机较强的银行,运用以后表现出前瞻性显著提升与信号传递动机变弱。因此,银行体系的逆周期调节仍需多政策合力调控。

Summary:Based on a manual collection of annual reports,news releases and policy announcements from various banks,this paper collects data on the implementation of the expected loss model by 160 commercial banks in China from 2013 to 2020.The empirical results indicate that after the adoption of the new model,the overall forward-looking practice of China's commercial banks has significantly improved,with non-listed banks and urban commercial banks showing sizable improvement in their foresight.Moreover,the effectiveness of this model is more prominent in banks with weaker risk resilience,which confirms that the new model has achieved some of its goals.At the same time,the improvement of foresight plays a role in mitigating the procyclicality of bank risk management,although it is not significant in some sub-samples.Further analysis shows that after the implementation of the new model,the expectation management of banks began to shift from future profits to future losses,thereby reducing the signal transmission motivation in loan loss provisions and affecting procyclicality.Banks with initially weak foresight and strong signal transmission motivation prior to the policy change witnessed a substantial improvement in foresight and a subsequent decrease in the signal transmission motivation.Additionally,this paper also conduct robustness tests using staggered DID dynamic effect test,sample replacement,event time replacements,explanatory variable replacements,placebo test and the inclusion of additional control variables.All of these tests consistently support the main findings.Our findings have important policy implications.Firstly,regulatory agencies should provide policy guidance to local small and medium-sized banks.The expected loss model,originally scheduled to be implemented among all commercial banks in 2021,has not yet been fully realized.The main reason is that local banks lack sufficient risk management and prediction capabilities.Secondly,it's essential to refine the specific terms of the expected loss model and adopted a multi-layered approach to enhance feasibility.Without clear and detailed guidelines,banks often have too much autonomy in the process of model implementation,and different types of banks have varying talent reserves and technical support in risk management.Finally,it is necessary to have a correct understanding of the relationship between forward-looking risk management and the accounting motivations behind loan loss provisions,and to address the limitations of countercyclical regulation.Achieving the goal of countercyclical banking regulation still requires more policy synergy.