With the current expected credit loss (CECL) effective date of 1/1/23 quickly approaching for many financial institutions (FIs), the question of whether or not a model validation is needed is on the forefront. CECL replaces the incurred loss model for estimating credit losses and the allowance for loan and lease losses (ALLL). This transition is expected to have a significant impact on the approach management utilizes to calculate this new allowance for credit losses (ACL). The complexity of this calculation has led many institutions to utilize third-party models. Given that CECL is a major change to calculating what is likely the most significant estimate in an FI’s financial statements, management and those charged with governance should be scrutinizing their chosen methodology prior to adoption.
What is Model Validation?
All significant models used by an institution should be subject to validation. Model validation is a set of processes and activities intended to verify that a model is performing as expected and is in line with its design objectives and business use. The frequency of validation depends on the complexity of the model and how often it’s used. The rigor of the validation is determined based on the potential risk.
Any validation should be performed by someone who does not have oversight of the model and is independent of the process. The validation can be done by someone internal to the institution or an external third party. Individuals who perform the validation should possess the skills, knowledge, and experience to accurately validate the model and ensure that the necessary changes are implemented.
Model Validation Expectations
The Office of the Comptroller of the Currency (OCC), Federal Reserve, and Federal Deposit Insurance Corporation (FDIC) developed the Supervisory Guidance on Model Risk Management which outlines sound practices for model risk management programs, including expectations for model validation. The guidance highlights that model validation is important to help prevent adverse consequences (including financial loss) of decisions based on models that are incorrect or misused. Although this guidance is primarily targeted to FIs with greater than $1 billion in total assets, it outlines best practices that should be considered by smaller banks and credit unions as well. The extent of validation should correspond to the FI’s complexity, the materiality of its models, and the size of its operations.
Most institutions now require their Bank Secrecy Act (BSA) and Anti-Money Laundering (AML) systems to undergo model validation. As the ACL is one of the most significant risks on FI’s financial statements, institution’s model risk management programs should consider leveraging outsourced solutions. For upcoming examinations, we expect that most regulators will ask about model validation of the CECL calculation. Perhaps a simple rule of thumb to follow would be your understanding of the calculation – are there complex theories and/or calculations that you don’t understand? If so, you should probably evaluate some level of model validation.
Beyond the Regulatory Requirements
Although CECL model validation may be done for regulatory reasons, there are other benefits to having a validation performed. Some benefits include:
Ensuring CECL models are mathematically accurate and are performing as expected.
Identifying any potential limitations or weaknesses in the CECL model.
Helping management and those charged with governance understand the CECL model, how it works, and how to interpret the outputs.
Gaining insights into model best practices and CECL modeling strategies used by other FIs of similar size and complexity.
Understanding the results of complex CECL calculations presents challenges to most institutions. Model validation will be an essential tool to ensure best practices are followed, a sound CECL model is implemented, and management is making informed decisions.
Learn More
