The Fleet

← All digests

Digest — The Entrepreneur, 2026-08-26

Why this crossed the escalation bar: the charter escalates when the SBLOC/SBA financing questions get answered, "since that reopens the whole candidate set." They did not get answered this run — the rulebook that governs them is being replaced, which is a larger event, and it lands on the lead candidate specifically.

No verdict changed. strong-yes is still empty. The acquisition thesis is still probably-yes. Land flipping is still investigate. The overall verdict is still not yet. Nothing was scouted.


1. There is one thing only you can do, and it takes about two minutes

Open this file and drop its Appendix 15 into inbox/:

legacy.sba.gov/sites/default/files/2026-08/SOP 50 10 8.1 effective 10.1.2026_0.docx

It is a public SBA document. It is blocked to every agent by robots.txt and opens fine in a browser. Reading Appendix 15 settles, outright, four questions this corpus has been circling for three weeks:

This is the highest-value action available in the business domain and no agent can do it.


2. SOP 50 10 8.1 takes effect 1 October 2026, and there is no deadline to race

SBA Information Notice 5000-880695, posted twelve days ago, issues a replacement for the SOP every SBA claim in this corpus was built against. Found because the standing procedural-notice check added last run was actually run — first application, and it caught a replacement of the entire governing document.

The transition turns on the SBA loan-number date. Getting one by 30 September would preserve today's rules. That is 36 days away and it is not reachable — the timeline is 6 months to 2 years of searching plus 60–90 days of financing, and there is no deal, no lender and no LOI. So 8.1 is simply the regime this operates under. Any framing that says "close before September 30" is wrong, and it is worth saying because that is the natural way to read this.

What changes for a business acquisition (all from secondary sources — six of them, published in a nine-day window, none quoting the primary text):

ChangeEffect on the thesis
Change of ownership can no longer be underwritten as a 7(a) Small Loan at any sizeEvery acquisition goes to full Standard 7(a): full credit memorandum, independent valuation, site visit. Your entire modelled band ($250K–$600K) moves.
DSCR for initial acquisitions rises to 1.25×, on historical or adjusted-historical earnings, projections barredSee §3 — this is the one that bites.
The 10% equity injection becomes non-waivable for a complete change of ownershipRemoves a flexibility that was never load-bearing here.
Equity injection moves to Appendix 15 and is described as "simplified"Puts the SBLOC-funded-injection argument back into unread. See §1.

3. Your realistic median deal probably does not clear the DSCR floor — and does not clear today's either

The corpus's financial model shows six scenarios, all comfortably above the DSCR floor. Checked against the new 1.25×, all six still pass, the weakest by ~30%.

That reading is almost certainly wrong. The model computes DSCR on unadjusted SDE — the "manager already in place" and "must hire a manager" rows carry identical ratios, because the replacement salary lands in cash flow and never touches the ratio.

Recomputed from the corpus's own published figures, treating replacement labour as an earnings adjustment:

ScenarioAdjusted DSCRvs. 1.25× (Oct)vs. 1.10× (today)
Median laundromat, you must replace the seller's labour1.05×failsfails
BizBuySell median, must hire a manager at $55K1.84×clearsclears

Three caveats, stated plainly. This is arithmetic on numbers already in the corpus, not new data. Whether a buyer-added manager's salary is actually deducted before the ratio is computed is exactly the unread question — it is the standard shape of an acquisition underwrite, but this corpus has no source for it and is not asserting it. And it changes no verdict.

But the consequence is useful, and it is the best thing this run produced. The corpus already said the search target is a business that already runs without its owner — an economic argument ($2,200/yr versus $26,200/yr on the same business at the same price). Under 8.1 that plausibly becomes an eligibility argument: a business whose owner's labour you must replace may not be financeable at all, not merely unattractive. That is a screening rule you can apply from the seller's P&L before ever signing an LOI, and it eliminates candidates earlier and more cheaply than anything else in the file.

The countervailing point: every acquisition moving to Standard 7(a) means more diligence cost, a longer timeline, and more lender judgment applied to a buyer profile the corpus already believes is an overlay problem. A model that passes the arithmetic and fails the overlay is still a decline.


4. The lender test — question 3 was obsolete before it was ever asked

Questions 1 and 2 stand (question 2 with an updated SOP reference). Question 3 is replaced — it asked how DSCR is measured for a change-of-ownership 7(a) Small Loan, a category that stops existing for acquisitions on 1 October. The new question 3:

Under SOP 50 10 8.1, an initial business acquisition carries a 1.25× DSCR floor on historical or adjusted-historical earnings, with projections unavailable to close a shortfall. Where the seller works in the business and I will have to pay someone to replace that labour, is the cost of that replacement deducted from historical earnings before the ratio is computed?

Still $0 and 2–3 hours via SBA Lender Match. Promote/demote criteria unchanged.


5. Land flipping: the verification cost is now priced, and it is worse than the model assumed

First genuinely neutral pricing data that candidate has ever had — state-filed title insurance rate schedules, which have nothing to sell.

Full verification runs 7–44% of the modelled $9,000 gross spread on a $10,000 parcel, before marketing cost. The screen-then-verify sequence designed one run ago is not economic at the bottom of the parcel band it models. Either the parcel band moves up — which cuts against the whole "too small for professionals to have bothered" argument — or verification is partial, which reinstates exactly the legal-access and boundary risk the checklist was built to retire.

Held at investigate and deliberately not demoted, because the leveraged alternative got harder in four ways in the same week and a bounded-downside fallback with no lender and no personal guarantee is worth more when the other path's regime is in flux.


6. The thing that has not moved in two weeks

Three consecutive runs have produced good financing research. None of it has touched the binding constraint.

You gave your risk-tolerance answer — willing to risk everything except the 401k and a rainy-day fund — believing the exposure was $19,500. It is not. The SBA personal guarantee is unconditional and uncapped (13 CFR 120.160, SBA Form 148 — re-verified against primary text on 2026-08-24, unamended since April 2023). On a $350K loan the real exposure is ~$350K plus interest and collection costs, enforceable at 180 days by Treasury administrative wage garnishment of up to 15% of disposable pay without a court order.

That was surfaced on 2026-08-12. Nothing found since reduces it, and no agent can answer it for you. The sequence is unchanged: re-answer the risk question → run the free lender test → then, and only then, a financial model with a named deal size.

A domain that keeps producing findings on the non-binding constraint is doing the comfortable work. Recording that here rather than in a run log, because it is the part that most needs saying out loud.


Also checked, nothing to do