You need to enable JavaScript to run this app.
By Travis Yule — CEO & Founder, Full Send Funding
Fed 2026 Small Business Credit Survey: 38% of employer firms applied, 42% of applicants got the full amount, 22% got none; most applied for operating expenses.
In one sentence: The small-business credit-access figures worth citing are the Federal Reserve Banks' own survey figures read with their frame attached: 38% of employer firms applied for credit, 42% of applicants got the full amount, 36% part and 22% none, and operating expenses, not growth, was the most common reason to apply.
In the Federal Reserve Banks' 2026 report on employer firms, 38% of small employer firms had applied for a loan, line of credit or merchant cash advance in the previous twelve months, and of those applicants 42% received the full amount they sought, 36% received some or most of it, and 22% received none.[1] Put on a base of 100 firms, that is 16 that got everything they asked for, about 14 that got part of it, about 8 that got nothing, and 62 that did not apply at all. The most common reason for applying was not growth. It was meeting operating expenses, cited by 56% of firms that sought financing, ahead of pursuing an expansion or new opportunity at 46%.[1]
Those are the small-business credit-access statistics that can be stated with a source attached, and this page is built on them, on two cash-flow challenge shares from the prior year's survey, and on nothing else. The number most often quoted in this territory — that some large share of small-business failures are caused by cash flow problems — is not on this page, because we cannot find the dataset it came from, and neither can anyone who quotes it. What follows is narrower and more useful: what the Federal Reserve's survey actually measured in the autumn of 2025, on what sample, what each figure can and cannot support, and what the arithmetic looks like when the shares are turned back into firms and dollars.
The frame comes first, because the numbers mean nothing without it.
The Small Business Credit Survey is an annual survey, run by the twelve Federal Reserve Banks, of firms with fewer than 500 employees.[1] The 2026 report on employer firms is drawn from the 2025 survey, which was fielded from September 3 to November 14, 2025 and produced 6,525 responses from employer firms with one to 499 employees across all fifty states and the District of Columbia. The report was released on March 3, 2026.[1]
Three features of that frame govern everything below.
It is a convenience sample. Respondents reach the survey through partner organizations and choose to answer. They are not drawn at random from a list of every employer firm in the country. A percentage from a convenience sample describes the firms that answered; how closely it describes all employer firms depends on who answered, which a random draw would have settled by design and a convenience draw cannot. The survey is the broadest regular measurement of small-firm finance in the United States, and it is still a description of respondents.
It is employer firms only. A firm with no employees other than its owner — most of the country's businesses by count — is surveyed separately and reported separately. Every figure on this page is about firms with at least one employee and at most 499.
It is self-reported and it is retrospective. Firms report what happened in the twelve months before they answered. The outcome of an application is the applicant's own account of it. Nobody audited the bank statements.
None of that makes the figures weak. It makes them specific, and specific is what a statistic has to be before it is worth citing.
Thirty-eight percent of employer firms applied for a loan, line of credit or merchant cash advance in the twelve months before the survey.[1] The other 62% did not.
That 62% is the largest single group in the data and the least discussed. The survey also asks non-applicants why they did not apply; we quote none of those shares here, because none of them appeared in the extracts we confirmed, and a number this page cannot show the source for does not go on it. What the 38% establishes is the denominator. Every outcome share that follows — full amount, some or most, none — is a share of applicants, not a share of firms. Reading "22% received none" as "22% of small businesses are turned away" overstates the turned-away population by a factor of about 2.6 — 22 against 8.4 — and it is the most common misreading of this report.
Our arithmetic on two published figures from the 2026 report on employer firms: 38% of employer firms applied, and of applicants 42% received the full amount, 36% some or most and 22% none. Each published share is rounded to a whole point, so every derived row carries about half a firm of rounding either way.
Read the bottom three rows against the top one. Of 100 employer firms, roughly 16 sought financing and got all of it, roughly 14 got some or most, and roughly 8 got none. The 8 are the firms that experienced the credit market as a wall. The 14 are the firms that experienced it as a partial answer — approved, but for less than the problem — and that middle group is the one the rest of this page is mostly about, because a partial approval is the outcome our own files see most often when a request is sized to a wish rather than to revenue.
Fewer than half of applicants received the full amount they sought; more than a third received part of it and just over a fifth received none.
The three outcome shares published in the 2026 report on employer firms for firms that applied for a loan, line of credit or merchant cash advance in the prior twelve months. The categories are exhaustive, so the three shares sum to 100%.
The three shares are the spine of the report for anyone in our business, and each deserves to be read on its own terms.
42% received the full amount.[1] This is the share of applicants for whom the process worked as advertised: they asked for a number and were approved for it. It is worth noticing that this is a minority. On the survey's own evidence, the normal experience of applying for small-business credit is not getting what you asked for.
36% received some or most.[1] The survey groups "some" and "most" into one category in the headline figure, and it does not publish the dollar shortfall behind it — a firm approved for 90% of its request and a firm approved for 30% sit in the same bucket. What the category does establish is that partial approval is a normal outcome, not an edge case: more than a third of applicants got a smaller number than the one they applied for.
22% received none.[1] Just over one applicant in five received none of the financing it sought. The survey does not attribute this to credit score, revenue, time in business or anything else in the headline figure, and we do not either. A decline is a decision made by a specific lender on a specific file against that lender's own bar — ours are listed, with the line behind each — and the survey records the outcome, not the reasoning.
Two things this figure cannot support are worth stating outright, because both get asserted on the strength of it.
It cannot support a comparison with any single lender's approval rate. Our own published rate — roughly 90% of applicants are approved — is measured on a different population, by a different method, against a different product. Applicants to a revenue-based working capital funder are not a random draw of employer firms, and "approved" on our side means an offer was made, not that the offer matched the request. The two numbers describe different things and should not be put in the same sentence as if one were the other. If you want to know what our bar actually is, how working capital underwriting works states it from the decision side.
It cannot support a claim about who is being turned away. The demographic and industry cuts exist in the survey's chartbooks, and they are real, but a share of one group receiving none of the financing it sought tells you the outcome for that group and nothing about the cause of it. The survey did not design a test of cause, and a statistic that travels without its sample definition attached does more harm than good. We leave those cuts to the report itself.
The most common reason firms sought financing was to meet operating expenses, at 56%; the second was to pursue an expansion or new opportunity, at 46%.[1] The two add to more than 100 because a firm could give more than one reason, which is the first thing to understand about them: they are not two populations.
That matters because the usual reading is that 56% is the share in trouble and 46% is the share doing well. The arithmetic of growth says otherwise. A business that grows faster than its margin can fund pays for the additional labour, materials and inventory before it collects for them, and the money it borrows to cover that is, on the survey's form, financing to meet operating expenses — in a firm that is expanding. How much of the 56% is that firm and how much is a firm with a genuine operating loss, the survey does not ask, and we do not know. The mechanism is worked through in why profitable companies run out of cash.
What the 56% does establish is the shape of the need: the most common reason a small employer firm seeks outside money is to meet operating expenses that fall due before the revenue covering them arrives — payroll every Friday against invoices paid in forty-five days. Where that revenue exists and is merely late, it is a timing gap, which is the case working capital funding is sized against; where the month's costs exceed the month's revenue, it is a margin problem, and the survey's question does not separate the two.
The prior year's report, on the 2024 survey, found that more than half of employer firms had experienced difficulty paying operating expenses (56%) or uneven cash flow (51%) as a financial challenge in the previous twelve months.[2] The survey's challenge question offers, among its options, paying operating expenses and uneven cash flow;[1] those are the two options that describe a cash-flow problem directly, and they are the two that half the sample ticked.
We quote the 2024-survey figures for those two options rather than the 2025-survey figures because the 2024 figures are the ones we could confirm in a report's own text. The 2026 report also publishes its account of the year's financial challenges and of what firms expected for the year ahead. We read summaries of both and could not confirm the figures in them against the report itself, so neither is quoted here — a page that carries a number it cannot show the source for is the thing this page exists not to be. The two shares that are quoted are a year older than the credit-access figures above, and the prose says so each time they appear.
What the two shares can support is narrower than it looks. "Difficulty paying operating expenses" and "uneven cash flow" are self-reported challenges, and the survey does not ask whether the difficulty was a timing gap — revenue that exists and arrives late — or a structural one, where the month's costs exceed the month's revenue and no timing fixes it. For a funder that distinction is the whole question. Borrowing bridges the first and worsens the second, because it adds a payment to a month that was already short. The survey cannot say which of the 56% is which, and neither can we from the outside; three months of bank statements can, and when borrowing is the wrong answer is the test to run before applying.
The firm's reading of the survey's design. The left column paraphrases the report's headline questions; the other two are the limits that follow from a self-reported, retrospective convenience sample of respondents.
The right-hand column is the one to read before quoting any figure from this page. Every row in the survey is a share of respondents to one question; none of them is a rate measured against a population, an amount measured in dollars, or a cause established by design.
Three absences deserve a sentence each, because each is a number people expect a page like this to carry.
There is no public dataset for the share of small-business failures caused by cash flow. The figure that circulates traces to no public series we can find. No Federal Reserve, Census Bureau, Bureau of Labor Statistics or Small Business Administration publication records why a closed business closed, or whether it was profitable when it did. A page that quotes a share anyway is quoting a number nobody can check.
There is no public small-business benchmark for receivable days by industry. The only official U.S. series of receivables and sales by industry samples corporations far larger than the firms in this survey, and the survey itself asks about financing and challenges, not balance-sheet days. The reasons, and the arithmetic you can do on your own books instead, are in the working capital cycle.
There is no dollar figure behind "some or most." The survey measures the outcome against the request, and it records the applicant's classification of it. A firm that asked for twice what its revenue supports and was approved for what its revenue supports is a "some or most" — which is why the next section exists.
The survey does not publish amounts, so the following is ours. It uses the firm's published sizing range and stated inputs, and any of the inputs can be replaced with a reader's own.
An electrical contractor deposits $150,000 a month — $1,800,000 a year. Its largest general contractor accounts for $100,000 of those monthly billings and has moved from paying at thirty days to paying at sixty. The contractor applies for $200,000 to carry operating expenses through the quarter in which that happens. Under the published sizing range of 80% to 150% of monthly revenue, the file supports between $120,000 and $225,000, with the position inside that range set by deposit consistency, negative days and existing positions rather than by the request.
Stated inputs: $150,000 of average monthly revenue, of which $100,000 a month is billed to one customer that has moved from thirty-day to sixty-day payment, and a $200,000 request. Approved = monthly revenue × the row's percentage, capped at the request; shortfall = request − approved; the share is shortfall ÷ request. The timing gap is one extra month of that customer's billings, $100,000, and the gap column is approved − $100,000. The survey category is how the applicant would classify each row on the survey's form, which measures against the request.
Three things follow from the table.
The applicant experiences four different outcomes; the survey records three, and two of them identically. An approval at $120,000 and an approval at $150,000 are both "some or most" on the survey's form. From the contractor's chair, one is 60% of the request and one is 75%, and the difference — $30,000 — is a fifth of a month's deposits.
Whether the partial approval is a problem depends on the gap, not the request. The timing gap is one extra month of that customer's billings outstanding: $100,000 billed each month and held for an additional thirty days is $100,000 the contractor is short, for as long as the slow payment lasts. A $120,000 approval covers the whole of that with $20,000 to spare. It is recorded as "some or most" because the request was sized to a round number, and the survey measures against the request. The question worth asking before applying is how much the gap actually is; how much your business can actually borrow is the arithmetic for it.
The decline is the only row that leaves the contractor short of the gap. It is also the row that a funder reading revenue rather than credit score produces least often on a file like this one, because the deposits are consistent and the shortfall is a timing gap with a named cause and a named end. Where the same file gets declined, in our experience, it is because of existing positions or negative days rather than because of the request — a finding from our own files, not from the survey, which records outcomes and not reasons.
Speaking for how the survey looks from the decision side of a working capital file:
What we fund and what it costs is stated on how we make money, and whether your revenue clears the bar takes a minute on the qualification check.
The survey is the best public measurement of its subject and it is still the wrong instrument for several jobs it gets used for.
In the Federal Reserve Banks' 2026 report on employer firms — a convenience sample of 6,525 firms with one to 499 employees, fielded from September to November 2025 — 38% had applied for a loan, line of credit or merchant cash advance in the prior twelve months, and of those applicants 42% received the full amount, 36% some or most, and 22% none. On a base of 100 firms that is 16, 14 and 8, with 62 not applying at all. The most common reason for seeking financing was operating expenses, at 56%, ahead of expansion at 46%; the prior year's survey found 56% of firms had difficulty paying operating expenses and 51% experienced uneven cash flow. None of it is a rate against a population, a figure in dollars, or a cause — and the number most often quoted about small-business cash flow, the share of failures it causes, has no public source at all.
If the reason you are reading this is a timing gap of your own, check what your revenue would qualify for — a minute, no credit impact — or call 518-312-0382 and ask directly.
Every survey figure on this page is quoted from two Federal Reserve Banks publications: the 2026 Report on Employer Firms, drawn from the 2025 Small Business Credit Survey,[1] and the 2025 Report on Employer Firms, drawn from the 2024 survey.[2] The survey is an annual convenience sample of firms with fewer than 500 employees; the 2025 wave was fielded from September 3 to November 14, 2025 and produced 6,525 responses from employer firms with one to 499 employees across all fifty states and the District of Columbia. The figures are self-reported and retrospective, and the article states that frame before any of them. Where a figure the report publishes did not appear in the text we were able to confirm — the reasons non-applicants gave, the 2025-survey shares for the operating-expense and uneven-cash-flow challenge options, the 2026 report's findings on costs and on expectations for the year ahead, and every demographic or industry cut — it is not quoted, and the prose says so rather than supplying a substitute.
Two things are derived rather than quoted. The per-100-firms table multiplies the published application share by each published outcome share (38 × 42%, 38 × 36%, 38 × 22%) and subtracts the application share from 100; each input is rounded to a whole point in the report, so each derived row carries about half a firm of rounding. The worked example applies the firm's published sizing range of 80% to 150% of monthly revenue to stated inputs — $150,000 of monthly revenue, of which $100,000 a month is billed to one customer that has moved from thirty-day to sixty-day payment, and a $200,000 request — and nothing else; its dollar figures are illustrations of how the survey's outcome categories map onto a request, not survey findings. The site's arithmetic audit recomputes every published value in both from those inputs before publication.
No causal claim is made on the survey's evidence. The survey records outcomes and self-reported challenges; it did not design a test of why any application was declined, why any firm experienced a challenge, or what happened after fielding. Where the article states a reason — that outright declines on the firm's own files are dominated by existing positions, negative days and open defaults, or that a partial approval usually follows a request sized above revenue — it is labelled as the firm's reading of its own files and is not attributed to the survey.
In the Federal Reserve Banks' 2026 report on employer firms, 38% of employer firms had applied for a loan, line of credit or merchant cash advance in the twelve months before the survey, which was fielded from September to November 2025. The figure describes a convenience sample of 6,525 firms with one to 499 employees, not every small business in the country.
Forty-two percent of applicants in the 2025 survey received the full amount of financing they sought, 36% received some or most of it, and 22% received none. Those are shares of applicants; measured against all employer firms, about 16 in 100 got the full amount, about 14 got part, and about 8 got nothing.
The most common reason in the 2025 survey was to meet operating expenses, cited by 56% of firms that sought financing; the second was to pursue an expansion or new opportunity, at 46%. Firms could give more than one reason, so the two overlap rather than describing separate groups.
In the Federal Reserve Banks' 2025 Report on Employer Firms, on the 2024 survey, 56% of employer firms cited difficulty paying operating expenses and 51% cited uneven cash flow as financial challenges in the prior twelve months. This page quotes the prior-year figures because they are the ones confirmed in a report's own text; the 2025-survey shares for those options, and the 2026 report's cost and expectations findings, are not quoted.
No. It is a convenience sample: firms reach it through partner organizations and choose to respond. Its percentages describe the respondents, and how closely they describe all employer firms depends on who answered. It is still the broadest regular public measurement of small-firm finance in the United States, which is why this page uses it and nothing else.
There is no public dataset that answers this. The figure that circulates traces to no Federal Reserve, Census Bureau, Bureau of Labor Statistics or Small Business Administration series, and none of those agencies records why a closed business closed or whether it was profitable when it did. A page that quotes a share anyway is quoting a number nobody can check.
It does not, and the two should not be put in one sentence. The survey pools every lender type and product into one distribution of self-reported outcomes among a convenience sample of employer firms. A single funder's approval rate is measured on its own applicants, by its own method, against its own product, and "approved" means an offer was made, not that it matched the request.
Travis Yule founded Full Send Funding in 2021 and leads it from Middle Grove, New York. He writes about working capital from the underwriting side of the table — what the numbers actually have to say before a business gets funded.