From credit risk analysts to modelers, the credit oversight functions at community banks under $1 billion in assets are working to enhance CECL reserve methodologies and modeling approaches before January 2023, when they must officially use the latest accounting standards that recognize expected loan losses over the maturity of a loan.
To understand the state of CECL adoption at community banks, the RMA Model Validation team developed the Community Bank CECL Range of Practice Benchmarking Survey. The survey was distributed to chief credit officers, chief risk officers, and chief financial officers of community banks across the United States in February and March 2022.
One hundred and three banks below $10 billion in total assets as of December 31, 2021, answered the survey, although not all participants completed every question. Banks below $1 billion in total assets represented over two-thirds of survey respondents.
Here are some key takeaways from the survey:
CECL models are now top of mind. The survey results reflect the urgency of CECL model validations among community banks. Over 70% of respondents said they plan to validate their CECL modeling framework by the end of 2022. Most community banks are conducting parallel runs of their CECL models now to evaluate the impact of the transition from incurred loss (ALLL) to expected loss (CECL). As these banks prepare to adopt CECL, four out of five rely on vendors for data, modeling, and support explaining results in regulatory reporting.
The majority of banks spend over two months to complete full CECL model validation. Fifty-five percent of respondents said the validation process took two to four months or longer. Specifically, about 42% of the respondents said they are conducting parallel runs and validations for CECL processes. Parallel runs and stress-testing in model validation means the process of testing the model to eliminate potential issues and errors. Given the complexity, the time, and cost associated with this transition, it would be beneficial for banks to start the validation process as early as possible.
Community banks most often use the WARM method. Participants said their banks use various calculation methodologies to estimate expected credit losses. Twenty-nine percent of community banks use the weighted average remaining maturity (WARM) method to determine credit loss estimates for at least one portfolio segment. Banks also use PD / LGD and discounted cash flow methods for their loan loss calculations. Notably, about 22% of banks reported their CECL methodologies are not yet finalized.
Banks perform CECL reserve calculations and analyses to understand and manage their credit exposures. To effectively do so, many use credit loss forecasting models. Notably, all participating banks use a mix of quantitative and qualitative methods to calculate their reserves.
Now that the transition to CECL is evident for community banks, many risk managers face the top model risk challenge: maintaining and validating working models while enhancing their banks’ credit measurement methodologies across overall risk management frameworks.
Full survey results are available to RMA members who participated in the survey. Contact us here to learn more about the RMA Model Validation Consortium and the ways we help community banks. Keep a look out for next month’s post, where we explore the effects and the challenges of the CECL transition on community banks’ reserving processes.