Accessible Version
The Empirical Relationship Between the Tier 1 Leverage Ratio and Community Bank Failures after the Global Financial Crisis, Accessible Data
Figure 1. Quarterly number of bank failures in the United States in the 21st century
This figure shows the number of bank failures in the United States in each quarter from Q1 2000 to Q2 2025. Two vertical red lines mark the beginning and end of the period from Q1 2008 to Q4 2012 that this analysis uses to model bank failures. Out of the 7,184 U.S. community banks that existed at the end of 2007, 389 (or about 5.4%) failed over this five-year period, which creates a large spike in the number of quarterly bank failures in the figure. Outside of this five-year period, the number of quarterly bank failures was generally under 10 and, on average, about 1.
Note: This figure shows the number of bank failures in the United States in each quarter from Q1 2000 to Q2 2025. The two vertical red lines mark the beginning and end of the period from Q1 2008 to Q4 2012 that this analysis uses to model bank failures. Out of the 7,184 U.S. community banks that existed at the end of 2007, 389 (or about 5.4%) failed over this five-year period.
Source: The Federal Deposit Insurance Corporation’s list of historical bank failures.
Figure 2. Estimated annualized probability of failure across all community banks (as of Q4 2007)
This figure shows the average annualized probability of failure across community banks in the sample as a function of their tier 1 leverage ratios, estimated using the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the annualized probability of failure over the five-year period from Q1 2008 to Q4 2012, using financial characteristics reported by 6,365 community banks at the end of Q4 2007. The figure shows a monotonically decreasing convex function for the annualized probability of failure starting at about 2.7% in the top left and ending at about 1.1% in the bottom right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis.
Note: This figure shows the average annualized probability of failure across community banks in the sample as a function of their tier 1 leverage ratios, estimated using the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the annualized probability of failure over the five-year period from Q1 2008 to Q4 2012, using financial characteristics reported by 6,365 community banks at the end of Q4 2007.
Source: Model-based estimates obtained using FFIEC call report data.
Figure 3. Estimated probabilities of failure for banks with different failure risks (as of Q4 2007)
This figure shows the average probability of failure across all, low-risk, and high-risk community banks in the sample as a function of their tier 1 leverage ratios, estimated using the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the probability of failure over the five-year period from Q1 2008 to Q4 2012, using financial characteristics reported by 6,365 community banks at the end of Q4 2007. The figure shows three monotonically decreasing convex functions for the annualized probability of failure, corresponding to all, low-risk, and high-risk community banks in the sample, where "low-risk banks" and "high-risks banks" are defined as the banks with a lower-than-average and higher-than-average estimated probability of failure, respectively. The function for high-risk community banks is the highest, starting at about 10% in the top left and ending at about 4% on the right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis. The function for low-risk community banks is the lowest, starting at about 1% in the bottom left and ending at about 0% in the bottom right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis. The function for community banks in the sample is in between the two other functions, starting from about 2.7% on the left and ending at about 1.1% on the right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis. (This function for all banks is identical to the one shown in Figure 1.)
Note: This figure shows the average probability of failure across all, low-risk, and high-risk community banks in the sample as a function of their tier 1 leverage ratios, estimated using the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the probability of failure over the five-year period from Q1 2008 to Q4 2012, using financial characteristics reported by 6,365 community banks at the end of Q4 2007.
Source: Model-based estimates obtained using FFIEC call report data.
Figure 4. Estimated probabilities of failure for community banks using 2007 and 2024 data
This figure shows the average probability of failure across community banks as a function of their tier 1 leverage ratios, estimated using their financial characteristics as of Q4 2007 and Q4 2024 in the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the probability of failure over the five-year period from Q1 2008 to Q4 2012. There are 6,365 and 3,436 community banks in the sample in Q4 2007 and Q4 2024, respectively. The figure shows two monotonically decreasing convex functions for the annualized probability of failure: one estimated using bank fundamentals as of Q4 2007, and another using bank fundamentals as of Q4 2024. The function for Q4 2007 is higher, starting at about 2.3% in the top left and ending at about 1.3% on the right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis. The function for Q4 2024 is lower, starting at about 0.6% in the bottom left and ending at about 0.3% in the bottom right as the tier 1 leverage ratio goes from 0% to 15% on the horizontal axis.
Note: This figure shows the average probability of failure across community banks as a function of their tier 1 leverage ratios, estimated using their financial characteristics as of Q4 2007 and Q4 2024 in the logistic model specified in Equation (1), including quadratic terms and state fixed effects. The model estimates the probability of failure over the five-year period from Q1 2008 to Q4 2012. There are 6,365 and 3,436 community banks in the sample in Q4 2007 and Q4 2024, respectively.
Source: Model-based estimates obtained using FFIEC call report data.