TREMOR publishes a quarterly dataset of US bank lending to nondepository financial institutions (NDFIs) - loans from FDIC-insured banks to mortgage intermediaries, business-credit lenders, private equity funds, consumer-credit lenders and other nonbank financial firms - extracted from FFIEC Call Report bulk downloads. This page documents exactly how that data is validated.
All NDFI data is independently validated against the FDIC BankFind API and the FDIC Quarterly Banking Profile (QBP), plus internal-consistency checks. Validations were run across all 6 quarters (Q4 2024 through Q1 2026) and every bank reporting NDFI loans - roughly 3,683 bank-quarter records. Q1 2026 was added in May 2026; the older quarters were force-refreshed at the same time to capture amended Call Report filings (deltas were minimal: +3 banks and +$2.4B NDFI across Q1-Q4 2025 combined).
For data sources, FFIEC field definitions and category methodology, see the companion NDFI methodology page.
Open the interactive US NDFI Banks tool on TREMOR for the live charts, per-bank explorer, and the Q4 2025 / Q1 2026 deep-validation toggle.
Part 1 - Bank-level validations (8 checks)
Validation 1 - FDIC cross-check: Total Assets (all banks)3,673 of 3,683 exact. Every bank found in FDIC; total assets match to the last thousand dollars except 10 minor amendment-timing diffs.
What we checked: for every bank in our dataset (594 to 630 per quarter), we compared the total assets figure extracted from the Call Report against the same bank's total assets in the FDIC BankFind API. Both report in thousands of dollars.
Result: across all 6 quarters, every single bank was found in FDIC (zero missing). Out of 3,683 bank-quarter records, 3,673 matched exactly. The 10 exceptions (all from the original 5-quarter run) are minor timing differences from late Call Report amendments. Q1 2026 added 622 records with zero exceptions.
| Quarter | Banks | Our Assets | FDIC Assets | Exact Matches |
|---|---|---|---|---|
| Q4 2024 | 594 | $20,746.3B | $20,746.3B | 594 / 594 |
| Q1 2025 | 603 | $21,231.2B | $21,231.2B | 603 / 603 |
| Q2 2025 | 608 | $21,820.8B | $21,820.8B | 607 / 608 |
| Q3 2025 | 624 | $22,033.3B | $22,033.3B | 624 / 624 |
| Q4 2025 | 630 | $22,211.4B | $22,211.4B | 621 / 630 |
| Q1 2026 | 622 | $23,082.1B | $23,082.1B | 622 / 622 |
The 10 asset exceptions (largest: Regions Bank Q4 2025, -$806M on $157.4B; Optum Bank -$5M; the rest under $1.2M) are caused by banks filing amended Call Reports after the initial submission - the FDIC reflects the amendment while our bulk download captured the original filing.
Validation 2 - FDIC cross-check: NDFI totals (55 largest banks)333 of 334 exact (89.8% of NDFI dollars). Our extracted NDFI total (RCFDJ454) matches FDIC's published LNNDEPC field for every bank FDIC reports it.
What we checked: the FDIC BankFind API publishes LNNDEPC (total loans to nondepository financial institutions) for ~55 banks per quarter (the largest). We compared our extracted NDFI total against FDIC's LNNDEPC for every bank where FDIC reports it, across all 6 quarters.
Result: 333 of 334 bank-quarter comparisons matched exactly to the last thousand dollars. These banks account for ~89.8% of total NDFI dollar value; the remaining ~10.2% comes from ~570 smaller banks where FDIC does not publish LNNDEPC. The single exception: Huntington National Bank Q4 2025, a $392K difference on a $15.3B position (0.003%), almost certainly an amended filing.
Example: JPMorgan Chase (RSSD 852218), Q1 2026 - our extraction $238,443,000K, FDIC BankFind API $238,443,000K. Exact match.
| Quarter | FDIC Banks | FDIC NDFI | Exact Matches | Coverage |
|---|---|---|---|---|
| Q4 2024 | 55 | $1,042.7B | 55 / 55 | 90.0% |
| Q1 2025 | 56 | $1,145.7B | 56 / 56 | 90.1% |
| Q2 2025 | 56 | $1,240.8B | 56 / 56 | 89.7% |
| Q3 2025 | 56 | $1,313.3B | 56 / 56 | 89.8% |
| Q4 2025 | 55 | $1,405.2B | 54 / 55 | 89.6% |
| Q1 2026 | 56 | $1,484.3B | 56 / 56 | 89.9% |
Coverage = share of our total NDFI (all banks) validated against FDIC. The unvalidated ~10% comes from ~570 smaller banks whose NDFI totals are individually small (average $0.3B each) and whose total assets are already validated exactly in Validation 1.
Validation 3 - Subcategory-to-total reconciliation654 of 776 exact; 122 rounding diffs (max $2K). For $10B+ banks, the five NDFI subcategories sum to the reported NDFI total within bank-side rounding.
What we checked: banks with $10B+ in assets report five NDFI subcategories (mortgage credit, business credit, private equity funds, consumer credit, other). The sum should equal the bank's reported NDFI total. We checked every bank reporting subcategories across all 6 quarters, reading directly from the raw FFIEC filing data.
Result: out of 776 bank-quarter records with subcategories, 654 matched exactly. All 122 exceptions are rounding differences of $1K or $2K - the bank rounds each subcategory individually, so the sum doesn't always match the independently-rounded total. No exception exceeded $2K. Example: Citizens Bank (Q4 2025) NDFI total $18,674,999K vs subcategories sum $18,674,997K - a $2K rounding artifact in the bank's own filing.
| Quarter | Banks w/ Subcategories | Exact Match | Rounding Diff ($1-$2K) |
|---|---|---|---|
| Q4 2024 | 126 | 115 | 11 |
| Q1 2025 | 128 | 108 | 20 |
| Q2 2025 | 128 | 104 | 24 |
| Q3 2025 | 130 | 104 | 26 |
| Q4 2025 | 131 | 109 | 22 |
| Q1 2026 | 133 | 114 | 19 |
Validation 4 - No negative valuesPASS. Zero negatives across 3,683 records x 7 fields.
All numeric fields (total assets, NDFI total, and all five subcategories) were scanned for negative values across all 3,683 bank-quarter records. Zero negatives found, as expected for loan balance data.
Validation 5 - No duplicate bank entriesPASS. Every RSSD ID appears exactly once per quarter.
Each bank is identified by its RSSD ID (Federal Reserve unique identifier). We verified no bank appears more than once in any given quarter. Zero duplicates found across 3,683 records.
Validation 6 - NDFI does not exceed total assetsPASS. Zero violations; highest NDFI/assets ratio 48.8% (warehouse lender).
A bank's NDFI loans should always be smaller than its total assets. Verified for all 3,683 records: zero violations. The highest NDFI-to-assets ratio observed is 48.8% (Northpointe Bank, Q4 2025), high but plausible for a bank specializing in warehouse lending. Q1 2026 highest: 34.8% (EverBank).
Validation 7 - Quarter-over-quarter asset stability5 flagged, all M&A. Five bank-quarter pairs exceeded a 50% QoQ asset swing - all explained by mergers/acquisitions, not data errors.
What we checked: large QoQ swings in a bank's total assets (more than 50%) could indicate a data error. We flagged all such cases.
Result: 5 bank-quarter pairs exceeded the 50% threshold over Q4 2024 - Q4 2025, all explained by M&A. Q1 2026 added no new banks crossing the threshold.
| Bank | Period | Before | After | Change |
|---|---|---|---|---|
| Foresight Bank | Q1 to Q2 2025 | $0.3B | $1.6B | +377% |
| Maine Community Bank | Q4 2024 to Q1 2025 | $1.2B | $2.7B | +131% |
| ChoiceOne Bank | Q4 2024 to Q1 2025 | $2.7B | $4.3B | +58% |
| Busey Bank | Q1 to Q2 2025 | $12.0B | $18.9B | +58% |
| Millennium Bank | Q3 to Q4 2025 | $0.5B | $0.8B | +54% |
Validation 8 - Bank continuity across quartersNormal churn. 23-37 banks added and 18-28 dropped per quarter - banks enter when they begin NDFI lending and exit when those loans run off.
We tracked which banks appear and disappear across consecutive quarters. Normal churn observed. The count grows from 594 (Q4 2024) to 630 (Q4 2025), then dips to 622 (Q1 2026) - natural churn (loans paid off or consolidated through M&A) plus possible late-filer drift. System-wide NDFI dollars still grew +5.1% QoQ to $1,650B.
Validation summary
| # | Check | Scope | Result |
|---|---|---|---|
| 1 | Total assets vs FDIC | 3,683 bank-quarters | 3,673 exact, 10 minor diffs |
| 2 | NDFI total vs FDIC | 334 bank-quarters (89.8% of $) | 333 exact, 1 diff ($392K) |
| 3 | Subcategory sum = total | 776 bank-quarters | 654 exact, 122 rounding ($1-2K) |
| 4 | No negative values | 3,683 records x 7 fields | Pass |
| 5 | No duplicate RSSDs | 3,683 records | Pass |
| 6 | NDFI <= total assets | 3,683 records | Pass |
| 7 | Asset stability (<50% QoQ) | ~3,100 quarter-pairs | 5 flagged (all M&A) |
| 8 | Bank continuity | 5 quarter transitions | Normal churn |
Validation sources
- Primary data: FFIEC Central Data Repository - bulk Call Report downloads (tab-delimited, one ZIP per quarter).
- Cross-validation: FDIC BankFind API (banks.data.fdic.gov/api/financials) - total assets (ASSET) and NDFI loans (LNNDEPC).
- Subcategory reconciliation: internal consistency check against raw FFIEC Schedule RC-C Part I filing data.
Validation last run: May 2026 (Q1 2026 added).
Part 2 - Deep cross-validation vs FDIC Quarterly Banking Profile
These checks compare our extracted system-wide totals (all FDIC-insured banks) against the FDIC Quarterly Banking Profile - an independent aggregation from the same underlying Call Report filings. Both Q4 2025 and Q1 2026 are shown below (the interactive tool has these behind a quarter toggle).
Q4 2025
Banking system totals - Q4 2025Close match. Total assets +0.073%, total loans +0.0006%, nonaccrual +0.016% vs FDIC QBP (FRED).
| Metric | Our Value | FDIC QBP (FRED) | Diff % |
|---|---|---|---|
| Number of institutions | 4,394 | 4,336 | +1.3% |
| Total assets ($M) | 25,276,419 | 25,257,913 | +0.073% |
| Total loans ($M) | 13,477,607 | 13,477,526 | +0.0006% |
| Nonaccrual loans ($M) | 90,240 | 90,226 | +0.016% |
| 90+ DPD accruing ($M) | 39,679 | 39,699 | -0.050% |
Institution count: our 4,394 includes all banks in the FFIEC CDR bulk download; FDIC QBP (4,336) is 58 lower (the exact reason - excluded filers, cutoff dates, or methodology - is not independently verified). The total asset difference between the two counts is $18.5B (0.073% of $25.3T). FRED series: QBPBSTAS (assets), QBPBSTASTLN (loans), QBPLNTLNNACSTAT (nonaccrual), QBPLNTLN90DU (90+ DPD). Source: FDIC QBP Q4 2025 PDF, Table IV-A.
Credit quality rates - Q4 2025Near exact. Noncurrent, 30-89 DPD and PDNA rates all within 0.0005pp of FDIC published values.
| Rate | Our Value | FDIC QBP | Difference |
|---|---|---|---|
| Noncurrent rate (NA + 90+ DPD) / loans | 0.96397% | 0.96401% | -0.00005pp |
| 30-89 DPD rate | 0.59652% | 0.59653% | -0.00001pp |
| PDNA rate (30-89 + 90+ + NA) / loans | 1.5605% | 1.5600% | +0.0005pp |
All three rates within 0.0005 percentage points (less than 1/20th of a basis point). Numerators from Schedule RC-N (RCFD1403 nonaccrual, RCFD1407 90+ DPD, RCFD1406 30-89 DPD); denominator from Schedule RC-C Part I (RCFD2122). Source: FDIC QBP Q4 2025 (PDNA 1.56% confirmed).
NDFI nonaccrual-rate methodology - Q4 2025PE funds 0.397% (highest). Subcategory NA rates from FFIEC RC-N memo items over RC-C balances.
| Subcategory | NA Amount | Balance | NA Rate |
|---|---|---|---|
| Mortgage credit (RCFDPV23 / PV05) | $0.7B | $353B | 0.192% |
| Business credit (RCFDPV24 / PV06) | $0.0B | $378B | 0.006% |
| PE funds (RCFDPV25 / PV07) | $1.5B | $369B | 0.397% |
| Bank-wide (RCFD1403 / 2122) | $90.2B | $13,478B | 0.670% |
PE-fund nonaccrual (0.397%) is the highest among NDFI subcategories, approaching the bank-wide rate (0.670%); business credit is near zero. No nonaccrual memo items exist for consumer credit or other NDFIs. Validation logic: since the bank-wide rate matches FDIC QBP to 0.00005pp and individual balances match the FDIC API exactly, the subcategory rates from the same filing data are reliable.
Capital adequacy - Q4 2025Close match; risk-based ratios +0.81pp (CBLR exclusion). Tier 1 +0.39%, equity ratio within 0.04pp; risk-based-ratio gap is structural, not error.
| Metric | Our Value ($M) | FRED QBP ($M) | Diff % |
|---|---|---|---|
| Tier 1 Capital | 2,309,984 | 2,301,055 | +0.39% |
| Total Risk-Based Capital | 2,406,516 | 2,375,481 | +1.31% |
| Total Equity Capital | 2,608,745 | 2,598,635 | +0.39% |
| Ratio | Our Value | FDIC QBP | Diff |
|---|---|---|---|
| Equity Capital Ratio | 10.32% | 10.28% | +0.04pp |
| Core Capital (Leverage) Ratio | 9.31% | 9.26% | +0.05pp |
| ACL / Loans | 1.65% | 1.65% | 0.00pp |
| CET1 Ratio | 14.84% | 14.03% | +0.81pp |
| Tier 1 RBC Ratio | 14.90% | 14.09% | +0.81pp |
| Total RBC Ratio | 15.53% | 15.29% | +0.24pp |
Risk-based ratio note: the FDIC QBP excludes banks with a Community Bank Leverage Ratio (CBLR) election - small community banks that opted for simplified capital reporting and do not calculate CET1 / Tier 1 RBC / Total RBC / RWA. Our data includes all banks, but CBLR banks contribute zero to the RWA denominator, so our computed ratios run slightly higher. None of the 629 NDFI-lending banks are CBLR filers, so NDFI-bank-level risk-based ratios are directly comparable to regulatory thresholds. Source: FDIC QBP Q4 2025 PDF, Table IV-A.
Q1 2026
Banking system totals - Q1 2026Close match. Total assets +0.089%, total loans +0.0004% vs FDIC QBP PDF (FRED ingests later).
| Metric | Our Value | FDIC QBP (PDF) | Diff % |
|---|---|---|---|
| Number of institutions | 4,335 | 4,278 | +1.3% |
| Total assets ($M) | 26,168,430 | 26,145,183 | +0.089% |
| Total loans ($M) | 13,692,553 | ~13,692,500 | +0.0004% |
Same ~1.3% institution-count gap pattern as Q4 2025 (+57 here vs +58). All 622 NDFI-lending banks match the FDIC BankFind API total assets exactly. Source: FDIC QBP Q1 2026 PDF, Table I-A.
Credit quality rates - Q1 2026Near exact. Noncurrent, 30-89 DPD and PDNA rates all within 0.0014pp of FDIC published values.
| Rate | Our Value | FDIC QBP | Difference |
|---|---|---|---|
| Noncurrent rate | 0.98080% | 0.98% | +0.0008pp |
| 30-89 DPD rate | 0.55057% | 0.55% | +0.0006pp |
| PDNA rate | 1.5314% | 1.53% | +0.0014pp |
All three within 0.0014 percentage points (less than 1/7th of a basis point), same residual pattern as Q4 2025. Source: FDIC QBP Q1 2026 PDF ("PDNA loans decreased 3 basis points to 1.53 percent").
NDFI nonaccrual-rate methodology - Q1 2026Mortgage-credit NA up to 0.307%. PE-funds softened to 0.261%; NDFI-bank-wide NA 0.656%.
| Subcategory | NA Amount | Balance | NA Rate |
|---|---|---|---|
| Mortgage credit | $1.1B | $368B | 0.307% |
| Business credit | $0.0B | $435B | 0.009% |
| PE funds | $1.0B | $383B | 0.261% |
| NDFI-banks (context) | $76.4B | $11,640B | 0.656% |
Q1 2026: mortgage-credit NA ticked up to 0.307% (from 0.192% in Q4 2025); PE-funds softened to 0.261% (from 0.397%); business credit essentially zero; NDFI-bank-wide NA dipped to 0.656% (from 0.670%).
Capital adequacy - Q1 2026ACL/loans exact (1.64%); risk-based ratios +0.81pp (CBLR). Same structural CBLR gap as Q4 confirms it is not a calculation error.
| Ratio | Our Value | FDIC QBP | Diff |
|---|---|---|---|
| Equity Capital Ratio | 10.06% | n/a | - |
| ACL / Loans | 1.64% | 1.64% | 0.00pp |
| CET1 Ratio | 14.66% | n/a | - |
| Tier 1 RBC Ratio | 14.72% | 13.91% | +0.81pp |
| Total RBC Ratio | 15.32% | n/a | - |
Same CBLR exclusion as Q4 2025 (+0.81pp on Tier 1 RBC, the exact same gap - confirming it is structural, not a calculation error). NDFI banks subset (622 banks): CET1 13.89%, Tier 1 RBC 13.93%, Total RBC 15.11%; none are CBLR filers. Source: FDIC QBP Q1 2026 PDF ("Tier 1 risk-based capital ratio declined 18 basis points to 13.91 percent").
See it live
This page is the static, readable companion to TREMOR's interactive US NDFI Banks tool, which renders the live charts, the per-bank explorer and drilldowns, the nonaccrual and exposure tabs, and the Q4 2025 / Q1 2026 deep-validation toggle. Open the interactive tool on TREMOR. TREMOR is part of tigzig.com - AI for analytics, databases and macro signals.