A simpler CECL

By Josh Stein and Mike Gullette
ABA Viewpoint

Josh Stein is VP for accounting policy, and Mike Gullette is SVP for accounting policy, at ABA.

Perhaps the biggest operational challenge of implementing the current expected credit loss standard is not the concept of lifetime loss estimation, but the expectation of precision in an environment where precision is not required by the accounting standard — if it’s even possible.

That challenge is most acute for community banks, which typically rely on smaller datasets and non-complex CECL methods. In this context, demanding point‑estimate precision often increases cost and complexity without improving insight into credit risk.

As the Financial Standards Accounting Board undertakes its post‑implementation review of CECL, this tension between flexibility in principle and rigidity in application — especially as enforced by auditing firms — is particularly important to confront.

Rethinking what ‘the CECL model’ is

FASB has acknowledged that estimating credit losses is inherently judgmental. Yet in practice, many community banks are expected to quantify and defend every adjustment layered onto model output. This becomes particularly problematic when community bankers report that qualitative overlays frequently meet or exceed the initial modeled allowance, in some cases by a wide margin.

This experience suggests the need to rethink what “the CECL model” is for non-complex institutions.

For many community banks, the CECL model is not a forecasting engine. Instead, it is a framework for organizing, interpreting and documenting information about credit risk in a consistent and auditable way — including economic conditions, portfolio characteristics and peer performance.

Community bankers consistently emphasize that:

• Changes in economic conditions are primary drivers of credit losses.
• Portfolio mix and local market dynamics matter more than national averages.
• Losses tend to materialize during adverse credit cycles.
• Peer performance helps identify blind spots and outliers.
• Each of these insights can be supported using publicly available data and industry benchmarks.

Public data and peer ranges as primary evidence

When public data signals market deterioration, credit loss expectations naturally increase. Call Report data, government economic releases and resources such as the Atlanta Fed’s Commercial Real Estate Market Index provide a robust foundation for scalable CECL analysis.

In practice, peer credit loss metrics are not merely corroborative. For many community banks, they represent the most relevant observable evidence available, especially when internal loss histories are limited or predominately benign.

Anchoring CECL estimates in public and peer‑based information does not eliminate judgment. Instead, it makes judgment more transparent, more comparable and more focused on changes in credit risk, rather than on compensating for perceived model limitations.

Enhanced SCALE as a practical foundation

The Federal Reserve’s Scaled CECL Allowance for Losses Estimator, or SCALE, methodology already offers a cost‑effective CECL approach to banks with under $1 billion in assets by applying peer‑based allowance coverage rates by loan type. ABA supports SCALE use and further enhancements so more community banks — including many banks above current SCALE thresholds — can rely on external data as the primary anchor for their CECL estimates.

An enhanced SCALE framework would:

• Support reasonable ranges of estimates driven by portfolio characteristics, macroeconomic risk signals and local market conditions.
• Maintain transparency, repeatability and comparability across institutions.
• Reduce reliance on opaque forecasting tools and vendor‑driven complexity.

Expanding the scope of SCALE would better align CECL with risk profiles rather than arbitrary size thresholds.

Reasonable ranges and reduced audit expectations

Rather than forcing banks to defend every basis point of allowance coverage, auditors and examiners should recognize uncertainty — particularly where data is limited and methods are intentionally non-complex.

A “reasonable range” approach would:

• Focus documentation on risks that differ from peer averages and medians.
• Allow reviewers to concentrate on outliers and meaningful changes in risk.
• Reduce audit and validation effort when allowances fall within reasonable peer norms.

This approach would shift external review away from immaterial variance and toward identifying emerging credit risk — precisely what CECL is intended to capture.

A clear opportunity for the CECL implementation review

Simplifying CECL for community banks should be a shared priority among bankers, auditors, regulators, and standard setters. As FASB conducts its post‑implementation review, greater recognition of third‑party data, wider use of peer ranges, expanded reliance on enhanced SCALE and reduced documentation expectations offer a practical and achievable starting point.

Together, these changes would move CECL closer to its original promise: credit loss estimates that are decision‑useful, proportionate and grounded in how community banks actually manage credit risk.