Health checks have long been recognized by the medical community as one of the best ways to diagnose and treat potential problems early before they can lead to serious health issues down the road. Annual health checks can also help individuals save money on care and maintain good habits that can lead to a longer, healthier life.
As we approach Q2 2026 reporting, many community financial institutions (CFIs) are treating this Current Expected Credit Losses (CECL) cycle as a health check on their process instead of a referendum on their model. After several years of living with CECL, the pain points that surface most often for CFIs tend to be less about calculation engines and more about governance, documentation and how well different parts of the organization line up behind the allowance story.
In our observations across CFIs nationwide, conversations between examiners, auditors and CFIs are reinforcing this shift. They are less focused on whether the lifetime loss methodology is “best in class.” Instead, they are zeroing in on whether management can show a repeatable process, clear qualitative framework and direct link between assumptions, portfolio performance and broader risk views.
"For most community financial institutions, CECL success has less to do with running the most complex model and more to do with making the allowance process understandable, supportable and repeatable," said Janet Leung, Managing Director of PCBB's Advisory Services.
While the standard itself has not moved much, expectations around judgment, overlays and third‑party oversight absolutely have.
What Recent Rate Moves are Really Testing
The evolving economic landscape has become an informal stress test of CECL discipline for CFIs. In a volatile economic climate with shifting demand patterns, regulators tend to look for management teams that can walk through:
- What changed in the portfolio and macro view.
- How those changes flowed into key assumptions and qualitative factors.
- Why the ACL moved the way it did from quarter to quarter.
When CFIs struggle with CECL, it’s rarely because the model can’t run the math. Instead, the friction usually shows up in softer spots such as:
- Reliance on the system output without a robust review.
- Nontransparent models that are difficult to validate under different economic conditions.
- Qualitative factors that are “always there” regardless of conditions.
- Documentation that reports the number but does not unpack the “why.”
In a more stable environment, these weaknesses can be hidden. In a volatile environment, they stand out.
Rethinking the Role of Third‑Party Models
Most CFIs now rely on some form of vendor or system‑based CECL solution as, it improves consistency, controls and access to analytics that would be hard to build in‑house. The risk comes when the presence of a model is treated as a substitute for management ownership.
A healthier posture for CFIs is: “The model helps us, but it does not speak for us.” This means:
- Knowing in plain language what the model is doing and where it is most and least reliable for your portfolio.
- Being able to reconcile model results to what you’re seeing in credit performance.
- Documenting when you lean into, lean away from or adjust model outputs and why.
When CFIs can’t explain period‑over‑period moves without falling back on “that’s what the system produced,” regulators could interpret this as a governance issue, not a technical one.
Qualitative Factors and Overlays: Fewer, Clearer, Better Supported
Qualitative factors and overlays remain where CECL judgment is exercised — and they continue to draw significant regulatory attention. At this stage in the cycle, most CFIs have a documented Q‑factor framework. However, the question is whether it is still fit for purpose.
Here are a few helpful patterns we see:
- Fewer factors, each clearly tied to specific risks and data points, instead of long lists that overlap.
- Adjustments that move when conditions change, rather than static “plug” amounts that sit in place for multiple reporting periods.
- Narratives that connect the dots from observed trends (internal and external) to a specific directional impact on the allowance.
It may be helpful to treat overlays the way you would any other risk control: Define why they exist, what would cause you to resize them and what would cause you to take them off. This will keep overlays from becoming permanent fixtures and make it easier to explain them to your board and exam teams.
Documentation: Telling the Story Once
Documentation continues to drive a meaningful share of findings, even at CFIs that are otherwise comfortable with their CECL approach. Often, the issue is not a lack of information but a lack of a coherent story. Different groups may describe the same change in different ways, or assumptions may be updated in practice without the rationale being captured in writing.
“What we consistently see is that governance is the real differentiator,” said Leung. “Teams that can explain key assumptions in plain language, connect them to portfolio performance and document a disciplined process that links CECL to how they manage credit and portfolio risk are better positioned to support their allowance decisions in a range of environments.”
A practical target for 2026 is to “tell the story once” across functions:
- Start with a concise narrative of what changed in the risk environment and portfolio.
- Link that narrative to specific model inputs, qualitative factors and overlays.
- Make sure explanations of CECL changes align with how you report ALM, stress testing, and credit.
We have observed that when an regulator can follow this chain without having to reconcile conflicting messages across committees and documents, your results will be judged more deliberate and less mechanical.
Pulling CECL Closer to ALM and Credit Risk
CECL is now firmly part of the broader risk conversation. Where examiner questions are popping up, they often sound like:
- “If these are the forecasts and prepay assumptions you are using in ALM and credit stress, why is the CECL forecast assuming something different?”
- “If you are highlighting rising delinquencies or concentrations to the board, where do we see that show up in your qualitative framework?”
- “If stress testing identifies a material downside case, how does that inform your allowance decisions?”
You do not need perfect one‑to‑one alignment, but you do need a thoughtful explanation when assumptions differ. Tightening that alignment is often more about getting the right people in the room at the right time than about changing the model.
Where to Focus CECL Efforts in 2026 and into 2027
There is still plenty of meaningful work to do for CFIs that are not looking to overhaul their models. Based on what we’re hearing from our CECL customers, priority areas that many institutions are tackling this year and into next year include:
- Refreshing qualitative frameworks to remove overlap, tighten support and make sure they move with conditions.
- Clarifying the governance of management overlays, including triggers, review cadence and exit criteria.
- Documenting how vendor models are configured, where they fit well and where management is applying judgment on top.
- Streamlining CECL documentation so finance, credit, and risk are all telling a consistent story.
- Using board and committee packages to connect CECL outcomes to the broader risk and performance picture.
For CFIs that lean into these governance and documentation upgrades, economic volatility becomes less a source of exam anxiety and more of an opportunity to show that CECL is integrated into how risk is managed — not just how an accounting estimate is made.
