Underwriting restaurants in 2026

Why restaurants are their own segment

A restaurant statement reads differently from any other industry. POS card processors deposit daily. Delivery platforms settle on their own cadence. Manual cash deposits persist as part of the operating mix. Payroll outflows arrive on a regular cycle and dwarf any individual deposit.

Standard underwriting metrics calibrated for retail or professional services miss this rhythm. Three POS platforms accounting for 90 percent of revenue is concentration in name only, because each platform represents thousands of underlying customers. Round-number deposits at 19 percent of total volume, which would be a strong manipulation signal in retail, is structural in food service.

The platform-mediated cash flow signature

The signature underwriters should recognize is not the amount or the timing. It is the counterparty mix. A healthy restaurant statement contains:

• Daily settlements from one or two POS processors. Names that recur every banking day.
• Periodic deposits from one or more delivery platforms. The cadence depends on the platform; weekly is typical.
• Manual cash deposits, often round-number, more frequent in older operators.
• Payroll outflows on a regular schedule.
• Supplier outflows to recognized food-service distributors.

Any deviation from this mix is a signal worth examining. A restaurant statement with no POS settlements is either a data gap or a misclassified business. A restaurant statement with platform names that do not resolve to known platforms is a candidate for stacking dressed up as platform payouts.

Where stacking concentrates

The 2021 and 2022 cohort of new restaurant openings is the fraud-risk pocket of the segment in 2026. Operators that expanded with cheap capital are carrying multiple facilities now. Stacking shows up as:

• Lender deposits described to mimic delivery-platform settlements (the most common disguise in this segment).
• Funding rounds spread across operating accounts, the catering account, and any LLC that holds a single location.
• Cash advances against future card receivables that arrive described as ordinary card settlements.

The detection signal is, again, counterparty resolution. Real platforms settle thousands of restaurants weekly. Disguised lenders settle a handful. The entity set separates the two.

Common signals that are not fraud

Restaurant underwriting in 2026 has a long list of false positives that an inexperienced underwriter should not over-weight.

• Round-number cash deposits. Manual cash mixing remains common, especially in older operators.
• Tip-share outflows that look like inter-account transfers.
• Vendor refunds that arrive as small inflows from a supplier counterparty.
• Seasonality. November and December run rates do not extrapolate.

Industry tuning calibrates each of these. The same engine applies tighter thresholds to a SaaS file and looser thresholds to a restaurant file, classified from the company domain when one is provided.

How underwriting changes

When the analysis is industry-tuned, the committee conversation changes. The 90-second re-underwrite produces an adjusted-revenue figure that already adjusts for platform mediation, manual cash mixing, and post-COVID cohort context. The underwriter spends time on the file's unusual signals, not on calibrating thresholds for the industry.

Restaurants are not the only segment with this shape. The industries page covers the other eight Vyaso supports. The pattern across all of them is the same: industry-aware detection, applied to every credit on the statement, with the calibration absorbed into the engine instead of into the underwriter's head.