This study estimates the natural rate of bank defaults, the threshold below which systemic banking crises are unlikely, using a threshold model based on bank default rates and macroeconomic indicators. Analyzing global data from major crises over the past 40 years, we identify a critical default rate of 0.25 %, equivalent to one default per 400 banks annually. Aligned with a 'BBB' rating, this benchmark supports the calibration of supervisory risk tolerance frameworks. Moreover, the study provides a replicable, data-driven approach to prudential policy design, linking acceptable bank failure frequency to key macroeconomic variables, especially inflation indicators.

中文翻译：

---

审慎政策的界限在哪里？洞察银行稳定性和风险承受能力

本研究使用基于银行违约率和宏观经济指标的阈值模型来估计银行违约的自然率，低于该阈值，系统性银行危机的可能性不大。通过分析过去 40 年重大危机的全球数据，我们确定了 0.25% 的关键违约率，相当于每年每 400 家银行中就有 1 家违约。该基准与“BBB”评级一致，支持监管风险容忍度框架的校准。此外，该研究提供了一种可复制的、数据驱动的审慎政策设计方法，将可接受的银行倒闭频率与关键的宏观经济变量（尤其是通胀指标）联系起来。