The fastest first pass is not a narrative review of every filing. It is a repeatable screen built from quarterly Call Report data.[1] Pull the same inputs for every bank, calculate the same ratios, rank each bank against a relevant peer set, and read the combinations instead of any single number.
The workflow in one view
- Define the peer set. Group banks by asset size, geography, business model, and loan mix before ranking anything.
- Pull the inputs. Use total loans, average loans, 90+ days past due loans, nonaccrual loans, gross charge-offs, recoveries, allowance for credit losses, tangible equity, intangibles, and other real estate owned when relevant. Core Call Report schedules include RC-C, RC-K, RC-N, RI-B, RI-C, RC, RC-R, and RC-M.[1][2]
- Calculate the ratios. Start with noncurrent loans to loans, annualized net charge-offs to average loans, allowance for credit losses to loans, allowance for credit losses to noncurrent loans, and the Texas Ratio as an optional capital-and-problem-asset stress check.
- Compare three ways. Rank against peers, compare against the bank’s own eight-quarter range, and check the latest quarter-over-quarter direction.
- Sort by pattern. Give priority to banks where problem loans are rising, losses are beginning to appear, and reserve coverage is flat or thin.
- Output a review list. The result should be a short queue of banks to research, not a pass-fail label.
The main insight is sequencing. Noncurrent loans often tell you where stress is visible now. Charge-offs tell you what has already become loss. Reserves tell you how much cushion management has recorded against expected credit losses. The screen is useful because those three signals rarely move at the same time.
Definitions that keep the ratios readable
- Noncurrent loans: loans that are 90 days or more past due plus loans in nonaccrual status, using Schedule RC-N terminology.[2]
- NCO: net charge-offs, or gross charge-offs minus recoveries.
- ACL: allowance for credit losses, the reserve recorded for expected credit losses under current accounting.
- CECL: current expected credit losses, the accounting model introduced by FASB ASU 2016-13 that shifted credit-loss reserving toward expected lifetime losses for many financial assets.[3]
- OREO: other real estate owned, usually property acquired through foreclosure or similar resolution.
- TCE: tangible common equity, commonly calculated as common equity less intangible assets and goodwill.
- ALLL: allowance for loan and lease losses, the older reserve term often used in pre-CECL historical analysis and traditional Texas Ratio definitions.
- SRC: small reporting company, a reporting-status term that can matter when reading accounting adoption timelines, not a credit-quality ratio.
Inputs, formulas, and denominator choices
| Measure | Formula | Why it matters |
|---|---|---|
| Noncurrent loan ratio | 90+ days past due loans plus nonaccrual loans / total loans | Shows loans already exhibiting visible stress. |
| Net charge-off ratio | Gross charge-offs minus recoveries / average loans, annualized for quarterly periods | Shows realized credit losses. Average loans are usually cleaner than period-end loans when balances move sharply. |
| ACL to loans | Allowance for credit losses on loans / total loans | Shows reserve depth against the overall loan book. |
| ACL to noncurrent loans | Allowance for credit losses on loans / noncurrent loans | Shows reserve cushion relative to currently stressed loans. If noncurrent loans are zero, mark the ratio as not applicable instead of creating a fake high coverage score. |
| Texas Ratio | Noncurrent assets plus OREO / TCE plus ACL or ALLL | Adds capital context to problem assets. Use it as a stress lens, not as a standalone failure predictor. |
Methodology rules for a reusable screen
- Source of record: use FFIEC Call Reports for institution-level data and the FDIC Quarterly Banking Profile for industry context.[1][5]
- Update cadence: refresh quarterly after new Call Report data is available. Use the latest quarter for ranking and at least eight quarters for trend.
- Peer-set logic: do not compare a credit-card-heavy bank, a CRE-heavy community bank, and a diversified regional as if their normal loss profiles should match. Start with asset bands, geography, and loan composition.
- Denominator discipline: use average loans for charge-offs, total loans for broad reserve depth, and noncurrent loans for stressed-asset coverage.
- Trend rule: treat one-quarter spikes as prompts for review. Give more weight to two or more quarters moving in the same direction.
Starting thresholds for triage
Thresholds should create a review queue, not a verdict. A practical first pass can flag banks when any of the following appear:
- Noncurrent loans to loans above 1.5%, or materially above the bank’s own recent range.
- ACL to noncurrent loans below 100%, especially if noncurrent balances are rising.
- Annualized net charge-offs accelerating for two consecutive quarters.
- Texas Ratio above 50%, particularly when capital is thin or OREO is increasing.
- Any two of the above occurring at the same time.
Those starting points are intentionally blunt. They should be tightened or relaxed for portfolio mix, collateral type, geography, loan growth, seasoning, and bank size. A lender with a high-yield consumer book can have normal charge-offs that would look alarming at a conservative commercial bank. A CRE lender can look calm until one large credit moves to nonaccrual.
The combinations that matter
| Pattern | Likely read | Next question |
|---|---|---|
| Rising noncurrent loans, low charge-offs, flat reserves | Early stress may be showing before loss recognition. | Which portfolio segment is driving the delinquency or nonaccrual move? |
| Rising noncurrent loans, rising reserves, modest charge-offs | Management may be provisioning ahead of realized losses. | Is reserve growth keeping pace with stressed balances and expected loss severity? |
| High charge-offs, lower noncurrent loans | The bank may have already moved problem credits through loss recognition. | Was this a cleanup quarter or an ongoing loss trend? |
| Low noncurrent loans, low charge-offs, low reserve coverage | This can be fine in a clean, low-risk book, but peer context is essential. | Does the bank carry lower-risk loans, stronger collateral, or just a leaner reserve posture? |
| Elevated Texas Ratio with accelerating NCOs | Problem assets, losses, and capital cushion should be reviewed together. | Is the pressure concentrated, and is tangible capital still adequate for the risk profile? |
The best use of the screen is pattern recognition. A single high ratio can be explainable. A cluster of adverse movements deserves work.
Caveats that prevent bad signals
CECL changed reserve comparisons
Pre-CECL and post-CECL reserve ratios are not perfectly comparable. ASU 2016-13 moved credit-loss accounting toward expected lifetime losses for many financial assets, so a reserve ratio after adoption can mean something different than a similar ratio under the older incurred-loss model.[3] If your trend crosses the adoption period, mark the break instead of treating the series as continuous.
TDR changes affect historical reads
ASU 2022-02 removed the troubled debt restructuring recognition and measurement guidance for creditors and replaced it with disclosures about certain loan modifications to borrowers experiencing financial difficulty.[4] That does not make noncurrent data useless, but it does mean older workout and modification comparisons may need a footnote.
Industry medians are context, not cutoffs
The FDIC reported that the industry reserve coverage ratio was 171.2% in its fourth-quarter 2025 Quarterly Banking Profile.[5] That is useful context, but it is not a universal target. A bank with different loan mix, collateral, geography, or loss history can deserve a different normal range.
What to do after a bank flags
A screen should lead to targeted questions. Before drawing a conclusion, check:
- Which loan category is causing the noncurrent increase.
- Whether the change is broad-based or tied to a few large relationships.
- Whether recoveries are masking gross charge-off pressure.
- Whether reserve growth is coming from real provisioning or from portfolio shrinkage.
- Whether acquisitions, loan sales, or charge-off cleanups distorted the quarter.
- Whether peer banks with similar loan mix are showing the same pattern.
This is where the screen earns its keep. It does not tell you which bank is safe or unsafe. It tells you where the public data is asking for a second look.
Where the workflow gets faster
For a faster implementation, the delinquency and charge-offs view in Banking can help build the same review list without stitching together quarterly filings by hand. Use it after you have chosen the peer set and ratio definitions; the tool should speed up the work, not replace the judgment steps above.
FAQ
Should I rank banks by reserve coverage alone?
No. Reserve coverage is useful only after you know what it is covering. High coverage can mean conservatism, but it can also mean problem assets are already elevated. Low coverage can be acceptable in a clean book or weak if stress is rising.
What if a bank has no noncurrent loans?
Do not force ACL to noncurrent loans into a giant percentage. Mark it as not applicable and rely more on ACL to loans, loan mix, charge-off history, and peer comparison.
How many quarters should the screen use?
Use the latest quarter for the ranking view, but review at least eight quarters before making a judgment. Credit deterioration is often a sequence, not a single datapoint.
Are the same thresholds valid for every bank?
No. The starting thresholds are triage rules. Adjust them for lending category, geography, bank size, growth rate, collateral type, and historical loss behavior.
Why annualize quarterly charge-offs?
Annualizing makes quarterly loss rates easier to compare across banks and periods, but it can exaggerate one-time cleanup quarters. Always pair the annualized figure with the underlying dollar change and recent trend.
Sources
- [1] FFIEC, FFIEC 051 Current Information, current Call Report forms and instructions: https://www.ffiec.gov/resources/reporting-forms/ffiec051
- [2] FDIC, FFIEC 051 Reports of Condition and Income Instructions, including schedules RI-B, RI-C, RC-C, RC-K, RC-N, RC-R, and related materials: https://www.fdic.gov/bank-financial-reports/ffiec-reports-condition-and-income-instructions-ffiec-051-report-form-1
- [3] FASB, Accounting Standards Update 2016-13, Financial Instruments – Credit Losses: https://storage.fasb.org/ASU%202016-13.pdf
- [4] FASB, Accounting Standards Update 2022-02, Financial Instruments – Credit Losses, Troubled Debt Restructurings and Vintage Disclosures: https://storage.fasb.org/ASU%202022-02.pdf
- [5] FDIC, Quarterly Banking Profile, Fourth Quarter 2025, industry asset-quality and reserve coverage context: https://www.fdic.gov/news/speeches/2026/fdic-quarterly-banking-profile-fourth-quarter-2025