The Null Institute | Independent Research Note | 13 November 2026
By Ilya Nikitin, PhD. Independent analysis of public SBA records; not peer-reviewed.
The small-loan decline is real. The idea that every lender is abandoning small borrowers is not. We independently examined 403,644 records in the U.S. Small Business Administration’s 7(a) loan-level file, covering fiscal years 2020–2026. Approvals for $50,000 or less fell 34.7% from FY2025 to FY2026, while approvals above $1 million barely changed.
Then we examined which lenders accounted for the change. Five lender names made up 88.2% of the national net decline in small approvals. Outside those five, the number of small approvals also fell, but their share of all approvals at those lenders actually rose from 17.6% to 19.4%. That is a very different picture from a uniform industry retreat.
What changed in SBA 7(a)?
| Approvals recorded by SBA | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| All sizes | 70,242 | 78,078 | 56,130 |
| $50,000 or less | 20,186 | 18,722 | 12,222 |
| Over $1 million | 7,743 | 8,958 | 8,872 |
| Small approvals as a share of all | 28.7% | 24.0% | 21.8% |
From FY2025 to FY2026, the number of approvals at $50,000 or less dropped by 6,500 (−34.7%), compared with just 86 (−1.0%) above $1 million. The shift towards larger approval amounts within 7(a) is clear. It had also begun before FY2026: the small-loan share was already down from 28.7% in FY2024 to 24.0% in FY2025.
Five lenders, two ways to describe the same decline
We ranked lenders by the absolute decrease in their number of 7(a) approvals for $50,000 or less. The five biggest drops were:
| Lender name in SBA file | FY2025 | FY2026 | Change |
|---|---|---|---|
| The Huntington National Bank | 2,802 | 1,079 | −1,723 |
| Northeast Bank | 2,115 | 435 | −1,680 |
| Wells Fargo Bank National Association | 1,058 | 82 | −976 |
| TD Bank, National Association | 1,875 | 1,029 | −846 |
| Readycap Lending, LLC | 1,326 | 816 | −510 |
| Five combined | 9,176 | 3,441 | −5,735 |
| All other lenders | 9,546 | 8,781 | −765 |
The five contribute 5,735 / 6,500 = 88.2% of the national net fall. But that is not 88% of all individual lender declines. Across lenders, decreases totalled 8,010 small approvals, partly offset by 1,510 increases at other lenders. On this gross-decline denominator, the same five account for 71.6%. Both numbers are accurate; they answer different questions.
We selected these five after seeing the changes. The concentration is a descriptive finding, not a causal explanation of why any institution changed its lending.
Outside those five, the small-loan share went up
| Group | FY2025: share of approvals ≤$50k | FY2026: share of approvals ≤$50k |
|---|---|---|
| Five lenders above | 38.6% | 31.4% |
| All other lenders | 17.6% | 19.4% |
For the other lenders, small approvals went from 9,546 to 8,781, a decline of just 8.0%. Their total 7(a) approvals fell even faster, from 54,311 to 45,181. That is why the small-loan share increased despite fewer approvals in absolute terms. Restricting this calculation to lender names present in both years produces virtually the same result: 17.6% to 19.5%.
Even the five selected lenders do not tell one uniform story. Readycap Lending’s small-loan share increased from 42.3% to 54.6%, although its small-loan count declined from 1,326 to 816. Its overall 7(a) activity contracted sharply. Counting fewer small loans is not by itself evidence that a lender is shifting away from small loans.
Across the five lenders combined, all 7(a) approvals fell from 23,767 to 10,949 (−53.9%), alongside the decline in their small approvals from 9,176 to 3,441 (−62.5%). Using their FY2025 small-loan share as a fixed benchmark, approximately 4,938 of their 5,735 missing small approvals can be attributed arithmetically to lower overall approval volume; the remainder reflects a different loan-size mix. That calculation is a decomposition, not proof of the economic cause.
Two checks against misleading comparisons
Government shutdown: The SBA reports a 43-day federal shutdown that halted lending activity early in FY2026. To avoid directly comparing those interrupted months, we also compared the identical January–September periods. Small approvals fell from 13,574 in 2025 to 9,794 in 2026 (−27.8%), while approvals above $1 million increased from 7,083 to 7,467 (+5.4%). The difference persists, but this does not remove delayed applications, rule changes or demand effects.
Cancelled approvals: The primary tables count all approval records, even those subsequently cancelled. Excluding records already marked CANCLD leaves 15,202 small approvals for FY2025 and 10,649 for FY2026 (−30.0%). But FY2026 loans have had less time to be cancelled, so these figures are not directly comparable measures of completed financing. Neither table measures applications rejected or money disbursed.
Is this happening across US business lending?
We do not know from this dataset. SBA 7(a) is a federal loan-guarantee programme. Its approvals are not a census of ordinary bank lending or a measure of every loan available to a small business. The FFIEC’s separate CRA series reports approximately 8.73 million business-loan originations in calendar 2024 and 8.56 million in 2025, about a 2.0% decline. That series has different reporting institutions, products (including eligible business credit cards) and years; it cannot be combined with these 7(a) figures.
Nor should the word microloan be used carelessly. A 7(a) approval of $50,000 or less is not a loan from the SBA’s distinct Microloan programme, which works through nonprofit intermediaries. And the amount approved is not a direct measure of a borrower’s business size.
What the evidence does, and does not, establish
Established: the 7(a) approval mix shifted away from the smallest amounts; five lender names account for a large part of the overall count decline; the remaining lenders’ small-loan share rose; and the direction survives two simple checks involving matched calendar months and cancelled approvals.
Not established: that banks made a deliberate decision to abandon small borrowers, that underwriting small loans became unprofitable, that rejected applications increased, or that business owners could not obtain equivalent financing elsewhere. To answer those questions, we would need application and denial records, comparable lender-level costs, changes in SBA rules and evidence on borrower demand.
Conclusion: small SBA-backed approvals are falling sharply, but the phrase “lenders are abandoning microloans” treats a complicated, concentrated shift as an industry-wide decision. The released records do not support that leap.
Data, methods and limitations
We analysed a user-supplied copy of the SBA 7(a) FOIA dataset, file FOIA_7a_FY2020_Present_asof_260930.csv, with 403,644 records and an as-of date of 30 September 2026. The file’s SHA-256 is b56194df6da2c22c318fcfc9f8f087152bc54e2be391c4cc70316514c37ef8e6. We checked internal annual totals and repeated the main calculations independently. We did not compare the file byte-for-byte against an official SBA checksum.
The calculations count approvals by ApprovalFY, GrossApproval, BankName, ApprovalDate and LoanStatus. Lender names are imperfect identifiers across legal reorganisations. Rankings of the five largest declines were chosen after examining the full dataset. We did not release borrower-level records. All percentage changes are descriptive measurements of this file, not causal estimates or forecasts.
Other primary references: SBA FY2026 statement (2 November 2026); FFIEC 2025 CRA release (5 November 2026); SBA Microloan programme definition.