Most buyers meet the Debt Service Coverage Ratio (DSCR) as one line in a term sheet: coverage of at least 1.25×. Then they open their model, divide EBITDA by the annual loan payments, get 1.7×, and assume the financing is settled. It is not. The lender is dividing a smaller number by a larger one, and on the same deal the lender's answer can sit well below the buyer's.
This article sets out the ratio the way a credit analyst builds it: what goes in the numerator, what goes in the denominator, where the floor comes from, and how one decision about a seller note moves a deal from fundable to declined. The example is a US deal financed with an SBA 7(a) loan, because the program lends up to $5M on a single loan and its rules are written down where anyone can read them.
The ratio, and why EBITDA is the wrong numerator
DSCR answers one question: for every dollar of debt payments due this year, how many dollars of cash does the business produce to pay them?
The shortcut version puts EBITDA on top. EBITDA is a reasonable starting point, and a poor finishing point, because a business cannot pay its lender with money it has already spent on other things it cannot avoid. Equipment wears out and has to be replaced. Tax is owed. A growing business ties up more cash in receivables and inventory. None of those appear in EBITDA, and all of them are paid before the lender is.
So the numerator a lender trusts is cash available for debt service (CADS):
Three points about the inputs:
- Adjusted EBITDA means EBITDA after the normalizing adjustments you can evidence, with a market-rate salary in the owner's seat. If the listing quotes seller's discretionary earnings, convert it first; SDE and EBITDA are different figures and coverage built on SDE overstates what is left for the lender by one owner's pay.
- Maintenance capex is what it costs to keep the business as it is, not growth spending you choose to make. Expect a lender to compare your figure with several years of fixed-asset additions, and to question a zero.
- Working capital is deducted when it grows. If the business releases working capital, a careful analyst does not add that back as if it will repeat.
The denominator is every required payment of principal and interest in the year, on every piece of debt the business owes after closing. Not the senior loan alone.
Where the 1.25× floor comes from
For an SBA 7(a) loan that finances a change of ownership, the floor is a rule. The SBA's lending rulebook, SOP 50 10 8.1 (published 25 September 2026 with technical policy updates, effective 1 October 2026), gathers every change-of-ownership rule into one place, Appendix 15, and sets the minimum coverage by transaction type (Appendix 15, Para. C.2.b):
- Initial Acquisition: 1.25:1
- Business Expansion: 1.15:1
- Owner Buyout: 1.25:1
- ESOP and Cooperative: 1.25:1
A buyer purchasing a business they did not already own is an Initial Acquisition, the SOP's default category, so the floor for the deal in this article is 1.25×. The ratio "must be satisfied using either the last fiscal year-end or an average of the last two fiscal year-end statements on either a historical or adjusted basis". Projections do not count: the lender "must evaluate the Applicant's post-closing financial projections but may not rely on them to meet the DSC requirement", with one exception for an owner-occupied special purpose property such as a hotel, where the appraised real estate fully secures the loan. The SBA publishes the full document.
On the numerator the rulebook is more generous than a credit analyst. Its historical ratio is "earnings before interest, taxes, depreciation, and amortization (EBITDA) divided by the combined debt service post-transaction", the shortcut this article opened with. Its adjusted ratio lets the lender add and subtract, with a written justification for each item in the credit memorandum: unfunded capital expenditures, non-recurring income, distributions, distributions for S-Corp taxes, seller discretionary expenses and ownership compensation. An adjustment without that analysis "will be ineligible". Capex and tax distributions are on the list by name. Working capital is not. A lender's credit policy usually goes further than the SOP minimum, and 1.25× on CADS is a stricter test than 1.25× on EBITDA. Expect the stricter one.
The 1.15× figure is still in the rulebook, as the general 7(a) credit standard for loans that do not finance a purchase (Section B, Chapter 1): coverage "equal to or greater than 1.15 on a historical and/or projected cash flow basis". Until 30 September 2026 acquisitions were underwritten to that standard. From 1 October, Appendix 15 governs any 7(a) loan that finances a change of ownership, and 1.25× on the business's own past results is the rule. Lenders set their policy at or above it; ask for theirs, and for the cash flow definition it is measured on, before you build the structure around a number. This article was updated on 15 October 2026 for the new SOP.
The arithmetic behind the floors is worth holding in your head. Coverage of 1.25× means cash can fall by 20% before the business can no longer make its payments. At 1.15×, the expansion floor, the cushion is 13%. At 1.50× it is a third. A lender choosing a floor is choosing how bad a year the loan has to survive.
A worked SBA 7(a) example
Every figure in this section is a worked example. The rate and term are assumptions for illustration, not quoted terms.
A buyer agrees to pay $2.0M for a commercial services business with Adjusted EBITDA of $420K. The proposed structure:
- $1.5M SBA 7(a) loan, 10-year term, fully amortizing, at an assumed 10.5%
- $300K seller note at 6%
- $200K of buyer equity
The 7(a) loan costs $20,240 a month, or $242.9K a year.
The shortcut says $420K ÷ $242.9K = 1.73×. The buyer stops here and is pleased.
The lender builds CADS:
− maintenance capex $40K
− cash taxes $45K
− increase in working capital $15K
= CADS $320K
On the senior loan alone, $320K ÷ $242.9K = 1.32×. Still above 1.25×, with less room than the buyer thought. Now the seller note.
If the seller note amortizes over five years, it adds $69.6K a year of required payments. Debt service becomes $312.5K and coverage falls to 1.02×. That is below the SBA's 1.25× floor for an acquisition, below the 1.15× an expansion would need, and nowhere near a lender's own policy. Even on the rulebook's narrower adjusted basis, with capex and tax distributions out and working capital left in, the same structure lands at about 1.07×. The deal as drawn does not get financed.
If the seller agrees to full standby, the note has no required payments while the 7(a) loan is outstanding, so it adds nothing to the denominator and coverage stays at 1.32×. The SOP is specific about what standby means: "no payments of principal or interest for the term of the 7(a) loan".
| Measure | Cash | Debt service | Coverage |
|---|---|---|---|
| EBITDA over the senior loan (the shortcut) | $420K | $242.9K | 1.73× |
| CADS, seller note on full standby | $320K | $242.9K | 1.32× |
| CADS, seller note amortizing over five years | $320K | $312.5K | 1.02× |
| CADS after a 10% fall in EBITDA, note on standby | $278K | $242.9K | 1.14× |
The same business, the same price and the same senior loan produce four answers between 1.02× and 1.73×. The gap between the first row and the second is the definition of cash. The gap between the second and the third is one clause in the seller note.
Two more Appendix 15 rules bear on that clause. A seller note that is not on full standby and is written interest-only is underwritten as if it amortized over no more than ten years, so its payment schedule does not shrink the denominator. And the total debt supporting the purchase, including seller debt not on full standby, is capped at the business valuation.
Standby takes the note out of coverage, not out of leverage
A note on standby does not disappear. The business still owes it. Lenders read a second measure beside coverage for exactly this reason: total debt against earnings.
In the example, total debt is $1.8M against Adjusted EBITDA of $420K, or 4.29× leverage, whether or not the seller note is on standby. Coverage tells the lender whether this year's payments are safe. Leverage tells them how much debt sits on the business in total, including the part that is quiet for now. A structure can pass one and fail the other, so compute both and expect to be asked about both.
Stress the number before the lender does
The last row of the table is the one to study. A 10% fall in EBITDA takes $42K out of CADS and moves coverage from 1.32× to 1.14×, below the 1.25× floor. A coverage ratio that clears the floor only when nothing goes wrong is not much comfort to a credit committee, and under the new SOP the lender cannot lean on your projections to argue otherwise.
The useful stresses are the dull ones: earnings down, capex up, taxes up, working capital up, the interest rate up, and then all of them together. Run each against the structure you intend to propose. If the combined case falls below 1.0×, you have learned where the deal breaks while you can still change the price, the equity or the seller's terms.
You can also run the ratio backwards. At a 1.25× floor, CADS of $320K supports $256K of annual debt service. On the assumed 10-year, 10.5% terms that is about $1.58M of senior debt. That figure, not the price you would like to pay, is the ceiling on what the business can borrow.
Educational guide, not advice. This guide is general information, not investment, legal, tax, accounting, lending or valuation advice. Rules, ratios, prices and worked figures are stated as of the date shown, are drawn from public sources and our own experience, and can be out of date or wrong; they are not a substitute for a lender's, regulator's or program's current rules. Verify any figure with its primary source and a qualified professional before acting on it.
How Acquiror AI does this
The Financing module in Acquiror AI computes coverage on CADS, not EBITDA, starting from the Adjusted EBITDA in your Quality of Earnings (QoE) work. Each debt instrument carries an "Included in DSCR" flag, so a full-standby seller note stays in leverage and out of coverage. It runs the standard downside shocks to earnings, capex, taxes, working capital and the rate, plus a combined case, solves for debt capacity at a target DSCR, and labels every ratio against bands your firm sets; the defaults are 1.50×, 1.25× and 1.10×. There is an SBA 7(a) instrument template for US-dollar deals, and the structure exports as a credit memo.
The takeaway
Compute coverage on cash, not on EBITDA. Put every required payment in the denominator. Know which floor is the rule and which is your lender's policy, and that from 1 October 2026 the rule for an acquisition is 1.25× on historical cash flow. Then stress the result, because the lender will. Every term in this article is defined in the glossary.