The Provision Lag — Q4 2025

FINASENSE Research · March 19, 2026

Data: Q4 2025 Fed Funds: 3.64% · 10Y: 4.18% · 2Y: 3.47%

The lag defers the credit cost — it does not erase it: The 60+ day delinquency ratio crossed 1% in Q4, to 1.03%, the highest in the dataset. Charge-offs did not follow: the net charge-off rate held near 0.77%, and full-year provision expense was essentially flat year-over-year. The gap between rising delinquency and flat losses has widened all year. The common worry — that under-provisioning is inflating earnings — is not what the data shows: provisioning is keeping pace with *realized* losses, and ROAA rose in 2025. The real issue is timing. Delinquent loans are aging without converting, which defers the eventual credit cost into 2026 rather than eliminating it.

The Widening Gap

Through 2025 the two lines came apart. System delinquency climbed from 0.80% in Q1 to 1.03% in Q4, while the annualized net charge-off rate stayed roughly flat — 0.83%, 0.79%, 0.76%, 0.77%. Early in the year charge-offs actually ran above delinquency; by year-end delinquency sat about 26 basis points higher. Past-due balances are accumulating in late buckets rather than resolving into losses.

Source: NCUA 5300 Call Report; FINASENSE analysis.

Provisioning tracked the realized losses, not the delinquency build. Full-year 2025 provision expense was about $14.4 billion, essentially flat (up roughly 1%) against 2024, and the allowance grew only modestly to $23.2 billion. That is a coherent response to charge-offs that are not rising — and a lagging response to a delinquency pipeline that is.

Not an Earnings Illusion

It is tempting to read flat provisioning against rising delinquency as an earnings illusion — net income flattered by reserves that should be higher. The data does not support that reading for 2025. Provision expense as a share of net interest income actually fell over the year, to roughly 17.9% from about 18.7%, because net interest income grew faster than provisioning. ROAA rose, not fell, on a year-over-year basis. Measured against the losses actually being taken, reserves are adequate and earnings are real.

The honest framing is forward-looking, not retrospective. 2025's earnings are not overstated relative to today's losses; they are vulnerable to tomorrow's. The provision lag is the gap between a delinquency rate that has already moved and the charge-offs and provisions that have not yet caught up.

Where the Conversion Has Already Started

The system aggregate hides an uneven leading edge. The over-$10B cohort is already converting: it carries both the highest delinquency (1.49%) and, unlike the rest of the system, the highest charge-offs (1.31%). The aggregate net charge-off rate stays low only because the mid-size and small tiers — charging off 0.47% to 0.60% — have not yet followed their delinquency higher. In other words, the largest, most consumer-concentrated credit unions are a quarter or two ahead in the conversion the rest of the system still has in front of it.

Delinquency vs. Charge-offs by Asset-Size Cohort, Q4 2025 (annualized)

No Results

The Deferred Cost

The mechanism is simple once the timing is clear. As the aged delinquencies across the mid-size and small tiers convert — the way the largest cohort's already have — net charge-offs rise, provisioning rises to match, and ROAA gives back part of 2025's gain. The lag does not avoid that cost; it moves it into 2026. That is the case for not extrapolating 2025's strong, margin-driven earnings: the credit bill is in the mail, just not yet delivered.

Connecting the Lines

This breakdown sits behind two of the Q4 2025 System Signals: the breach of the 1% delinquency mark (Signal 1) and the earnings discussion (Signal 3). The provision lag is the bridge between them — the reason a quarter can post a record margin and rising full-year earnings while the credit pipeline that will eventually weigh on both is still building. Watch the delinquency-to-charge-off gap: it closes when conversion arrives, and that is the quarter earnings turn.

This report is provided for informational and educational purposes only and does not constitute investment, legal, regulatory, or examination advice, nor should it be relied upon as the basis for any decision.

FINASENSE is not affiliated with the National Credit Union Administration (NCUA). Financial data is sourced from NCUA 5300 Call Report filings as submitted by individual credit unions and is not guaranteed as to accuracy or completeness. Ratio definitions and account classifications reference the NCUA Financial Performance Report (FPR) Chart of Accounts. All aggregation, analysis, and derived metrics are independently computed by FINASENSE and may differ from NCUA-published figures. Interpretations reflect the views of FINASENSE and not those of the NCUA.

This report does not consider the specific circumstances of any individual credit union and is not tailored advice. FINASENSE has no financial relationship with, and receives no compensation from, any institution referenced.

All information is provided "as is," without warranty of any kind, and FINASENSE disclaims liability for any decisions made in reliance on this report. Historical metrics are not indicative of future financial condition. This report is proprietary to FINASENSE, a publication of IP Foundries, LLC (Arizona), and may not be reproduced, distributed, or reused without prior written consent.