Sellers claim a lot of add-backs. Some of them are legitimate. Some of them are hope dressed up as accounting. And the gap between those two categories is where acquisition deals go to die.
If you're buying a business and financing it with an SBA 7(a) loan, the lender isn't going to value the business off the broker's offering memo. They're going to reconstruct cash flow from three years of tax returns, run their own version of seller's discretionary earnings (SDE), and compare it to what the seller's broker is claiming. When those numbers diverge β and they often do β the deal either reprices, restructures, or falls apart.
Here's what actually survives that process.
What is an SDE add-back, and why does it matter for SBA financing?
SDE add-backs are adjustments made to a business's reported net income to approximate what the business would earn under a typical owner-operator. The idea is that private business owners run personal expenses through the company, pay themselves above- or below-market wages, take one-time charges, and do other things that distort reported income. Add-backs strip those distortions out.
For SBA 7(a) acquisition financing, the adjusted SDE is the number the lender uses to test debt service coverage. Most lenders want a global DSCR of 1.25Γ or better β some will live with 1.15Γ on a strong deal, but 1.25Γ is the working standard. If the add-backs holding up that ratio don't survive underwriting scrutiny, the coverage fails, and the loan doesn't get approved at that loan amount.
This is why add-backs aren't just an accounting conversation. They're a valuation conversation, a loan sizing conversation, and a deal-survival conversation rolled into one.
Which add-backs do SBA underwriters actually accept?
The ones that are well-documented and non-recurring. That's the short version. Here's what that looks like in practice:
Owner's compensation above or below market. This is the most defensible add-back and the one underwriters are most comfortable with. If the current owner is paying themselves $250,000 a year and a replacement manager could be hired for $90,000, the $160,000 difference is a legitimate add-back β provided you can document market comp for that role. The reverse matters too: if the owner is paying themselves $40,000 and working 60-hour weeks, the lender will deduct a market-rate salary before calculating coverage, which can hurt the buyer.
Owner's personal expenses run through the business. Cell phones, auto expenses, insurance premiums, travel with a personal component β these are common and generally accepted if they're on the return and can be identified. The lender wants to see them on Schedule C or the corporate return, not just on a list the broker emailed over.
Depreciation and amortization. Standard. No controversy. These are non-cash charges and every lender adds them back. What does get scrutinized is whether the business has a meaningful capex requirement that should be netted against that add-back β a business with aging equipment may show high D&A and also need real cash for replacement.
One-time, non-recurring expenses. A lawsuit settlement. A one-time equipment repair. A flood loss in year two. These are legitimate if you can prove they were genuinely non-recurring. The lender will look at whether something similar showed up in an adjacent year. If legal fees appear on every return, that's not a one-time item β that's a chronic cost of operating this business.
Owner's personal health insurance and retirement contributions. Typically accepted. These run through the business for tax purposes but wouldn't apply to a new owner in the same way, and they're well-recognized by SBA underwriters.
Which add-backs do underwriters reject β or heavily discount?
This is where broker memos and lender overlays diverge most sharply.
"Adjusted EBITDA" that doesn't tie to the tax return. A lot of broker packages will show an "adjusted" number built from a mix of bank statements, internal P&Ls, and the owner's verbal explanation of anomalies. Underwriters work from the tax return, full stop. If the adjusted figure isn't reconcilable to a signed return, it doesn't exist in the lender's model.
Owner's salary replaced with an unrealistically low management cost. This is the most common miss I see. A seller doing everything β sales, operations, client relationships, delivery β claims a buyer can hire a replacement manager for $50,000. That may be true in narrow cases. But if the business is owner-dependent and the replacement cost is clearly understated, the underwriter will substitute their own figure. The gap comes directly out of the SDE.
Revenue from contracts that haven't transferred. Not an add-back issue exactly, but it functions like one in the SDE analysis. If a meaningful portion of revenue depends on a personal relationship with the seller, or on a contract that requires client consent to assign, the lender may haircut projected income. This is especially common in professional services and government contracting.
One-time revenue bumps claimed as normalized earnings. A seller who had a $200,000 equipment sale in year two, or won an unusually large contract that didn't repeat, may present the average of three years as "normalized." Underwriters spot this fast. They'll strip the anomalous year or weight it lower.
Discretionary expenses the seller describes but can't document. "We ran about $30,000 of personal expenses through the business" is not an add-back. It needs to appear somewhere in the financials, traceable to a line item. Verbal add-backs don't exist.
How do lenders treat add-backs differently across SBA programs?
The standard SBA 7(a) underwrite β whether done by a Preferred Lender (PLP) or a non-PLP going through the SBA directly β follows SOP 50 10 8 guidelines on cash flow analysis. The SBA doesn't prescribe a specific add-back list; it gives lenders discretion to use a "reasonable method" for calculating repayment ability. That discretion is where lender overlays come in.
Some lenders are conservative and will only accept add-backs that appear on the return and are supported by a third-party document. Others are more willing to work through a well-built memo from a CPA or broker. SBA Express loans β which have a lower guarantee percentage and faster turnaround β tend to get less hands-on cash flow reconstruction, which can cut both ways.
The SBA 504 program, used mostly for real estate and heavy equipment, splits the financing between a bank first mortgage and a CDC/SBA debenture. Both the bank and the CDC will underwrite cash flow, and they don't always reach the same conclusion. On an acquisition using 504 debt, you may end up defending your SDE analysis to two different underwriting teams.
What should buyers and their advisors do before the deal gets to a lender?
Build the SDE model from the tax returns first, not from the broker memo. Use Schedule C, Form 1120-S, or Form 1065 K-1s β wherever the income actually lives β and work from there. Every add-back should have a corresponding line item or a document you'd be comfortable putting in a lender's credit file.
Then ask: if the lender rejects the two largest add-backs, does the deal still work? Run the coverage ratio at the lower number. If DSCR falls below 1.15Γ in that scenario, you're either overpaying for the business or you need a deal structure that reduces debt service β seller financing, a longer amortization, a lower acquisition price.
CPAs who do this work regularly know which add-backs lenders accept and which ones they won't touch. Engaging one early, before the letter of intent is signed, saves everyone time. The seller finds out their broker memo was optimistic. The buyer finds out what they're actually financing. And the lender gets a file that's been stress-tested before it hits their desk.
That's not a perfect process. But it's a lot cleaner than building a deal around SDE adjustments that don't survive the first phone call with underwriting.