The macro picture is calm enough to lull a casual observer. Unemployment was 4.2 percent in September, according to the Bureau of Labor Statistics, and layoffs are low. Most large companies tell surveyors that their margins look normal. Yet in the Federal Reserve Banks' 2026 Small Business Credit Survey, owners describe higher input costs, trouble finding staff and growing doubt about whether expansion is worth the risk. Both pictures are accurate. The question is how a small firm can be 'fine' in the aggregate data and still be financially cornered.

The usual answer is that tariffs raised costs and owners are gloomy. That is true but incomplete. The less-examined mechanism is the link between two bills that arrive at the same time: the higher price of imported inputs, and the higher price of the short-term credit used to carry those inputs until they are sold. For a small firm, the cost of the goods and the cost of financing the goods stack on top of each other. The result is not a wave of closures. It is slow erosion of cash, owner capacity and willingness to invest.

A resilient headline, a strained ledger

The 2026 Report on Employer Firms, drawn from the 2025 Small Business Credit Survey, is notable for what it does not show. Revenue and employment growth held steady year over year, and firms were only slightly more likely to report falling revenue than rising revenue. What deteriorated was the outlook. The revenue expectations index fell from 39 to 33, and the employment expectations index fell from 26 to 23, both the lowest since 2020. Reaching customers and growing sales was the most common operational challenge. Hiring or retaining qualified staff came second.

The cost side is more specific. Rising costs of goods, services or wages was the most common financial challenge, and more than four in ten firms said tariff-related costs were a problem. Seventy-seven percent reported one or both. Tariff pain was concentrated in retail (69 percent) and manufacturing (62 percent). The survey was fielded from September to November 2025, so it captures a period when tariff exposure was near its peak.

The imported-input cost is also a financing cost

Forty-eight percent of firms sourced at least some inputs from outside the United States, and 14 percent sourced more than half. A large majority of those firms saw input prices rise. Seventy-six percent passed some of the increase to customers, but 60 percent absorbed some of it. Few changed their supply chains: 13 percent moved to domestic suppliers, 8 percent to different foreign suppliers and just 3 percent relocated production to the United States.

That last set of numbers matters for the financing argument. If firms cannot easily switch suppliers, a higher landed cost becomes a standing working-capital requirement. Customs duties are generally owed when goods clear the border, well before the finished product is sold. An importer that once financed $100,000 of inventory now finances more, for the same physical volume, for the same number of days. This is my inference rather than a finding of the survey, but it is consistent with what firms report borrowing for. Fifty-six percent of applicants sought financing to meet operating expenses, versus 46 percent for an expansion or new opportunity. Eighty-six percent of firms use financing regularly, most commonly credit cards and loans.

The cost of that financing is not uniform. Applicants who went to online lenders were far more likely to be surprised by the price: 60 percent said actual borrowing costs were higher than expected, versus 37 percent at small banks and 32 percent at large banks. Online fintech lenders drew 29 percent of loan, credit line and cash-advance applicants in the 2025 survey, up from 17 percent in 2020. Small banks fully approved 57 percent of their applicants, more than other lender types. The firms most likely to need fast cash for operating expenses are therefore also the ones most likely to take the most expensive route to it.

Absorption before amputation

What does a firm do when costs and financing charges rise together, but demand has not yet collapsed? The evidence points to a sequence in which layoffs and closures come late.

First, firms trim margins. Sixty percent of foreign-input users said they absorbed some of the increase. Second, they draw on cash. In the Richmond Fed's CFO Survey commentary of September 23, 2026, Zach Edwards and Daniel Weitz report that nearly half of unconstrained firms said sufficient cash let them rely less on external financing. Third, owners put themselves behind the loan. Among SBCS firms with debt, 59 percent used a personal guarantee and 51 percent used business assets. Fourth, they delay investment and slow hiring rather than cut staff. That is the pattern the falling expectations indices describe.

This sequence explains why labor data look healthy. The September jobs report showed payrolls up only 29,000, a weak reading, but the unemployment rate has stayed in a narrow band. The JOLTS release for August shows the layoff and discharge rate at about 1.0 percent. Small firms do not need to shed workers to stay solvent in the near term. They can absorb the shock in the owner's equity, the owner's credit score and the business's cash cushion, none of which appears in a payroll report.

Who bears it: size, sector and collateral

The burden is not evenly spread. Four dividing lines stand out in the official data.

Size. The CFO Survey found that roughly double the share of smaller firms (under 500 employees) report being financially constrained compared with larger firms. Overall, only 19 percent of respondents said financing access or cost constrained their plans, down from nearly 30 percent in the second quarter of 2023. Among constrained firms, more than 60 percent said the constraint stopped them from pursuing new business opportunities. In other words, financial strain shows up as opportunities not taken, not as defaults.

Industry and import reliance. Retail and manufacturing report the highest tariff-cost challenges. These are also sectors where inventory and materials must be bought in advance, which is where a financing cost bites hardest.

Financing access. Of SBCS applicants, 42 percent received all the financing they sought, 36 percent received some or most, and 22 percent received none. Firms rejected or partly funded are the ones forced into cards and online products.

Pricing power and collateral. The survey does not measure pricing power directly, so I use pass-through as a proxy. The 76 percent that passed costs on are the firms with some pricing power. The 24 percent that did not pass any of it on, plus the firms that passed on some but absorbed the rest, are the exposed group. On collateral, the heavy use of personal guarantees suggests that many firms lack business assets sufficient to secure credit by themselves. The Federal Reserve's July 2026 Senior Loan Officer Opinion Survey sharpens the point. Banks reported unchanged standards for commercial and industrial loans to small firms, and standards were easier than their historical midpoint. Standards for very small firms were near the midpoint, and standards on credit cards tightened. A firm with bankable collateral faces a benign market, while a thin-collateral firm that relies on a card faces a tighter one.

The vacancy paradox

The apparent contradiction is that firms report unfilled positions but little confidence in expansion. The BLS data suggest that the two are less opposed than they seem. JOLTS shows 7.1 million openings in August, a rate of 4.3 percent, against 5.2 million hires, a hires rate of 3.3 percent. That is about 1.36 openings per hire (my calculation from those two figures). The quits rate was only 1.9 percent. A market with high vacancy-to-hire ratios and low quits is not a hot, expanding market. It is a stuck one, where workers stay put and firms struggle to fill the seats that do open.

For a small firm, then, an opening is more likely a replacement or a retention problem than a growth plan. Hiring or retaining qualified staff is a real challenge in the SBCS, but a firm can want a machinist back without wanting to build a second shift. Expansion requires capital, demand visibility and tolerance for debt, and those are the inputs under pressure. The smallest establishments also show volatility. In August, the openings rate for establishments with one to nine employees fell from 6.0 percent to 4.3 percent, a reminder that posted vacancies at tiny firms can disappear when finances tighten, not only when jobs are filled.

Evidence that cuts against the thesis

The financing-squeeze argument has real counterevidence, and it deserves weight.

The CFO Survey finds that about half of firms report 2025 and 2026 gross margins within one percentage point of what they call normal, and three-quarters are within three points. The authors write that a sustained, broad-based margin squeeze 'is not borne out in the cross-section of survey responses'. Most firms reported higher spending over the past three months, the largest share in four years. Net hiring intentions are near 70 percent, and only about 7 percent of firms plan workforce reductions. Even among smaller firms, roughly 80 percent said credit costs are not a constraint.

The tariff backdrop has also changed. The Supreme Court struck down the IEEPA tariffs on February 20, 2026. Using USITC DataWeb figures, the Penn Wharton Budget Model estimated an effective tariff rate of 7.1 percent in March 2026, the lowest since March 2025, against 2.3 percent in January 2025. A temporary Section 122 surcharge then expired on July 24 and was replaced by new Section 301 duties, with the Yale Budget Lab putting the average statutory rate near 11 percent afterward. The cost shock therefore did not vanish. It changed legal form and may have eased for some importers while hardening for others. Bank credit supply, meanwhile, is not tight for firms with collateral, and the SBCS shows 31 percent of firms carry no debt at all.

The honest reading is that the squeeze is real but concentrated. It is a story about a minority of firms, probably 20 to 25 percent, that are import-dependent, collateral-light and short of pricing power. The mass of small firms is not in distress.

Limits of the data

Several limitations should temper any strong conclusion. The SBCS is a convenience sample of 6,525 firms, not a random one, and the Federal Reserve itself cautions that results may be biased. Its input-sourcing question refers to 2024, while the survey was fielded in fall 2025, before the tariff regime changed. The public findings do not cross-tabulate import reliance with financing outcomes, so the link between imported-input costs and credit-card or online-lender use is inferred, not measured. I did not extract Census Bureau Business Trends and Outlook Survey results by employment size class, which could test the size gradient directly, so readers should consult that biweekly dataset. The CFO Survey respondents are finance executives, a group that likely skews toward firms with dedicated finance staff, so it may understate stress at the smallest firms. JOLTS size-class data are noisy month to month, and the August drop for micro-establishments could partly reflect sampling variation. Finally, margins 'near normal' can mask the cost of financing if owners define normal after years of adjustment.

What owners can do with this

The practical implications follow from the mechanism. Treat the carrying cost of imported inventory as a line item. A product's true cost includes duty, freight and the interest on the days between payment and sale. Firms that price on landed cost alone will keep absorbing a financing charge they cannot see. Move recurring working-capital needs off cards and merchant cash advances where possible. The SBCS shows small-bank applicants had the highest full-approval rate and the fewest cost surprises, so a relationship line of credit, arranged before it is needed, is cheaper than emergency borrowing. Shorten the cash cycle through supplier terms, smaller and more frequent orders, and faster invoicing. Finally, separate retention hiring from expansion hiring. Filling a critical vacancy can protect revenue even when growth capital is scarce, while expansion should wait until pricing and financing costs are known.

For lenders and policymakers, the lesson is that stress at small firms appears first as foregone investment and rising owner exposure, not as job losses. By the time it shows up in payroll data, the cushion that was hiding it is already gone.