Bank Lending Pullback Signals: A Call Report Screening Method

For bank credit analysts testing peer risk appetite, the practical question is simple: is a bank quietly moving away from risk-sensitive lending before losses become obvious?

Executive summary: A lending pullback is a meaningful decline in a selected loan category, or in that category’s share of gross loans, that is not explained by the whole loan book shrinking at the same pace. The signal is stronger when weaker credit quality, higher provisions or charge-offs, capital pressure, funding pressure, or management commentary points in the same direction. A one-period balance decline by itself does not prove credit stress, especially around mergers, reclassifications, loan sales, or seasonal runoff.

Last reviewed: April 23, 2026. Author/editor credentials: Prepared by the Deep Digital Ventures banking-data editorial team, which builds analysis tools from public Call Report, FDIC, and regulator datasets. Methodology note: This article describes a public-data screen, not an investment recommendation, credit rating, or supervisory finding.

A lending pullback does not need to show up as a headline decline in total loans. A bank can keep gross loans nearly flat while reducing construction loans, non-owner-occupied commercial real estate, C&I exposure, credit cards, auto loans, or other consumer balances. The point is not to label a bank as stressed from one quarter of data. The point is to find the loan categories where management may already be changing risk appetite, then use filings, enforcement databases, and management commentary to decide whether the move is disciplined balance-sheet management or a warning sign.

Which Categories to Watch

The first screen should start with loan categories that connect to either supervisory concentration guidance or credit-cycle sensitivity. The current FFIEC instructions are the source documents to check before mapping fields across banks of different size and complexity.^[2][3]

• Construction and land development loans: compare the balance with total risk-based capital when reviewing CRE concentration. The 2006 interagency CRE guidance uses construction, land development, and other land loans at 100 percent or more of total capital as one supervisory screening criterion.
• Non-owner-occupied commercial real estate: focus on income-property exposure rather than owner-occupied operating-company property. The same guidance uses total CRE loans at 300 percent or more of total capital, together with 50 percent or more CRE growth during the prior 36 months, as another supervisory screening criterion.
• Commercial and industrial loans: read management commentary for whether runoff is borrower demand, pricing discipline, line utilization, or tighter underwriting.
• Credit card loans: compare balance movement with charge-offs, recoveries, and delinquency trends.
• Auto loans: compare balance movement with whether loss content is changing faster than the portfolio balance.
• Other consumer loans: watch the line separately, especially at consumer-focused banks, but keep it outside the example aggregate below unless the products are comparable enough to combine.

Start with the legal bank, not the parent company. Use bank search and individual bank profiles to identify the institution, then confirm it in FDIC BankFind or the National Information Center.^[4][5] A parent-company investor deck may describe strategy, but the Call Report shows the insured depository institution’s balance sheet.

A Simple Pullback Screen

A useful pullback screen has to separate three cases: broad loan contraction, targeted rotation, and data noise. The cleanest version compares the same bank over at least two reporting dates, uses the same Call Report definitions, and tests loan movement against capital, credit quality, funding, and public commentary.

For the public-data example below, the bank universe was U.S. insured depository institutions with Call Reports available for both June 30, 2025 and December 31, 2025. The selected loan group was construction and land development, non-owner-occupied CRE, C&I, credit card, and auto loans. A bank was flagged when that group fell by at least about $500 million and by at least 2.0 percentage points as a share of gross loans over the six-month comparison. Other consumer loans were monitored in the watch list but excluded from this aggregate because that line can mix personal, installment, and specialty consumer products that are not always comparable across banks.

Metric	Primary source	Why it matters
Selected-group dollar change	Loan category balances	Shows whether the bank actually reduced funded exposure in the selected loan group.
Selected-group share of gross loans	Loan category balances and total loans and leases	Controls for banks that are growing or shrinking overall.
Total loan growth	Loan totals and balance-sheet totals	Separates broad contraction from targeted rotation.
CRE concentration ratios	CRE loan balances plus total risk-based capital	Applies the 100 percent construction-and-land and 300 percent CRE screening criteria from the 2006 interagency CRE guidance.
Provision, charge-offs, and noncurrent loans	Income, loss, allowance, and credit-quality schedules	Tests whether the pullback is accompanied by emerging credit stress.
Deposits and funding mix	Deposit and brokered-deposit schedules	Checks whether loan runoff may be tied to liquidity, brokered deposits, or funding cost pressure.
Management and regulatory context	Earnings materials, SEC filings where available, and regulator enforcement databases	Distinguishes planned portfolio remix from reactive tightening.

The arithmetic is simple. Pull two Call Reports for the same legal bank from the FFIEC Central Data Repository.^[1] Sum the selected categories. Divide that sum by gross loans. Then compare the percentage-point change in the loan group with the change in total loans, noncurrent loans, provision expense, net charge-offs, allowances, and capital ratios.

Worked example: under that six-month screen, UMB Bank, National Association reported that the selected loan group fell by about $1.2 billion from June 30, 2025 to December 31, 2025, while its share of gross loans moved from 64.3 percent to 57.9 percent and total loans increased. That combination meets the dollar and share thresholds while avoiding the simplest explanation, a shrinking total loan book. It still needs confirmation in credit quality, losses, allowances, and management commentary.

For CRE-heavy banks, add the supervisory concentration check. The FDIC FIL-104-2006 interagency CRE guidance points analysts to two screens: construction, land development, and other land loans at 100 percent or more of total capital, and total CRE loans at 300 percent or more of total capital when CRE loans have grown 50 percent or more during the prior 36 months.^[6] Those criteria are screens, not automatic limits or conclusions.

Examples From a Public-Data Screen

The table below shows selected large-dollar flags from the June 30, 2025 to December 31, 2025 screen. It is not a ranked list of weak banks and not a statement that any listed bank is under stress. The screen used construction and land development, non-owner-occupied CRE, C&I, credit card, and auto loans as the selected group; it was not the same definition as the interagency CRE concentration screen.

Bank	Selected-group change, June 30 to Dec. 31, 2025	Share of gross loans	Criteria-based read
BMO Bank National Association	Down about $4.3 billion	43.5 percent to 41.2 percent	Met the dollar and share thresholds; total loans also moved lower, so broad contraction remains possible.
Stifel Bank	Down about $2.8 billion	57.8 percent to 20.3 percent	Met the thresholds by a wide margin; review legal-entity, classification, and strategy context before drawing a credit conclusion.
UMB Bank, National Association	Down about $1.2 billion	64.3 percent to 57.9 percent	Met both thresholds while total loans increased; follow-up should test whether the move reflects rotation, pricing, runoff, or classification.
Webster Bank, National Association	Down about $868 million	28.0 percent to 25.1 percent	Met both thresholds alongside overall loan growth; compare credit and loss metrics before assigning a cause.
Raymond James Bank	Down about $536 million	19.2 percent to 16.9 percent	Met both thresholds with total loans moving higher; confirm whether the change reflects strategy, runoff, or classification.

These rows are screening outputs from public Call Report loan data, not investment conclusions and not statements that any listed bank is weak. Mergers, portfolio sales, line reclassifications, specialty-lending runoff, and parent-company reporting differences can distort a simple period-over-period read. The next check is whether the same period shows rising noncurrent loans, higher net charge-offs, allowance changes, or capital pressure.

The peer question matters too. A $500 million decline has a different meaning at a large national bank than at a smaller specialty bank. Use peer comparison to line up the same category, the same reporting dates, and the same denominator before deciding whether a bank is unusual.

How to Avoid False Signals

Loan categories can move for reasons that have little to do with credit fear. A merger can change the legal entity. A portfolio sale can create a one-quarter drop. A bank can reclassify loans after an internal review. A specialty lender can see seasonal runoff. A public-company parent can also discuss loans at a consolidated level that does not match the insured bank’s Call Report.

• Confirm the legal bank first. Use FDIC BankFind or NIC before comparing BMO Bank National Association, Webster Bank, National Association, or any other insured bank with a parent-company filing.^[4][5]
• Use the current Call Report instructions. The December 2025 FFIEC 031, 041, and 051 instructions explain the schedule definitions; do not mix categories from an old mapping without checking the instructions.^[2][3]
• Check at least three reporting periods when the signal is large. A June-to-December move can identify the issue, but March, June, September, and December help separate a trend from quarter-end noise.
• Pair dollar movement with share movement. A dollar decline in construction loans is less informative if total loans fell by the same proportion.
• Test credit confirmation. If selected exposure falls while noncurrent loans, provision expense, and net charge-offs all worsen, the pullback deserves a more skeptical read.
• Check capital and concentration. For CRE, compare loan balances with total risk-based capital and the 100 percent, 300 percent, and 50 percent screens in the 2006 interagency CRE guidance.^[6]
• Review enforcement and supervisory context. Search the FDIC enforcement decisions and orders database, the OCC enforcement actions page, and the Federal Reserve enforcement actions page before treating a balance-sheet move as purely strategic.^[7][8][9]
• Use amended or final filings when available. Filing timing and later amendments can change a period-over-period read, especially when a bank has unusual transactions or reporting changes.

The screen should also be tied back to source data. If a number is going into a board packet, diligence memo, or article, cite the exact reporting period, schedule, and field family. For background on what the site is using, start with the underlying public-data context view, then verify the source filing directly through FFIEC or FDIC.

What the Signal Means

A lending pullback can mean risk discipline, weak borrower demand, tighter pricing, funding pressure, a loan sale, a merger-related cleanup, or early credit concern. The useful decision rule is to classify the signal by confirmation, not by the loan decline alone.

Pattern	Most likely read	Next action
Selected-group share falls, total loans rise, and credit metrics stay stable	Targeted rotation or pricing discipline	Read management commentary and compare peers.
Selected-group share falls, total loans fall, and funding pressure rises	Balance-sheet contraction or liquidity management	Check deposits, brokered deposit data, and liquidity discussion.
Selected-group share falls while noncurrent loans, charge-offs, or provisions rise	Possible credit-cycle response	Review credit-quality schedules, loss schedules, allowances, capital, and recent enforcement records.
One category drops sharply without related credit or funding pressure	Possible portfolio sale, reclassification, or business-line change	Look for transaction disclosure, Call Report footnotes where available, and parent-company filings.

For a credit analyst, the signal is a map for the next document review. Start with the category movement, then move to credit quality, losses, allowances, capital, enforcement databases, and management commentary. If those sources point in the same direction, the pullback is worth writing up. If they conflict, the right conclusion is narrower: the bank changed mix, and the cause is not yet proven.

Secondary Uses

The same screen can support other workflows, but they should stay secondary to the lending-mix question. A fintech founder can use bank profiles to identify the legal sponsor bank and ask whether concentration limits or approval standards are changing. A journalist or board member can use peer comparison and public-data context to decide whether a move is unusual before turning to filings, enforcement records, and management commentary.

Deep Digital Ventures Banking is built around this public-data workflow, but the site output should be treated as a starting point. The final memo still needs source filings, exact dates, and a clear statement of what the screen can and cannot prove.

FAQ

Is a decline in CRE or C&I lending always bad?

No. A decline can be a healthy response to pricing, concentration limits, or weak risk-adjusted returns. It becomes more concerning when the same period also shows higher noncurrent loans, higher net charge-offs, higher provision expense, or weaker capital ratios.

Which Call Report data should I start with?

Start with loan categories and gross loans. Then use past due and nonaccrual loans, charge-offs and recoveries, allowances, provision expense, and regulatory capital to test whether the loan movement is backed by credit or capital confirmation. For average balances, use Schedule RC-K.

Why do CRE screens use capital instead of total assets?

The 2006 interagency CRE concentration guidance uses total risk-based capital as the denominator for its construction-and-land and total-CRE supervisory screens. That is why a CRE pullback screen should connect loan balances to capital, not only to total assets.

How many periods should I compare?

Use at least two reporting dates for the first screen and at least three when the signal is large. A six-month comparison can identify the issue, but a four-quarter view is better for separating a true change in appetite from quarter-end movement, seasonality, or a one-time transaction.

Sources

1. FFIEC Central Data Repository Call Report bulk data: https://cdr.ffiec.gov/public/PWS/DownloadBulkData.aspx
2. FDIC-hosted FFIEC 031 and 041 Call Report instructions, December 2025 edition: https://www.fdic.gov/bank-financial-reports/ffiec-reports-condition-and-income-instructions-ffiec-031-and-041-report-3
3. FDIC-hosted FFIEC 051 Call Report instructions, December 2025 edition: https://www.fdic.gov/bank-financial-reports/ffiec-reports-condition-and-income-instructions-ffiec-051-report-form-1
4. FDIC BankFind legal-entity lookup: https://banks.data.fdic.gov/bankfind-suite/bankfind
5. FFIEC National Information Center institution lookup: https://www.ffiec.gov/npw
6. FDIC FIL-104-2006 interagency CRE concentration guidance: https://www.fdic.gov/index.php/news/financial-institution-letters/2006/fil06104.html
7. FDIC enforcement decisions and orders database: https://orders.fdic.gov/s/
8. OCC enforcement actions page: https://www.occ.gov/topics/laws-and-regulations/enforcement-actions/index-enforcement-actions.html
9. Federal Reserve enforcement actions page: https://www.federalreserve.gov/supervisionreg/enforcementactions.htm

Continue reading→