CECL Model Validation

Confirm your CECL reserve model can withstand the test of time.

Maintaining a reasonable and supportable reserve estimation through varying economic conditions is critical for every financial institution. As these models become increasingly complex, highly scrutinized oversight from management, risk management, auditors, regulators, and investors is the norm for institutions of all sizes. Whether you have implemented a straightforward, in-house model or a more sophisticated third-party model, DCG’s team of CECL model validation and credit risk management experts have the right combination of mathematical rigor, credit modeling expertise, and horizontal perspective to meaningfully validate and effectively challenge your CECL model and related governance.

“Managing your CECL model assumptions is not a set-it-and-forget-it exercise.”

Mike Guglielmo

Managing Director

Transitioning to CECL

Making the move from an existing Incurred Loss model to a CECL compliant model can be a daunting undertaking. By requiring that institutions consider reasonable and supportable forecasted economic conditions affecting collectability of assets, the new accounting standard opens the door for a bevy of potential econometric methodologies. This can often require new or more data, modeling capabilities, talent, and education. Our team has significant horizontal perspective to help you assess these potential hurdles on the way to CECL compliance.

Expert CECL Model Validation

Our CECL model validation team includes former credit model developers, CECL model implementers, and MRM experts. With our combined breadth and depth of knowledge, we provide comprehensive and insightful model validations tailored for your institution’s size and complexity and appropriate for any CECL approach. DCG has experience with all the major vendor provided solutions, as well as internally developed models of varying complexity.

Reserve Benchmarking

Consistent with MRM best practices and as a complement to our CECL model validations, DCG offers a CECL benchmark analysis to provide additional independent challenge. As regulators and auditors become more familiar with existing methodologies and outcomes, they expect strong internal documentation of performance monitoring and outcomes analysis. DCG’s independent benchmark can significantly augment your internal efforts.

CECL Model Documentation

With over 3,000 pages of documentation written over the past 5 years, DCG has the technical expertise and business acumen to produce polished documentation for all types of models. With a deep understanding of regulatory requirements and the ability to make complex topics easier to understand, we can help elevate the quality of model documentation at your institution.

Lessons Learned from DFAST

Our team has validated over 500 credit-related models, relying on a deep history with DFAST, credit, and ALLL validations to develop a robust CECL validation framework. Through unmatched industry experience and horizontal perspective, DCG has developed a 12-part CECL validation scope based on our proprietary research and experience conducting trial validations on CECL test models.

Talk With Us

How We Help Clients

Examples of Client Results

Created institution-specific CECL documentation for $10B+ bank with both technical detail for modelers’ understanding and practical discussion for non-statisticians’ understanding.
Provided effective analysis and a detailed technical review of complex CECL vendor approach and documentation to give management of $25B+ institution confidence that calculations were working as intended.
Identified shortcomings of vendor macroeconomic forecasting regressions used at <$10B community bank. Helped management understand the practical implications of independent variable transformations on reasonable and supportable CECL forecasts.
Validated complex ACL/CECL solutions comprising multiple modeling components, including (proprietary and vendor) econometric forecasting regressions, migration matrixes, loan-level loss rate modeling, portfolio-level loss rate modeling, aggregator models, and qualitative adjustment frameworks.