There is no single line item in bank financial statements getting more attention these days than the loan loss reserve, due to both CECL and COVID-19.  With the second quarter of 2020 now in the books, we decided to use our BankGenome™ intelligence system to analyze how banks with less than $50 billion in assets have handled their loan loss reserves through the end of the June quarter.

This article summarizes our findings. And good news: To help your bank going forward in this uncertain environment, we are making the data and analytics available via a free download. We built a tool that allows you to plug in your bank (using your RSSD number from the front page of the Call Report) to see how it ranks in your state and among your peers.

Our study focuses on the “Reserve Rate,” which we are defining as the loan loss reserve as a percentage of the loan balance. Please note that we have adjusted the loan balance to exclude both held-for sale loans and PPP loans. Our most interesting findings:

1. Loan loss reserves for CECL banks are not that much higher than those banks using the incurred loss method. The median Reserve Rate for the 164 community banks and regional banks that adopted CECL was 1.48 percent. This is only 17 basis points or 13 percent higher than the Reserve Rate of 1.31 percent for the 4,788 banks using the incurred loss method. Several likely reasons account for the lack of a material difference. Community and regional banks still adhering to the incurred loss method are far more likely to be private, so they are not feeling the same pressure from the public markets to maintain a dividend policy or manage their quarterly earnings. They also tend to be far more conservative with their loan loss reserving in general. I also think that the lines are blurring between an “expected” loss and a “probable” loss due to COVID-19. Many banks still adhering to the incurred loss method have been increasing their reserve via their qualitative factors, specifically the “economic q-factor”. Many banks are even using their stress tests to help them do so, under the assumption that the losses likely exist in their portfolio, but they just don’t know exactly which loans are problematic because the pandemic-fueled losses haven’t yet occurred. Nevertheless, it is surprising that the difference in the median reserve is so insignificant.
2. But when we look closer at a breakdown by bank asset size, the variance widens between the incurred loss filers and the CECL filers. For example, the median Reserve Rate for a CECL filer with between $10 billion and $50 billion in assets was 1.54 percent, which is 28 basis points greater than the Reserve Rate for an incurred loss filer in the same asset class. A similar trend occurs with banks between $5 billion and $10 billion in assets. The Reserve Rate for CECL filers is 1.53 percent versus 1.2 percent for incurred loss filers in this bracket. The trend reverses when looking at smaller banks. Smaller CECL filers had lower Reserve Rates, but smaller incurred loss filers had higher Reserve Rates. 

3. The median ACL reserve for a CECL filer has increased by a whopping 83 percent compared to its ALLL reserve at the end of 2019. At first, this finding feels completely contradictory to the first finding mentioned above.  However, it all becomes clear when we look at the starting point at the end of 2019:

Amazingly, CECL filers ended last year with a much lower loan loss reserve rate than the incurred loss filers. In a sense, the increase in their reserve was one way to catch up to the incurred loss filers. We should NOT mistake the CECL filers as being under-reserved as of the end of 2019. It was more likely that an incurred loss filer was over-reserved at the end of last year prior to any inkling of the pandemic. Banks simply do not have many losses in good times. The ALLL is supposed to be a proxy for probable losses, and probable losses should emerge within one to two years. You would have to add up 5 to 10 years of net-charge offs before you would reach the loan loss reserves for most of the incurred loss filers at the end of 2019. Don’t forget that many of the CECL filers were active in M&A. Under the incurred loss method, purchase accounting did not require a significant ‘on balance sheet’ reserve against acquired loans.  However, CECL irrationally includes the ‘double dip’ concept, which requires banks to carry a loan loss reserve for acquired loans in addition to the purchase accounting marks. Makes no sense, but this has forced these banks to increase their reserves substantially.

COVID-19 has also had a greater impact on the CECL filers, and in many ways, they will have an easier time justifying the increase in their reserves to auditors. If a bank is following the letter of the law with respect to generally accepted accounting principles it would not have to technically wait for a loss to become probable. In fact, this concept mostly likely explains why 164 CECL filers rejected the opportunity to defer compliance with CECL. They were looking for reasons to support their reserve with more quantification, and in many cases, recognized that “the street” was more interested in their loan loss reserve than their earnings.

4. CECL filers increased their reserves more aggressively than incurred loss filers in the June quarter relative to the March quarter. CECL filers increased their reserves by 19 basis points compared with only 6 basis points for ACL filers. The second quarter of 2020 is the “calm before the storm” quarter. The losses have not happened yet, but they are coming, likely later this year and into next year. What we do not know yet is the magnitude of those losses. The CECL filers continued to increase their reserve rather aggressively in the second quarter. Many incurred loss filers, more likely to be over-reserved at the end of last year, are using that cushion to grow their reserves at a slower rate.
5. Nearly 1 out of 4 Incurred Loss Filers decreased their Reserve Rate in the June quarter, versus less than 1 out of every 10 CECL filers. This does not mean that these banks recorded negative provisions, nor did they experience massive charge-offs that exceeded their reserve build. In fact, most of the 1,104 incurred loss filers that reduced their Reserve Rate did so because loan growth (excluding PPP loans) exceeded their reserve growth via provision expensing. In some cases, this was likely unintentional as smaller, private banks have the habit of recording the same provision expense (say $300K) every quarter. It is an interesting time to have this happen, but some banks can support this with their stress testing, even if this was more likely to occur by accident than by design. Nevertheless the regulators are watching, and those banks should be prepared to support and defend their reserves.

The loan loss reserve will continue to garner attention from all stakeholders as we move forward.  Modified loans will hit their 180-day check point in late September and early October, which could cause a spike in classified loans. As banks file their Call Reports and 10-Qs in late October, we will know a lot more.

To find out how your bank’s Reserve Rate compares to other banks, click here to download our BankGenome-powered™ excel tool.