Sam Rosati and Drew Eckman walked 150+ buyers through SBA SOP 50 10 8.1. Here's what changes on October 1, what doesn't, and what to do if you're under LOI.
The SBA released SOP 50 10 8.1 on Friday, August 14. It takes effect October 1, 2026. That gave the market 48 days to digest a 500-plus-page document, and the takes started flying before most people had opened it.
So we did what we usually do. Drew Eckman, our Head of SMB Loan Support and a transactional attorney who also brokers SBA loans, spent his Saturday reading the whole thing and running a redline against the old version. The following Tuesday he and Sam Rosati walked 153 buyers through it live, then came back a week later to work through the questions they didn't get to. Twenty-six questions the first time, fourteen the second.
This post is the written version of both sessions. Here's what changed, what didn't, and what Sam and Drew are telling buyers to do about it.
If you have a live deal, the only thing that matters right now is whether you get a PLP number from the SBA before October 1. Drew was direct about it: get the number and you close under the current rules. That means through underwriting and to a bank commitment before the effective date, not just an application submitted. If you're not under LOI already, assume you're playing by the new rules.
Every 7(a) change of ownership now lives in Appendix 15, and the SOP says that where anything else in the document conflicts, "the rules contained in this Appendix shall govern." When someone quotes you a general SOP provision, ask whether Appendix 15 says something different. It often does.
Appendix 15 sorts every acquisition into one of four buckets:
Sam's read: the SBA has decided first-time buyers are the risky bucket and wrote the rules accordingly. If that's you, read everything below with Initial Acquisition in mind.
The 10% minimum is calculated on total project cost, not purchase price, and for Initial Acquisitions it "cannot be reduced or eliminated." Sam's reminder from the second session: project cost is always more than purchase price. Gross it up before you model anything.
The new part is the floor underneath. At least half of the 10% has to come from unlimited sources: your own unborrowed cash, documented gifts, a personal loan to a guarantor repaid from outside the business, or a true grant. The limited sources, meaning seller debt on full standby, other standby debt, and minority investor money, can fill the other half at most, and they share one cap.
This generated the most follow-up questions of anything in either session, mostly variations on "is it half of 10%, or half of whatever I inject?" Drew settled it on the second call: the simplest way to think about it is that the guarantor has to bring 5% of project cost. If you inject 15%, your personal minimum is still 5%, not 7.5%. On a $3M project cost, that's $150,000 of your own money before a seller or investor helps with the rest.
A few things attendees asked that are worth knowing:
Sam's prediction, made on the first call and repeated on the second: this produces more partnered searches. Stanford's Search Fund Study already showed partnered searchers outperforming. Now the math points the same direction.
When investor equity is used to meet the injection, distributions to that investor beyond what covers their tax obligation are prohibited until the 7(a) loan is paid off.
Two things the SOP does not prohibit, per Drew: a preferred return accruing and catching up at payoff, and ordinary distributions on investor dollars above the injection amount. The lender may add DSC covenants to your investor documents to police the second one. Several attendees asked how you give one investor tax-only treatment on part of their money and normal treatment on the rest. Drew's answer: separate classes of equity, which is standard corporate drafting but adds cost and complexity to a deal that didn't have it before.
Sam's view is that this matters less than it looks. Model one of these deals with no exit assumption and it never looked good on distributions alone. The practical change is who is willing to hold injection capital that sits in tax-distribution-only status for a decade. Both of them expect the market to try tranched structures with different pricing on the restricted piece, and neither would guess yet at how lenders will react.
Want to buy 85% and have the seller keep 15%? That's an Owner Buyout now, and for Owner Buyouts, "individuals not currently employed by the business may only acquire less than 50% of the total equity and may not become the largest direct or indirect shareholder." The seller also has to provide a full guaranty for at least two years after final disbursement.
If it doesn't fit Owner Buyout, it defaults to Initial Acquisition, which requires a 100% purchase and says the seller "may not remain as an officer, director, stockholder, or employee."
Sam was clear this is his read and not a rule, but he thinks rollover is effectively dead for outside searchers. It is still open to employees. The GM buying 85% with the seller keeping 15% is fully compliant, needs no QofE at any price, and the lender can reduce or eliminate the injection.
That opens a path Drew flagged both weeks. "Currently employed" has no minimum tenure and no measurement date. A searcher who joins the company as an operator first can buy it later with a true seller roll. The job has to be a job, and someone on the second call asked the obvious question: what if the seller changes their mind after you've spent a year on payroll? Drew's answer was that this is exactly the risk, and it's why the path is real but not for everyone. Sam mentioned he'd heard from a searcher in the last few days who did precisely this and is now in diligence.
One related change: the seller transition period moved to 24 months, consultant only. No W-2, no board seat, no equity. In Owner Buyouts, the seller can stay indefinitely.
Any Initial Acquisition or Business Expansion with a business purchase price of $3M or more requires a QofE. The threshold is measured before equity, seller debt, or other financing, and appraised owner-occupied real estate comes out first. A $4M deal with a $1.5M building is a $2.5M business and needs no QofE. Owner Buyouts and ESOPs are exempt at any size.
The report has to include a reconciliation of financials, cash proof, and documentation of every add-back. And it "may not be prepared by or for the borrower or seller." It's the lender's report. Re-addressing letters don't cure that, and a sell-side QofE in the data room fails the same sentence. The lender "must use the earnings from the QoE" for DSC, and if the number comes in light, "the loan amount must be reduced accordingly." Your equity is the bridge.
This is where the second session spent the most time, and Sam's summary was that the cost question is annoying but the timing question is the one that kills deals.
On cost: Drew expects most banks to require an approved vendor from their list, the way they do with appraisers. He expects a number of approved vendors to emerge, and he expects lenders with expensive vendors to start losing deals to lenders with cheaper ones. He also raised the possibility that a buyer who still wants their own QofE ends up paying twice. Sam's counter: two high-quality QofE firms can look at the same business and land in different places, because add-backs are gray, and having two reports on one deal is its own problem.
On timing: the current playbook is to engage QofE the week the LOI signs, pay half up front, and if showstoppers surface mid-engagement, retrade or walk before legal spend starts. Under the new SOP the lender is the client and you're paying for an engagement you can't halt. Nothing in the text stops the lender from engaging in week one. Late ordering is a habit banks imported from appraisals. Ask your lender before you sign the LOI whether they'll engage early and whether findings can be phased.
Initial Acquisitions move from 1.15x to 1.25x, calculated on the last fiscal year-end or a two-year average at lender discretion. Business Expansions stay at 1.15x. Documented add-backs count by rule. Projections are evaluated, but lenders "may not rely on them to meet the DSC requirement."
Drew's take is that this changes less than people think, because most lenders were already underwriting to 1.25x on historicals. The mistake he sees constantly: a searcher shows him TTM numbers at 1.5x and projections that never dip, but the last full fiscal year is at 1.2x. That deal didn't work before and it doesn't work now. Sam's modeling guidance for Bootcampers is unchanged: underwrite to 1.25x on the last fiscal year, and treat anything better as cushion.
A related point Drew raised in the follow-up: the lender's business valuation has to support the purchase price. If total debt, including the seller note, exceeds the valuation, the gap has to be made up in equity. A full standby seller note may count as that equity, though Drew expects lenders to read that provision differently. A conventional seller note does not.
Under the old SOP, if real estate was 51% or more of loan proceeds, the entire loan rode a 25-year amortization. That's over. The business piece now caps at 10 years and only the real estate portion gets 25. You structure as separate 7(a) and 504 loans or one blended weighted-average note.
Drew ran the math on a $1.5M business with a $2M building: roughly a 19-year blended amortization instead of 25, and at an illustrative 9.25%, monthly debt service goes from about $29,975 to about $32,600. Call it $32K a year more on the same deal.
There's no headline rule here, which is why you haven't heard much about it. Drew thinks it has the largest underwriting impact of anything in the document, and Sam agrees. Three provisions interact:
Put those together and the conventions lenders have offered for years, "24 months standby and we won't count it against DSCR," "36 months, excluded," are going to be challenged. A $500K note at 6% interest-only shows $30K a year. Imputed at a ten-year amortization, the lender has to count about $66,600. More than double.
How Sam is modeling deals right now: every note not on life-of-loan standby is amortizing debt within ten years. If a lender offers better, great. The deal should clear DSC without it.
Separately, seller note refinancing now requires 36 months "in place and current, not on standby," up from 24. Standby time doesn't count toward the clock.
Most of the document. The 7(a) program still funds acquisitions at terms nobody else offers. The core thesis from our earlier post on the April 2025 changes holds: buyers who bring more liquidity than the minimum, document everything, and pick lenders carefully were already winning, and this SOP widens that gap.
Two things Sam asked attendees to carry out of both sessions. The SOP is a floor, and lenders can hold you to more than it requires. And two equally strong lenders will read the same paragraph differently. Lender selection matters more this quarter than it did last quarter.
Sam's closing line on the second call is the honest summary: for the next six months, we'll learn more by watching deals go through underwriting than any other way.
LIVE—Denver | October 21 to 23, 2026
Denver runs three weeks after the new rules take effect. The financial modeling, deal structuring, and SBA sessions with Drew will be on the current SOP, not the one everyone trained on. Three days with Sam Rosati, Adam Markley of PROX Capital Group, and the Deal Team. Applications close October 7.
Educational only. Not legal, tax, or financial advice. SOP references reflect SOP 50 10 8.1, effective October 1, 2026. Loans with a PLP number issued before that date are governed by SOP 50 10 8.
Matt Beckham is Operations & Growth Manager at SMBootcamp, where he runs the programs, webinars, and community that have trained 400+ searchers toward 70+ acquisitions.
