Skip to main content

Looking for Valuant? You are in the right place!

Valuant is now Abrigo, giving you a single source to Manage Risk and Drive Growth

Make yourself at home – we hope you enjoy your new web experience.

Looking for DiCOM? You are in the right place!

DiCOM Software is now part of Abrigo, giving you a single source to Manage Risk and Drive Growth. Make yourself at home – we hope you enjoy your new web experience.

Portfolio Risk & CECL CECL Accounting CECL Models CECL Regulation

Effective CECL model validation: A framework

Abrigo
August 7, 2023
Read Time: 0 min

Applying model risk management to CECL

What's involved in CECL model validation? Learn what banks, credit unions, and others subject to CECL accounting can expect from this risk management process.

You might also like this webinar, "Conquering CECL model validation: Prepare for success."

LISTEN

Takeaway 1

An area of risk financial institutions can’t afford to overlook is risk posed by models used for developing estimates and making decisions, including CECL models.

Takeaway 2

Regulators say management should periodically validate the loss estimation process for the allowance for credit losses (ACL) and any changes to it.

Takeaway 3

A proper model validation will have a comprehensive scope, clear findings and mitigation actions, and open communication between parties.

Internal risk

Model risks: losses, poor decisions

Many banks are hyper-focused on risks they face related to external forces, such as rising interest rates. But an internal area of risk they can’t afford to overlook is risk posed by various models used for developing estimates and making decisions.

While models can be valuable management tools, the models themselves can pose risk if their output incorrectly informs financial institution decisions. Regulators have noted such risks can involve financial losses, poor business and strategic decision-making, or damage to a bank’s reputation.

During a recent Abrigo webinar, about two-thirds of participants said their financial institutions had a model risk management process in place, as well as an inventory of models. About 30% did not have these but were planning on them in the next two years, and the remaining attendees did not and weren’t planning to implement these.

Model validation is a crucial aspect of model risk management. The OCC handbook describes a model as “a quantitative method, system, or approach that applies statistical, economic, financial, or mathematical theories, techniques, and assumptions to process input data into quantitative estimates.”

Management & directors

Oversight for CECL validation

Models for the current expected credit loss (CECL), too, should undergo model validation given the material estimate they produce. Regulatory guidance says management should periodically validate the loss estimation process for the allowance for credit losses (ACL) and any changes to it to confirm its appropriateness for the institution’s size, complexity, and risk profile.

As the Interagency Policy Statement on Allowances for Credit Losses in April 2023 states:

In reviewing the appropriateness of an institution's ACLs, examiners may….[v]erify that models used in the loss estimation process, if any, are subject to initial and ongoing validation activities. Validation activities include evaluating and concluding on the conceptual soundness of the model, including developmental evidence, performing ongoing monitoring activities, including process verification and benchmarking, and analyzing model output.

The statement adds that the board of directors’ oversight activities, subject to review by examiners, include “requiring management to periodically validate, and, when appropriate, revise loss estimation methods.”

Access a brief guide to CECL model validation here. 

What validations review

The CECL model validation process

What’s involved in CECL model validation, and what can banks expect from this risk management process?

Model risk management guidance (FRB SR 11-7, OCC Bulletin 2011-12, FDIC FIL-22-2017) outlines that the guiding principle for validation is an effective challenge to the model design, implementation, and use. A proper model validation will have a comprehensive scope, clear findings and mitigation actions, and open communication between parties.

In the case of CECL modeling, having an effective challenge to the model design, implementation, and use means the validation will review the following:

  • Model conceptual soundness

  • Data integrity and reconciliation

  • Governance and internal controls (including model documentation)

  • Assumptions and limitations (for example, prepayment speeds and recovery period length)

  • Outcome analysis (e.g., reviewing data development related to the correlation between losses/defaults and various economic variables)

  • Sensitivity and scenario analysis

  • Qualitative adjustments

  • Ongoing monitoring

Specifically related to a CECL model, a model validator is likely to:

  • Evaluate the CECL methodology or CECL methodologies chosen: PD/LGD, vintage, lifetime loss rate, etc.
  • Evaluate the calculation and observance of PD and LGD if using the PD/LGD methodology.
  • Assess the economic indicators chosen for the forecasts and ensure they align with industry practices and are correctly implemented.
  • Evaluate the sensitivity analysis performed and identify studies missing that might have a material impact in the model.
  • Evaluate the methodologies for calculating prepayments, curtailments, and funding rates.
  • Evaluate methodologies and the process for calculating reserves for individually assessed loans.
  • Evaluate the qualitative methodology, including factors considered, thresholds, and scaling.
  • Evaluate whether current assumptions and limitations are relevant, given current economic conditions, business strategy, portfolio, or different model methodologies.
  • Review CECL model documentation for compliance with guidelines provided by model risk management.
  • Inspect the institution’s loan policy to ensure key components of CECL methodology are considered and reviewed, and that it aligns with implementation.

Model validation testing should be performed by a team with no involvement in the development or day-to-day use of the model to avoid conflicts of interest. Independent model validation “helps ensure that incentives are aligned with the goals of model validation,” according to FRB SR 11-7.

CECL models can involve significant management judgment. In addition, the relative recency of CECL adoption means in-house model risk management teams might not have extensive CECL experience. Involving an experienced third party in the organization’s CECL model validation is well worth the investment if it provides assurance of a well-designed and reliable model.

Bankers trust Abrigo advisors to help them with all kinds of CECL-related issues, including model development and purchase accounting.

LEARN MORE
Webinar

Crafting an effective Q factor framework for CECL

Read More
Webinar

Navigating uncertain times: Strategies for effective risk management and compliance

Read More
Webinar

Understanding audit and regulatory expectations for CECL

Read More
About the Author

Abrigo

Abrigo enables U.S. financial institutions to support their communities through technology that fights financial crime, grows loans and deposits, and optimizes risk. Abrigo’s platform centralizes the institution’s data, creates a digital user experience, ensures compliance, and delivers efficiency for scale and profitable growth. Make Big Things Happen.

Full Bio

About Abrigo

Abrigo enables U.S. financial institutions to support their communities through technology that fights financial crime, grows loans and deposits, and optimizes risk. Abrigo's platform centralizes the institution's data, creates a digital user experience, ensures compliance, and delivers efficiency for scale and profitable growth.

Make Big Things Happen.