Under $10m of EBITDA the market lends 3.25x. A year ago it lent 4.00x.

What question is a lender actually answering?

Two questions, and the answer is the smaller of them. The first is how large a loan the earnings support as a multiple: debt to EBITDA, which sets the size of the facility. The second is whether the cash the business produces services that loan at today's rate, after the things it cannot stop paying: coverage, which decides whether the facility survives credit committee. A borrower is offered the lower of the two answers, and which one binds depends almost entirely on size.

It is worth being blunt about what the first number is a multiple of. Reported EBITDA and underwritten EBITDA are different quantities, and the gap is the most contested item in any financing. The Financial Stability Board notes that stripping adjustments out of private credit borrowers moves leverage from a reported 5x to 6x range to something closer to 7x (Financial Stability Board, Report on Vulnerabilities in Private Credit, 6 May 2026). A capacity answer computed off an add-back schedule the lender has not agreed is not an answer.

The second thing worth naming early is that the answer moved. Total debt for a business under $10m of EBITDA cleared 2.50x to 4.00x in July 2025 and clears 2.50x to 3.25x today, with senior capacity down half a turn over the same year (SPP Capital Partners, Market At A Glance, July 2026, with the July 2025 comparator from the same series). A capacity figure carried over from a conversation eighteen months ago overstates what is available by roughly three-quarters of a turn.

What is the market clearing by size?

Senior debt for a borrower below $10m of EBITDA clears at 2.00x to 2.50x with total debt at 2.50x to 3.25x. Above $10m the bands are 2.25x to 3.75x senior and 4.00x to 5.50x total. Above $25m they are 4.25x to 5.25x senior and 5.00x to 6.50x total (SPP Capital Partners, July 2026). Alongside the debt, lenders require at least 40% base equity capitalization with at least 60% of that in new cash, and independent sponsors are expected to show investment beyond rolled deal fees.

A second grid circulates and it is more generous, so it is worth knowing which one applies to you. Lincoln International prints unitranche leverage of 4.00x to 5.50x for borrowers below $15m of EBITDA and 4.75x to 6.25x at $40m to $100m (Lincoln International Private Credit Snapshot, as of 1 May 2026). The two are not in conflict. They survey different populations: SPP reports actual lower middle market placements, while Lincoln values a portfolio whose median company carries $60.6m of last twelve months adjusted EBITDA and skews heavily sponsor-backed. For a founder-owned business below $10m of EBITDA, SPP is the honest number.

What actually gets written at the small end is more conservative than either grid. One first-lien lower middle market lender reported new platform deals at 2.8x weighted average senior leverage and 29% weighted average loan to value (Capital Southwest, Form 8-K, for the quarter ended 30 June 2026). Another reported a portfolio at median net senior debt to EBITDA of 2.5x on an average portfolio company EBITDA of $11.2m (Main Street Capital, Form 8-K, as of 31 March 2026). Unitranche loan to value sits at 50% below $100m of EBITDA and 55% above it (Houlihan Lokey, as of 30 April 2026).

Which test gives the answer: leverage or coverage?

It depends on size, and the crossover is the useful thing to know. Take a business with $8m of EBITDA. At the top of its leverage band it borrows $26m, and at the top of the pricing band that debt costs 11.15% all-in, which is $2.9m of cash interest. Interest coverage lands near 2.8x and fixed charge coverage near 2.1x. Neither is close to a lender's floor. The leverage multiple is the constraint, and no amount of coverage headroom will buy a fourth turn.

Now take a business with $30m of EBITDA. At the top of its band it borrows $195m at 9.40%, which is $18.3m of cash interest. Interest coverage lands at 1.6x, exactly the median across 2,785 middle-market borrowers and more than $1.2 trillion of debt (KBRA, Q2 2026 Middle Market Compendium, for the twelve months ended 30 June 2026, published 28 July 2026). Fixed charge coverage lands near 1.2x, below the 1.3x that a large middle-market sample was running in the first quarter of 2026 (Lincoln International, as of 31 March 2026). Here the last turn of leverage is not available at any price, because coverage will not carry it.

The practical consequence is that the two ends of the market should prepare differently. A sub-$10m business improves its answer by improving the multiple: cleaner earnings, defensible add-backs, contracted revenue, an asset base that supports a borrowing base. A $30m business improves its answer by improving the denominator of the coverage test: less amortization, more of the coupon in cash-pay senior rather than junior capital, a revolver sized so it does not spring a covenant, and a rate hedge.

What moves the answer?

Revenue quality first. Contracted and recurring revenue supports more debt than transactional revenue at the same EBITDA, because the lender is underwriting the persistence of the cash flow rather than its size. The reverse is currently true of one specific category: software carries 75 to 100 basis points above the standard pricing matrix per one survey, or 150 to 300 basis points above comparable non-software credits per another, with interest-only periods cut from three years to two or less and liquidity minimums and cash controls now standard (Houlihan Lokey, as of 30 April 2026, and SPP Capital Partners, July 2026).

Then the asset base, because it opens a different lane. Asset-based facilities size off collateral rather than earnings, with eligible receivables advancing at 80% to 85%, blended inventory near 50% and machinery and equipment at 50% to 75% of orderly liquidation value (ABF Journal, 1 June 2026). A business with a real borrowing base can often carry more total debt than its EBITDA multiple suggests, which is precisely why middle-market borrowers are being pushed toward asset-based and non-bank lenders as bank hold sizes in leveraged lending have contracted from $75m to $100m down to $30m to $50m (ABF Journal, 19 March 2026 and 1 June 2026).

Then customer concentration, cyclicality and sector, which currently act as a hard filter rather than a pricing adjustment. SPP describes a market traveling in two directions, with high-quality issuers seeing among the most aggressive terms in years while macroeconomic uncertainty disproportionately affects lower middle issuers below $5m of last twelve months EBITDA, storied credits, highly cyclical borrowers and increasingly anything touching software, where leverage capacity is contracting and lender scrutiny is intensifying (SPP Capital Partners, July 2026).

And sponsor backing, which changes the answer in both directions. A sponsor adds equity capacity and a track record the lender can underwrite, and sponsored borrowers are less likely to progress from delinquency to outright default because a sponsor can inject liquidity (Financial Stability Board, 6 May 2026). It also changes what the lender expects. Sponsors currently account for 74% of institutional maturity-extension amendments but only 44% of new-money activity (PitchBook LCD, as of 30 June 2026), so a lender will ask what the sponsor has committed rather than what it could do.

What if the answer is not enough?

There are three honest options and one that is usually a mistake. The first is junior capital, which buys turns at a price: mezzanine and other junior instruments run 13.00% to 16.00% all-in for borrowers below $10m of EBITDA, against 11.00% to 12.00% above $25m, and the small end moved 100 basis points wider at both ends of the range over the year (SPP Capital Partners, July 2026). The second is a government-guaranteed lane, where the Small Business Administration caps 7(a) pricing at base plus 3.0%, roughly a 9.75% ceiling at a 6.75% prime rate, and where the cumulative 7(a) and 504 limit doubled from $5m to $10m effective 4 July 2026, the first increase since 2010, with the single-loan maximum holding at $5,000,000 (computed from the SBA 7(a) program page, August 2026, and SBA, May 2026 announcement).

The third is more equity, and it is often the cheapest of the three once the coverage arithmetic is done. Lenders already require at least 40% base equity capitalization with 60% of it new cash (SPP Capital Partners, July 2026), and a business trying to close the gap with junior capital at 16% is buying an obligation that the fixed charge test will register in full.

The mistake is taking the maximum. Borrowers financed in the 2021 and 2022 vintages now carry leverage around 0.9x higher than at underwriting and adjusted cash interest coverage around 0.4x lower, with refinancing risk for those vintages elevated through 2027 (VRC, Q2 2026). They did not do anything wrong afterward. They took the number the market offered at the top of a cycle, and the interest calendar did the rest.