Strong frontage, fragile first year.

The project was a ground-up express exterior tunnel: a 130-foot conveyor on a 1.3-acre pad, eighteen to twenty-two vacuum stations, free-vacuum format, with an all-in development budget of approximately $5.4 million — roughly $1.0 million land, $1.6 million building and site work, $1.9 million tunnel and payment equipment, and the balance in soft costs and working capital. The site fronted a six-lane suburban arterial carrying 40,000 vehicles per day, signalized at the adjacent intersection, in a three-mile trade area that had added rooftops steadily for a decade.

The deal was structured for SBA 7(a) construction financing at approximately $4.8 million, 85 percent loan-to-cost, twenty-five-year amortization. Car wash is named special-purpose property under SBA SOP 50 10 8, which triggered the third-party feasibility requirement and the heightened documentation the SOP expects for single-use collateral.

The sponsor's pro forma assumed a 1.4 percent capture rate against the frontage traffic, producing a year-one volume that cleared the lender's coverage threshold from the first month of operation. The analytical question was whether that assumption survived contact with the way express car wash demand actually behaves.

The intuitive reading of a car wash site starts with the traffic count. More cars past the door reads as more washes, and a 40,000-vehicle arterial reads as a premium location. A capture rate applied to that count — here, 1.4 percent — produces a daily volume, and the daily volume drives the revenue line. At 40,000 vehicles per day and 1.4 percent capture, the projection produced roughly 560 cars per day, a figure that supported the requested loan at a comfortable stabilized coverage from day one.

A feasibility study that accepted the traffic-times-capture logic at face value would have endorsed the sponsor's volume, confirmed the coverage, and cleared the file. It would also have been wrong about the one thing the credit committee most needed to be right about.

Capture rate does not hold constant as traffic rises. It moves inversely. The relationship, consistent across operator data and published industry analysis, places capture near 1.7 percent at 10,000 vehicles per day, falling toward 0.8 percent at 30,000, and toward 0.5 percent at 60,000 and above. The mechanism is straightforward: high-count arterials carry through-commuters at speed, with less moment-of-decision conversion than a slower road serving local errands. A 1.4 percent capture assumption belongs on a 15,000-vehicle road, not a 40,000-vehicle one.

There is a second, deeper point. The correlation between traffic count and wash volume is weak on its own — a documented regression analysis attributes only about six percent of the variance in wash volumes to traffic count. Traffic is one input, not the forecast. The defensible volume is built by triangulation: a traffic-based capture estimate at the correct, count-adjusted rate; a population-based estimate from the registered-vehicle base in the trade area; and a comparable-site benchmark from operating washes in similar markets, validated against mobility data.

Re-rated at a count-appropriate 0.7 percent capture, the frontage traffic supported roughly 280 cars per day, not 560. The population method, run against the trade area's registered-vehicle base and a defensible per-vehicle annual wash frequency, produced a similar figure. The honest stabilized volume was approximately half the sponsor's assumption. And at the honest volume, the project did not fail — it stabilized to a coverage the lender could underwrite. It simply could not do so on day one.

Count-adjusted capture rate. The capture assumption was re-rated from 1.4 percent to a count-appropriate 0.7 percent using the documented inverse relationship between annual average daily traffic and capture. The adjustment was the single largest swing in the analysis.

Triangulated volume build. Three independent methods — traffic-capture at the corrected rate, trade-area registered-vehicle demand, and comparable-site benchmarking validated against mobility data — were run to a converged stabilized volume rather than a single traffic-derived point estimate.

Site-quality adjustments. The site earned documented upward adjustments for signalization, the going-home side of the arterial, and an above-threshold three-mile population, and a downward adjustment for the through-commuter traffic profile. Each adjustment was sourced and quantified rather than asserted.

Membership ramp, built bottom-up. The Unlimited Wash Club projection was constructed from a member count, not a percentage of revenue. The ramp followed the documented curve for new express tunnels: roughly 1,000 active members by month twelve, 2,000 by month twenty-four, stabilizing near 2,750 by month thirty, at a blended membership rate inside the prevailing twenty-five to thirty-five dollar range. Membership was modeled to reach roughly seventy to seventy-five percent of wash revenue at stabilization, consistent with the mix disclosed by the largest public express operators.

The coverage gap. The fixed debt-service line was overlaid on the climbing revenue curve. The analysis isolated the month-eighteen-to-twenty-four window — the period in which a new express tunnel's revenue has not yet caught its debt service — and quantified the working-capital reserve required to carry the project through it.

Stress scenarios. The model tested a twenty percent reduction in stabilized revenue, an elevated membership churn assumption, and a ramp extended by six months. Under the combined stress, stabilized debt-service coverage held at 1.22x, and the project cleared its obligations through the trough provided the working-capital reserve was funded at close.

The credit file presented two volume figures and explained the gap between them. The re-rated capture analysis replaced the sponsor's day-one coverage with an honest ramp: a first-year coverage below the covenant floor, a working-capital reserve sized to bridge it, and a stabilized coverage that cleared the threshold by a defensible margin. The lender's SBA specialist treated the funded reserve, not the stabilized number, as the condition that made the loan bankable. The SOP 50 10 8 expectation for special-purpose collateral — independent documentation that the specific project, at the specific scale, with the specific sponsor, can be operated profitably — was answered at the level of the year-one cash position, which is where special-purpose risk actually concentrates.

The financing closed with a funded working-capital reserve sized to the documented coverage gap. The structure converted a project that looked strong on a site memo and fragile in its first year into one the lender could carry with a defined, bounded downside. The inflection was not that the market was weak. It was that the right number to underwrite was the first-year cash position, not the stabilized return.

• 01Capture rate moves inversely with traffic count. A capture assumption that is reasonable on a low-count road is aggressive on a high-count arterial. The headline traffic figure flatters the forecast if capture is held flat.
• 02Traffic count explains only a small share of wash-volume variance. Defensible volume is triangulated from traffic, trade-area vehicle demand, and comparable-site benchmarks — not derived from a single capture-times-count calculation.
• 03For a ground-up express tunnel, the binding constraint is rarely stabilized coverage. It is the month-eighteen-to-twenty-four ramp window, where revenue has not yet caught fixed debt service. The working-capital reserve is the structural answer.
• 04Unlimited Wash Club revenue belongs in the model as a bottom-up member count and ramp curve, not a flat percentage of revenue. The shape of the ramp, not the stabilized share, determines whether the project survives its first two years.