The MCA refi cliff: what restaurants with merchant cash advance stacks do now that SBA 7(a) is closed to them

A meaningful slice of the independent restaurant operators who took merchant cash advances in 2023, 2024, and the first half of 2025 did so on a working assumption that turned out to be wrong. The assumption was that the MCA was bridge capital — expensive, but short-duration money that could later be rolled into a 7(a) SBA loan once the business cleared a qualifying revenue or time-in-business threshold. Operators with two or three MCAs stacked across funders had a real path to consolidating the lot into one ten-year SBA note. That path closed on June 1, 2025, with the effective date of SOP 50 10 8. The MCA stack is now structurally permanent until paid off.

Restaurants aren’t the largest users of MCAs as a category, but the segment has unusually high penetration for its size — bank-decline rates on independent operators stayed high through 2022–2024, and MCAs were one of the few products willing to underwrite on six months of daily card volume and a sub-650 personal credit profile. The same segment now has the narrowest set of exits from a stack it was told would be temporary.

What SOP 50 10 8 actually says about MCAs

SOP 50 10 8, effective June 1, 2025, and reinforced under Procedural Notice 5000-872050, narrowed the definition of refinanceable debt under the 7(a) program to “true loan obligations.” The carve-out is the operative language. A merchant cash advance is, in its legal structure, a purchase-and-sale agreement: the funder buys a defined percentage of future card receipts at a discount. It is not, in the SOP’s reading, a loan, and therefore cannot be the debt that a 7(a) refinance pays off.

The direct consequence: an operator carrying $180,000 in MCA balances cannot apply for a 7(a) loan with the proceeds earmarked to retire those balances. The indirect consequence is what most operators underweight. Existing MCA payments continue to count against debt-service capacity in any subsequent 7(a) underwriting. An operator carrying $180,000 in MCA balances at 14% daily holdback against $2.4 million in annual card volume is showing the underwriter roughly $336,000 a year in MCA payments — a line that sits in the debt-service stack the underwriter has to clear before approving any new 7(a) capital. Even applications for unrelated purposes — equipment, expansion, real estate — get squeezed by the MCA payment line that the same applicant could have refinanced away under the prior SOP. The first MCA an operator takes in 2026 is materially more consequential than the first MCA they took in 2024.

The math of an MCA stack you can’t refi

The arithmetic of an MCA is opaque on the contract and brutal on cash flow. Restaurant-specific factor rates currently land between 1.15 and 1.45. A factor of 1.30 on a $100,000 advance means the operator owes $130,000, repaid through daily card-receipt holdbacks. On a typical 9-to-12-month payback, that translates to a 40–80% APR — the conversion the CFPB asks lenders to compute and most MCA funders aren’t required to disclose. Shorter paybacks push the APR higher, sometimes well above 100%.

The daily holdback is where the cash-flow drag shows up. An 8% to 18% slice of every card deposit is routed to the funder before it reaches the operator’s account. For a restaurant doing $200,000 a month in card volume at 14% holdback, that’s $28,000 a month — roughly $930 a day — pulled off the top before payroll, food cost, or rent. Stack two MCAs at 14% each and the operator has sold 28% of card receipts forward to funders they cannot consolidate. Stack three and the holdback approaches or exceeds the operator’s contribution margin.

The structural problem is compounding. MCA number two was underwritten against card volume net of MCA number one’s holdback, so its payment is sized against a reduced base — meaning its implied APR runs higher than the contract states. MCA number three repeats the exercise against a twice-reduced base. The point at which the operator can no longer cover payroll without a new advance is when the stack becomes a dependence, and it arrives sooner than the headline factor rates suggest.

The four exit ramps that still exist

With 7(a) refinance off the table, four exits remain. None of them are clean.

Reverse consolidation. A new advance structured to make the operator’s existing MCA payments on their behalf — one funder paying the others’ daily holdbacks in exchange for a single, larger payment. The mechanics smooth cash flow in the short term. The economics are nearly always worse: the reverse-consolidation funder takes on the credit risk of an operator already in a stack and prices accordingly. Effective APRs frequently exceed 100%, and the new advance extends total time in the stack. It delays default, not exits it.

Term loan refinance through non-SBA lenders. A handful of non-bank lenders will underwrite a true term loan to retire MCA balances, typically priced 18–35% APR with 24–48 month amortization. The qualifying bar is meaningful — clean personal credit (usually 660+), 18-plus months in business, demonstrated revenue stability — and many operators in the deepest stacks don’t clear it. For those who do, a term loan is the closest functional substitute for the SBA refi: the rate is materially worse, but the structure converts daily holdback into monthly amortization. Lenders here underwrite against current restaurant loan eligibility criteria — bank-statement cash flow, time in business, personal credit, AR mix — rather than a single credit-score gate, but the threshold is still real.

Negotiated payoff with the original MCA funder. Some MCA funders will accept a discounted lump-sum payoff early, particularly when the operator has cash on hand and the funder reads the alternative as default risk. Discounts of 10–25% off the remaining balance are not uncommon, but the operator has to write the check. Bringing fresh capital — a private investor, refinanced equipment, a real estate cash-out — can fund the payoff, but the math has to clear the new capital’s cost as well as the discount.

Pure cash flow grind. The fourth option is to ride out the existing stack without taking on anything new. This is the right answer more often than the industry’s pitch books suggest. An MCA six months into a nine-month payback is not the problem it was at month one. For operators near the end of a stack, the discipline of refusing the next advance is the actual exit. The cost is the cash-flow tightness in the meantime, which kills restaurants that don’t price it correctly.

What changes in MCA underwriting now that 7(a) refi is off the table

MCA funders priced their books, in part, on an implicit assumption that a meaningful percentage of borrowers would refinance out via SBA. That assumption is broken, and the implications for new underwriting are starting to show.

Funders are paying closer attention to the operator’s existing stack at the point of application. A restaurant applying for a new advance while carrying two existing MCAs is now read as higher-risk than the same applicant would have been in 2023 — there’s no clean off-ramp and no third party to refinance them out. Specialist funders in the MCA funding for restaurants with thin credit history segment can still underwrite operators with sub-600 personal credit and as little as six months in business — a profile that has no other product available to it — but stack-position questions are sharper, and implied factor rates on second and third advances have moved up to reflect the absence of an SBA exit.

Some funders are also offering longer-amortization products that look more like term loans — six- to twelve-month timelines stretched to eighteen or twenty-four months, with daily holdback reduced accordingly. These structures don’t solve the underlying cost, but they reduce cash-flow drag on any single advance. Operators weighing them should compute the equivalent APR against the longer timeline, not the original product’s headline.

The third change is regulatory. The FTC and several state attorneys general — New York, California, Virginia, Utah, Georgia — have expanded MCA disclosure requirements through 2025 and 2026. Operators applying in those states now see APR disclosure on the contract itself, not only the factor rate. That doesn’t make the product cheaper, but it removes one of the asymmetries that funded the segment’s growth.

Before you take a fresh MCA in 2026

For operators weighing a new advance with one or more existing MCAs on the book, three questions clarify the decision.

First, whether the proceeds fund a specific, time-limited cash flow gap that can be repaid out of identifiable revenue — a payroll cycle before a known deposit clears, a seasonal stocking-up, a refrigeration replacement that pays for itself in food-cost savings. Advances against identifiable revenue events are the use case the product was built for. Advances to cover an operating shortfall with no identified close are the use case that builds stacks.

Second, whether the operator has computed the equivalent APR against actual payback timeline and read it next to the factor rate. A 1.30 factor at ten months is a 56% APR. A 1.45 factor at six months is a 144% APR. Restaurants comparing advances on factor rate alone systematically underprice the short-payback option.

Third, whether existing MCA payments are factored into the new application’s cash flow model. A new advance underwritten against gross card volume is being underwritten against revenue the operator does not actually have access to. The resulting holdback combines with existing holdbacks to push the total above what the operating account can sustain. The right denominator is card volume net of existing holdbacks, not gross.

The bottom line

The MCA is not the problem when it’s used surgically for the narrow case it was built for. It is the problem when it’s used as a substitute for working capital that should have come from elsewhere, or when each advance is taken on the assumption that the SBA refi will eventually clean it up. That assumption is now structurally false. Restaurant operators carrying MCAs in 2026 are carrying them to maturity, paying them off discretely, or routing through reverse-consolidation and non-SBA term-loan exits that cost more than the SBA refi did and qualify fewer applicants. The operators who avoid regret are underwriting their own use case against the four-exit reality, not against the pre-2025 menu where a fifth, cheaper exit existed at the end of the road.