CECL Implementation is Here: What it Means for Financial Institutions
By David St. Yves, Partner, Assurance Services
After several years of delays, Current Expected Credit Loss (CECL) implementation is now right around the corner. In 2016, ASU 2016-13 was issued by the Financial Accounting Standards Board (FASB). It was amended by ASU 2019-10, which adopted the CECL model for credit losses on loans and certain debt securities. This standard applies to private and small public filers for years (and interim periods) beginning after December 15, 2022.
For calendar-year financial institutions, the CECL standard will be adopted on January 1, 2023, and reported in the first quarter call report for 2023. Any initial adjustment required to the allowance for loan losses for CECL adoption can be recorded as an adjustment to equity as of the adoption date. Institutions will likely have to disclose the amount of the adjustment in the 2022 financial statement footnotes.
By now, all financial institutions should have started planning for adoption by evaluating models that are appropriate for their asset size and specific loan portfolios. Many larger institutions have purchased third-party software, such as Abrigo (Sageworks) or Moody’s, to download loan information and perform the analysis and calculations. While these higher-end models can be expensive, the ability to perform the evaluation and calculations for more complex loan portfolios often makes them a cost-effective solution.
For smaller institutions with less complex loan portfolios, there are numerous models from other vendors that are simpler and more affordable. The Federal Reserve published the SCALE model1, which can be a solution. Many of these simplified models use peer group information derived from call reports as the starting point for historical losses. Some institutions have found that the peer group loss information may not represent the institution’s actual historical loss rates on their specific loan portfolio, especially if the institution has not incurred any recent losses.
There can also be inconsistencies in the peer group loss rates on a quarter-to-quarter basis, which produces inconsistent calculation results. This can result in calculations that require quarter adjustments even if there are no charge-offs incurred and no changes in the qualitative factors during the quarter. Another challenge with these simplified models is that the peer group loss information is only broken down by loan type in total. If the institution needs to segregate a specific loan portfolio by risk characteristics, such as industry concentration, credit quality indicator, or credit score, the peer group loss rate may not provide adequate information. Modifying the base calculations (such as by using the average peer group loss rates) can most likely help overcome these challenges.
Institutions should evaluate a couple of alternative models and perform simulations for a quarter or two to determine which model aligns with their portfolio risk and reporting needs. The model should produce results on a quarterly basis that are appropriate for the specific loan portfolio and consistent with changes in the portfolio during the quarter. CECL allows institutions to use different models for different segments of the loan portfolio, so some institutions may find that a simple Excel model produces more accurate and consistent results for smaller, less complex segments (such as consumer loans).
Marcum’s Financial Institutions Services Group members have been actively assisting institutions in evaluating various models. We are available to discuss individual institutions’ implementation process and assist in identifying effective, efficient, and appropriate solutions for CECL implementation.
- More information can be found at https://www.marcumllp.com/insights/federal-reserve-releases-cecl-scale-model