Buying an accounting practice? The asset can quit.
An SBA 7(a) will finance it the same way it finances a machine shop, on cash flow rather than on collateral. What differs is what you are actually buying: a list of clients who hired a person, not a building full of machines. That one fact decides the whole structure of the deal.
Free 30 minutes. Bring the fee register and leave knowing what the coverage really is. Reviewed against the September 2026 rules.
Nothing here is bolted to the floor
A machine shop comes with a building, presses and a truck. If the business fails, a bank can sell the metal. An accounting practice comes with a client list, a lease, three laptops and a software subscription. If the clients leave, there is nothing left to sell. That sounds like the reason a bank would say no, and it is not: SBA 7(a) underwrites the cash flow, not the collateral, which is exactly why a business whose value is almost entirely goodwill gets financed at all.
So the loan is available. What changes is where the risk sits. In an equipment deal the downside is a slow year and a discount on used machines. In a practice the downside is a name on an engagement letter: the clients hired a person, and that person is leaving on closing day. Every structuring decision on the rest of this page exists to deal with that single fact.
A 12 percent client loss is a 40 percent cash flow loss
Numbers make this concrete. Take a practice billing $1,000,000 a year in fees and spending $700,000 on staff, rent, software and everything else, which leaves $300,000 of seller's discretionary earnings. Put a price of $900,000 on it. Whether that is the right multiple is a separate argument; what matters here is what happens to the money afterwards.
The equity injection runs about 10 percent of total project cost, so $90,000, and up to half of that can be a seller note on full standby, meaning the seller collects nothing while the SBA loan is outstanding. Call it $45,000 of your own cash and an $810,000 7(a) over ten years. At 9.5 percent, the planning rate this site uses until a term sheet says otherwise, the payment is $10,481 a month, or $125,774 a year. Now watch what happens to the coverage ratio in three steps.
Before you pay yourself, 2.38x. $300,000 of earnings against $125,774 of debt service. On paper that is a comfortable file, and it is the number a listing will lead with.
After you pay yourself, 1.43x. You have to live. Take a $120,000 salary out of the same $300,000 and $180,000 is left to carry the loan. That still clears the 1.25x the SBA requires for a first-time purchase on loans numbered from October 1, 2026, which most lenders wanted anyway, and it is roughly where a deal should be structured.
After twelve percent of the clients leave, 0.48x. $120,000 of fees walk during the transition year. The cost base does not walk with them, because the staff, the lease and the software renewals are already committed, so the whole $120,000 comes off earnings. That leaves $180,000, and after your salary there is $60,000 to service $125,774.
Read the last line again. A 12 percent client loss produced a 40 percent cash flow loss and put the loan under water, and 12 percent is not a catastrophe in a business built on personal relationships. It is a normal transition year. The multiplier is not a coincidence either: it is one divided by the earnings margin. At 30 percent margins, every point of fees you lose costs you three and a third points of the line that actually pays the bank.
Only one piece of the deal can still move after closing
That arithmetic is why retention protection is the deal rather than a detail in it. The problem is timing. The bank sizes and funds the loan against a price fixed at closing, and the $810,000 moves on day one whether the clients stay or go. Nothing about the loan can be reopened in month eight when the third biggest client follows the seller into retirement.
The only piece of the consideration still in the room after closing is the piece the seller is holding. So that is where the protection has to live: a seller note carrying an agreed right to reduce what is owed if named clients leave inside a defined window. Not a handshake about being reasonable. Written, with the measurement spelled out, before the letter of intent goes back.
Measure fees, not logos. Clients retained is the wrong test. A client who stays and halves the scope has half left, and a headcount test scores that as a win. Compare billed fees over the same months a year apart, client by client, and agree the comparison in writing while everyone is still friendly.
Keep the offset legal against the standby. If part of that note is counting toward your equity injection, it is on full standby and the lender will read those terms line by line. The offset language and the standby language have to be drafted together, or you win the argument with the seller and lose the injection credit. How the standby wording actually works is worth reading before you draft anything.
Then the transition itself. The seller stays through one full cycle, and in this business a cycle is a filing season, because clients meet their accountant when they need something rather than at a handover lunch. Put the introductions on a schedule with names and dates on it, and hand the lender the same schedule. A documented transition is financeable. A promise is not.
Five reads that are specific to a practice
The general work still applies, and the due diligence checklist covers it. These five are the ones a practice hides.
1. The fee register, three years side by side, by client. Not a revenue total. You are looking for concentration, and for which relationships arrived with the seller personally rather than with the firm. Those two facts price the retention risk, and neither of them appears on a profit and loss statement.
2. Realization, not billings. Fees billed and fees collected are different numbers, and the gap between them is the practice's real pricing. A firm that writes off ten percent at the invoice and another five at the payment is earning less than its rate card claims, and you are buying the collected number.
3. The shape of the year. A practice that earns six months of its fees in ten weeks does not have a profit problem, it has a working capital problem, and the loan payment is due in the quiet months too. Size the working capital line for the trough rather than the average, and do it before the loan is sized, not after.
4. Who signs. If you are not licensed yourself, someone who is has to sign the returns, and that person is both a cost and a dependency. Price them into the model before the offer, and find out whether they intend to stay once the seller has gone.
5. Where the files live. Client history sitting on an old desktop install, or in the seller's head, is a migration project you inherit during your first busy season. Ask to see the actual system in use, not a screenshot of it.
Most of this reads the same way for any practice built on recurring fees and a licence, from bookkeeping to insurance to physical therapy. The word on the door changes. The arithmetic does not.
Can you use an SBA loan to buy an accounting practice?
How much cash do you need to buy an accounting practice?
What happens if clients leave after the practice sale?
How do you record an SBA loan in the practice's books after closing?
Buying a practice, or the book of clients inside one?
Thirty minutes, free. Bring the fee register and we will tell you what the coverage really is after your own salary, what a twelve percent transition year does to it, and what the seller note has to say so that the answer is not simply your problem. New to the numbers? Start with what SDE is or what DSCR is.
Talk through your dealPrefer the self-serve rung first? the Deal Teardown is a written read on your live deal in three business days, 48-hour rush available.
No retainer required to talk. hello@clariqadvisory.com
