Litmus Testing Your Allowance for Loan and Lease Losses (ALLL)
March 28, 2011
By: Dave A. Anderson
The ALLL level and methodology mechanics are common criticisms in regulatory examination reports during these difficult times of excessive past due and nonaccrual loans and commercial real estate concentrations. If regulators believe your ALLL is underfunded, it can lead to revising your Call report, wiping out your income for the year, negatively impacting your "Management" component rating, and finally, total frustration.
Accountants and regulators will tell you that if you use the appropriate ASC 310 (formerly FAS 114) and ASC 450 (formerly FAS 5) measurement methods, your ALLL balance should be adequate. But the real truth is that you have to start with what the regulators are really looking for, and then justify that balance using the proper ASC 310 and ASC 450 mechanics. A word of warning, it is probably not wise to use these litmus/reasonableness tests in your ALLL methodology, but attaching them as addendums shouldn't be a problem and could assist in convincing them that your ALLL is adequately funded. These tests are only going to tell you if you are in the ball park and may prevent you from hearing this from the regulators at your next exam, "We believe, based on the inherent risk in your loan portfolio, the ALLL is underfunded by at least $XX million". Of course that night you can't sleep, you fear for your investment or job, and you look at yourself in the mirror and think....they can't be right!
What ALLL balance are they expecting? First, let's start with the most important litmus test ratio, because this ratio is what the supervisory staff in Washington look at in their frenzied, surveillance driven jobs. Try to explain any reason for an aberration to a Washington type of why their coveted ratio may be inaccurate results in examiner frustration, since most examiners won't be able to, or willing to, explain why your bank's ratio relative to the ALLL balance is justifiable when it is so low. It has something to do with examiner job security, pay raises, and promotions...you get the point. The second reasonableness test will be familiar to most seasoned bankers, and the third test, while probably the best, requires specific data from every bank in the U.S. making it difficult to conduct by most community banks.
To initially introduce the first litmus test I want to explain that the applicable variables involve the level of your nonperforming loans relative to the balance of your ALLL. So, turn to page 7 of your UBPR and let's assume you don't have any loans that are over 90 days past due and still accruing interest. If you do, be able to explain why, but it really won't make a difference in our first litmus test, because these loans (90 + and accruing) need to be used in this test. As of 12/31/10 it is the 9th ratio on page 7 and is titled "LN&LS Allowance to Nonaccrual LN&LS". Let's assume you do have some loans that are more than 90 days past due and still accruing. Turn to page 8a of the UBPR and note a ratio about two thirds of the way down of the schedule under "Other Pertinent Ratios" titled "Non-Curr LN&LS to - LN&LS Allowance". If the aforementioned ratio on page 7 is over 40 percent (assuming no loans 90+ and accruing) and the ratio noted on page 8a is under 250 percent, you have passed the first and most important litmus test. However, if you have a large volume of loans 30-89 days past due or special mention/watch loans, you may want to increase this litmus test ratio threshold by 5 to 10 percent.
We will use a methodology most seasoned bankers are familiar with as our second litmus test. Historically bankers have used many different ratios for each of these categories, but I am going to use the ratios I believe were most commonly used prior to the implementation and enforcement of FASB statements 5 and 114. Get out your watch list, current balance sheet, and calculator. Assuming all loss in your loan portfolio is charged off prior to the examiner's "Asset Quality Date" (hint), take 50 percent of the balance of your internally classified doubtful loans, 15 to 20 percent of your substandard loans (depending on collateral protection), 3 percent of your watch or special mention loans, and 1 percent of all other loans. Don't tell me CD secured loans are without risk, because they may not be when the taxing authorities discover a CD exists at your bank, and your borrower hasn't been paying some of its taxes. If your ALLL balance is in excess of the total of these amounts, you have passed litmus test number two.
The third litmus test is my favorite, but the most difficult to calculate unless you have access to software/data available from a vendor. The regulators have access to this data/software and use it to prove your ALLL is underfunded. The FFIEC website may have the tools capable of helping you conduct this test if you have a staff member who is technological savvy by using available information at https://cdr.ffiec.gov/public/PWS/DownloadBulkData.aspx. This example will best explain the conclusion you are attempting to reach. It involves comparing specific bank metrics of your bank's credit/asset quality/capital condition to those similar to yours, and then comparing those results to the ratio I always dislike using, but helpful in this exercise, ALLL/ total loans. I would always refrain from using this simple ratio in justifying your ALLL, since it usually has little relevance when considering your credit risk profile compared to the general population of banks. So what we are attempting to accomplish in this test is to find banks in the U.S. with a similar range of metrics and compare their ALLL/total loans ratio to yours.
Let's assume your bank has some asset quality problems, a CRE concentration, and well, it could use more capital. As of 12/31/2010 some key ratios at your bank were:
Total risk-based capital: 10.10% (9.60% to 10.60%)
Nonperforming (NA + 90 days past due and still accruing): 7.20% (6.00% to 8.40%)
CRE concentration: 450% (340% to 560%)
Construction & Land Development concentration: 210% (160% to 260%)
ALLL/total loans: > 0% (this ratio is what we are seeking, so we want all banks in the general population that meet all the other criteria listed above).
So, let's run some queries which may require the help of a vendor or technological savvy staff member. Our query will include a range of the following ratios listed above in parentheses, and then we will key in on the ALLL/total loans ratio to see if you pass this test. Remember, the ranges listed above that we will query will help us find banks in our similar predicament. If your bank's ALLL/total loans ratio is in in the lower 20% of banks with similar metrics, you have failed this important litmus test.
So there you have it, fail all three of the tests and rest assured your ALLL will be deemed underfunded with the regulators typically pointing to inadequate ASC 450 allocations as the reason why. If you fail two tests, chances are less that the regulators will think the ALLL is underfunded. Fail only one (and you have several loans moving into OREO in the next two to three months), you may be safe.
Remember, these are only litmus tests, but it's worth your time and effort to understand them and realize how the regulators rely on them as benchmarks/reasonableness tests to measure the adequacy of your bank's ALLL. There are always exceptions, but as you all know, determining an appropriate ALLL level is not an exact science.
To be added to our email list, email Jim@bankstrategiesllc.com with your name and email information.