A Partner Cuts SBA Buyer Failure by 44%, the Loan Data Shows

A word on scope

This analysis draws on SBA 7(a) loan records, restricted to SBA-financed acquisitions approved FY2010–2017 — a seasoned, post-recession cohort chosen so outcomes are settled rather than in flight. "Default" here means charge-offs among resolved loans. As a validity check, the overall default rate came out at 6.8%, almost exactly the SBA's own published figure. It also references Noam Wasserman's Harvard study of more than 10,000 founders.

Two limits matter. Cash and seller-financed deals are not in the SBA data, so this describes financed Main Street acquisitions only. And "partnership" is measured as the legal entity type — a proxy for buying with a partner — so genuine two-person deals that incorporated as an LLC or corporation sit in the "solo-looking" bucket, which would only understate the effect. The figures are correlational and measure business survival, not partnership longevity.

Where does "most partnerships fail" come from?

From a real study about a different thing. The warning is usually delivered with confidence — 70%, sometimes 80% — right before the advice to buy alone and keep control. The number traces to Noam Wasserman at Harvard, who studied more than 10,000 founders and found that roughly 65% of high-potential startups were derailed by co-founder conflict. That is sound research about venture-backed startups and the people who found them.

It is being repurposed to warn against buying an established, cash-flowing business with a partner. Along the way, 65% became 70 became 80, "startups" quietly became "all partnerships," and "conflict causes trouble" turned into "partnerships fail." Depending on the source, "most partnerships fail" lands anywhere from 50% to 80%, with no landmark study, no clean dataset, and no recalculated rate underneath it — conventional wisdom repeated until it hardened into fact.

What do the SBA outcomes actually show?

The opposite of the folklore. Split by who bought the business, solo owners in the cohort defaulted at 6.8% and partnerships at 3.8% — partnered buyers kept the business at nearly twice the rate of those who went alone. The pattern holds in every industry examined, without exception:

The mechanism is straightforward. Small acquisitions rarely fail from a bad idea; they fail from a bad month that becomes a bad quarter with no one to absorb it — a key manager quitting, a concentrated client leaving, an owner sidelined by illness in the same stretch. A partner is a second checkbook, a second set of hands, and a second person legally on the note. What the folklore calls a liability functions, in the data, as a shock absorber.

Where does a partner matter most?

On the smallest, most fragile deals — the opposite of the usual assumption. On acquisitions under $150,000, the range first-time buyers gravitate to because it looks affordable, solo owners defaulted at 7.6% and partnered owners at 4.5%, a three-point gap. As deals grew larger the gap nearly vanished: at $1M and above it was 2.8% solo versus 2.1% partnered, close to a rounding error.

A partner is not a luxury reserved for large, sophisticated transactions. Large deals carry their own depth — cushion, management, margin. It is the small business, where a single bad month can end everything, that benefits from a second person most, and it is precisely that business that first-time buyers most often take on alone, mistaking small for safe. Small does not mean safe; it means no room to be wrong.

How much does structure change the odds?

Enough to swamp most other decisions. Stacking the favorable choices — an established business, bought with a partner, on a long twenty-year note — produced a historical default rate of 0.75%, fewer than one in a hundred. The opposite stack — a startup, bought alone, small and cheap, on a short note — ran 10.3%, more than one in ten. Same lender, same program, same market, a roughly fourteen-fold difference in whether the buyer keeps the business.

Every input in that stack — who is on the deal, how old the business is, how large, how long the note — is knowable before signing. The failure that gets attributed to bad luck is, to a large degree, a set of choices made at the closing table: resilience or fragility, priced in from day one.

Is the partner the cause, or just a marker of a better buyer?

That objection is fair and cannot be fully dismissed. This is correlation, not a controlled experiment, and the sharpest version is that the kind of person who can attract and keep a partner is simply a more capable, more creditworthy buyer who would have succeeded regardless. The records identify that a partnership bought the business; they say nothing about the partners themselves.

Two facts constrain the objection. The advantage does not come from partners buying bigger or safer deals, because it holds inside every size band, industry, and age group tested. And it is not an artifact of a calm decade — the same gap, wider still, appears in the recession-era cohort. The legal-entity proxy also means many real two-person deals are miscounted as solo, which understates rather than inflates the effect. The causal share is uncertain; the direction is not.

What separates partnerships that work from those that fail?

Structure, not chemistry. None of this argues for grabbing the first person willing to sign — bad partnerships are real and costly. But the difference between the ones that endure and the ones that implode is usually built before closing, in three unglamorous decisions:

1. Split by skill, not ego. One partner owns operations, the other owns the money and growth. Overlapping responsibility is where conflict starts.
2. Sign the buy-sell agreement early. How a partner exits, what a share is worth, and who decides are settled on a good day, in writing, before they are contested.
3. Require real capital on both sides. A partner with nothing at risk is the liability the folklore warns about; shared exposure is what makes the second checkbook reliable.

The most valuable thing a buyer brings to a small acquisition is often not a larger down payment or a cleaner credit score. It is a second person who cannot afford to let the business fail either. The warning about partnerships is an unsourced number doing real damage; the loan record underneath it points almost exactly the other way.

Sources & method

Figures are correlational and describe SBA-financed Main Street acquisitions with settled outcomes:

• SBA 7(a) loan records — SBA-financed acquisitions only, seasoned cohort (approvals FY2010–2017), with default defined as charge-offs among resolved loans. The baseline default rate of 6.8% validates against the SBA's published figure. "Partnership" is the legal entity type, used as a proxy for buying with a partner; some tail segments have smaller samples. The figures measure business survival, not partnership longevity.
• Noam Wasserman, Harvard — study of more than 10,000 founders finding roughly 65% of high-potential startups derailed by co-founder conflict, cited to trace the origin of the "most partnerships fail" claim.