ALCO, CECL, and Benchmarking: Closing Balance Sheet Gaps

Are You Saving Margin or Creating Margin?

Three Conversations That Should Be One

The ALCO meeting is running on rate sheets, maturity schedules, beta and liquidity sensitivity reports, and macroeconomic forecasts. Down the hall, loan review is working through a credit picture--rising watch-list credits, a CRE concentration approaching regulatory attention, and a fog of information that is new to many who have not experienced this cycle before. Those two rooms are assessing the same institution, while using different data. They are not talking.

At most community banks, balance sheet governance runs as three separate conversations: funding, reserves, and peer benchmarking. In time, each function developed its own workflow, data cadence, and reporting rhythm. That evolution was reasonable when each function had a different regulator, a different calendar, and a different team accountable for the output. The problem is that these three functions are analytically dependent on each other in ways the separation obscures. The institutions that figure that out before the exam cycle forces the conversation are the ones that come through without lasting franchise damage.

Too often, funding decisions are made without knowing where you stand

Most institutions know their own cost of funds. While publicly available data like call reports provide insightful information, they are not strategic. Cumbersome, vague and generalized, this data has a meaning lag time to the real-time decisions being made. Furthermore, it is challenging to map to the medians for peer institutions of comparable size, geography, and business mix.

That gap matters because rate matching is not the same as optimal pricing. For example, an institution consistently 15 basis points above the peer median on CD rates may believe it is being competitive. Yet it may be overpaying for deposits it could have retained at lower cost. I've seen both scenarios play out in the same market, at institutions operating on the same street corners, and the difference almost never traces to competitive intensity. It traces to whether anyone had looked.

Call this the funding-context gap: the distance between what an institution knows about its own cost structure and what it would know if it could see that cost structure against a true peer cohort. It is the most common source of preventable margin drag in community banking. This is rarely identified as the root cause because the institution does not know what it does not know.

Rate posting is not purely a rate decision. It carries implicit assumptions about where the institution stands relative to peers, what the credit outlook looks like over the life of that funding, and whether the cost of that deposit is consistent with where NIM needs to land. Every rate posted, every maturity structured, reflects a view of the balance sheet — whether that view is made explicit or not.

When those assumptions are left implicit, the funding strategy defaults to the simplest available signal: what competitors are posting. That is a reasonable starting point. It is a poor ending point.

What the credit picture adds to the CECL calculation

Community institutions with clean historical loss records face a well-documented challenge in CECL: the problem of zeros. When internal loss history is negligible, that history is not predictive of the next credit cycle. The technical answer is in peer-enhanced data — using stressed peer loss rates at the 75th or 90th percentile to justify qualitative factor adjustments in benign cycles.

That peer data is only useful if it is current, comparable, and accessible at the time the reserve calculation is being made. Examiners have shifted their focus from mathematical precision to conceptual soundness. They want to see that the assumptions are coherent and well-documented. The math alone does not satisfy that anymore.

The CECL reserve calculation is also where the credit context from ALCO and benchmarking converges. An institution running CECL off its own loss history, isolated from peer performance data and from the credit migration trends inside its own portfolio, is making reserve decisions with two variables missing. I have reviewed CECL models at community institutions where the qualitative factor adjustments were reasonable, but the peer anchoring was two quarters stale. The math was fine. The story it told was empirically accurate, but woefully incomplete.

Why the ALCO team needs the credit picture

Community bank NIM reached a six-year high of 3.75% in Q1 2026*. That number looks strong in isolation. It looks different when you factor in approximately a trillion dollars of CRE repricing still in the pipeline, declining consumer credit scores across the industry, and what I am seeing described as the first meaningful rise in non-performing loans in over a decade.

Community banks are repricing deposits faster than they are repricing loans. That creates NIM pressure that is partly rate-driven, partly mix-driven, and partly a function of credit performance inside the book. Most CFOs can decompose the rate effect. Fewer can isolate which portion is structural deterioration in the credit book versus a cyclical funding cost adjustment. That decomposition requires credit context the ALCO conversation rarely has.

The ALCO team discussing deposit strategy needs to be working in concert with the credit risk picture from loan review. When those two conversations do not share data, the institution is running two separate models of its own risk. Each driver of NIM movement — rate effects, volume effects, mix effects, credit performance effects — calls for a different response. The ALCO team missing the credit variable is, by definition, solving the wrong version of the problem.

What connecting these conversations actually looks like

The institutions I have seen do this well are not necessarily larger or better resourced. They run these three conversations from shared data. Peer benchmarking data sits inside ALCO discussions — not as a quarterly report reviewed after the fact, but as a live reference when rate decisions are being made. CECL reserve calculations are anchored to current peer performance data, not the potentially sugar-coated internal loss history alone. The credit migration picture — where the portfolio is moving, what stress scenarios show, what examiners are likely to focus on — is part of the context when funding decisions happen.

The discipline here is operational, not technological. Institutions doing this well have designated someone accountable for making sure the CECL team and the ALCO team are looking at the same peer data set in the same reporting cycle. They have removed the quarter-lag between when credit signals emerge and when they reach the funding conversation.

That is the specific change. Not a new system. Not a restructured committee. Rather, it is a shorter distance between three data sets (Funding, Reserves, Benchmarking) that already exist inside the institution.

There is a useful distinction between the margin you save and the margin you create. Saving margin is a defensive exercise: hold rates, watch peers, don't overpay. Creating margin means understanding your funding-context gap, connecting your credit outlook to your ALCO assumptions, and making rate decisions with the full balance sheet in view. Most community institutions are adept at the first. The institutions I have seen pull away from peer median on NIM do the second. The gap between them opens gradually, over two to three rate cycles, and by the time it is visible in the income statement it is two years old in the decision record.

The institutions that close it do not wait for the next exam to force the conversation. They have it now, with the data they already have.

*FDIC Quarterly Banking Profile - Q1 2026