CECL: A Half-Baked Cake

I began my career auditing and consulting with credit unions in 1977 and vividly recall my mentor, Gene O’Rourke, teaching me allowance for loan losses (ALL) theory in those early days. Credit unions had adopted ALL accounting procedures a few years earlier and were still steep in the learning curve for this important accounting concept. Before adoption of the ALL, credit unions simply charged-off uncollectible loans directly to the regular reserve, with no impact to earnings. Yes, that was the major purpose of the regular reserve, which became pretty useless subsequent to ALL accounting.

Over the course of my career, credit unions — and all banking institutions — often found themselves having to revise their ALL methodology based on economic conditions and underwriting quality issues that became apparent as loan portfolios seasoned. These economic events played havoc with earnings; further, since ALL accounting is somewhat subjective, it wasn’t unusual to see some level of earnings management result from the ALL methodology implemented.

Fast forward to 2024. The current expected credit loss model (CECL) has totally changed the method by which credit unions measure loan losses. Credit unions adopted CECL throughout 2023, and virtually all had implemented it by Dec. 31. Now that the first quarter 2024 call report data is available, some important metrics provide a critical perspective on adoption methodology.

ALL Versus CECL Basic Theory

ALL theory — also referred to as the “incurred loss model” — essentially requires institutions to recognize loan losses when loan impairment characteristics have already developed. The balance of the ALL is not a measure of total lifetime loan impairment.

From a practical standpoint, credit unions calculated a historical loss ratio as a proxy for impaired loans. For example, if net charge-offs averaged 0.5% of loans outstanding for the past several years, that same ratio was applied to the balance of the portfolio. A $100 million portfolio might have had an ALL balance of approximately $500,000. That same portfolio would experience lifetime losses significantly exceeding this amount.

Under CECL standards, the allowance for credit losses (ACL) must measure the amount of expected lifetime losses in the portfolio regardless of whether an impairment event has already occurred. Lifetime losses result in an amount much greater than what would be derived using the loss ratio approach described above.

The magnitude of the difference between ALL versus ACL is highly dependent on the composition of the loan portfolio. For example, if a credit union’s loan portfolio was composed mostly of consumer loans — auto and credit card — and small amounts of residential real estate, then the lifetime losses of an existing portfolio could easily be two to three times greater than 0.5% (1.0% to 1.5%). The previously calculated ALL balance of $100,000 could increase to $300,000 or more under CECL, depending on other economic trends. However, if the portfolio were highly concentrated in residential real estate loans with minimal loss exposure, the difference between ALL and CECL would be comparatively lower.

Table 1 summarizes total ALL balances by peer group as of Dec. 31, 2022 — the last ALL reporting period before credit unions began to adopt CECL — to ACL balances as of March 31, 2024. This table indicates the largest credit unions realized a much more significant increase in their ACL balance as compared to the smaller credit unions. But as noted in Tables 3 and 4, delinquency and charge-offs for the larger credit unions increased by significant amounts in the first quarter of 2024, perhaps accounting for much of this difference.

TABLE 1: TOTAL ALL/ACL BALANCE

Setting Basic Expectations

Based on the above basic understanding of ALL versus CECL standards, the following expectations are reasonable.

TABLE 2: EXPECTATIONS AND RESULTS

Rising Delinquency And Charge-Off Impacts CECL Calculation

It is important to note both delinquency and charge-off trends have increased since the inception of CECL accounting as noted in Graph 1 and 2 below. Of particular interest is the significant rise in both delinquency and charge-off levels for credit unions with more than $20 billion in assets.

Given the rise in both of these metrics, it is reasonable to assume ACL balances should have increased since CECL inception, assuming no major changes in the amount or composition of the loan portfolio. As noted in Graph 3, it does appear the ACL has increased since CECL inception, so from a directional perspective, this expectation of increasing ACL has been achieved. But this provides directional perspective only, and the magnitude of the increase is not reflected in this metric.

ACL Delinquency Coverage Ratio

This ratio measures the amount of coverage for delinquent loans inherent in the ACL balance (see Graph 4). The delinquency coverage ratio could be misleading when a credit union reports past due residential real estate loans that are well secured and don’t represent loss exposure. Conversely, large amounts of unsecured loans that are past due with little or virtually no collateral protection versus large amounts of auto loans that have reasonable collateral protection could have differing results.

ACL Net Charge-Off Coverage Ratio

The ratio of the ACL balance as a percentage of annualized net charge-offs is one of the most critical metrics every credit union should be measuring. As noted in Graph 5, this ratio is in steep decline for all peer groups and averages approximately 160% for all peer groups displayed. In other words, the average credit union portrayed in this graph has an ACL balance that can absorb 1.6 times the current amount of annual net charge-offs. This ratio was approximately 240% as of Dec. 31, 2023.

On the surface, this ratio looks low and the negative directional change could be indicative of an understated ACL balance.

The Averages Don’t Tell The Full Story

There are 296 credit unions with assets greater than $100 million in the United States that report an ACL net charge-off coverage ratio of less than 100%. Unless there is something unusual in the current charge-off levels, this low coverage ratio is very problematic and likely to be challenged by auditors and regulators.

There are 258 credit unions in the United States with assets greater than $100 million that report an ACL net charge-off coverage ratio of greater than 500%. This could be indicative of excess reserves that could lead to earnings management.

On the surface, the aforementioned comparatively low and high coverage ratios appear significantly out of line with the averages noted in Graph 5. There are many possible explanations for these variances, but those credit unions that are far outside the average range should review their methodology to ensure they comply with applicable accounting standards.

Is Your CECL Strategy Fully Baked? Dig into your loan performance and asset quality trends to determine if your credit union has enough reserved in its allowance for credit losses. Callahan & Associates can help you perfect your performance research, identify areas of improvement, and prepare for exams. Request a free performance scorecard session diving into the metrics of your choice today. I Want My Free Scorecard.

Other CECL Pitfalls

Credit union management and those charged with governance should be aware of the following:

1. The amount of ACL related to residential real estate loans should be carefully analyzed. Many credit unions have significant portfolios of such loans that reflect very low current loan/value metrics (LTV) and therefore minimal loss exposure. Such portfolios should be analyzed to ensure loans with low LTVs have been segregated from the higher LTV loans for loss exposure purposes.
2. Member business loans and other commercial loans should be carefully analyzed and segregated by loan type. For example, MBLs secured by office buildings have much different risk characteristics than a multi-family residential building.
3. Those credit unions that purchase participation loans from other credit unions should inquire of the lead lender what ACL metrics are applied to the residual portfolio held by the lead lender. Significant differences between the lead lender CECL requirement versus the participant credit union should be investigated.
4. Credit unions should ensure loans are charged-off in a timely manner and the credit union has appropriate title to any collateral securing the loan.
5. Credit unions should have regular meetings with their CECL vendors to ensure they understand the complexities of the model used and to take advantage of any insights the vendor has on best practices for their respective CECL systems.
6. Credit unions should ensure appropriate Q&E factors are considered for loss exposure purposes. For example, if there was a significant forecasted drop in residential real estate values, such a forecast should be considered in the determination of lifetime losses on this slice of the loan portfolio.

As with any new accounting standard, various compliance complexities will continue to surface. These complexities will be compounded by changing economic circumstances that will provide much-needed perspective on the efficacy of the lifetime loss estimates used in the early days of adoption.

It is incumbent on management to continuously evaluate such loss estimates, and — where significant changes to such estimates result — determine whether such changes are indicative of an insufficient evaluation methodology that requires modification or of changing economic conditions that were beyond management’s ability to estimate in the normal course of events.

Michael Sacher is a seasoned credit union accountant and advisor, having spent his entire career in the credit union sector. He continues to consult with credit unions in various areas such as CECL and is also available for temporary CFO assignments. Mike can be reached at mike@sacherconsulting.com   (310) 880-5323