Car Wash Feasibility Study: The Six Assumptions SBA Lenders Challenge First

The express car wash was institutional capital's favorite Main Street asset from 2020 to 2022, and it is now the asset class where SBA credit committees ask the sharpest questions. The reasons are public. The largest operator went private at a fraction of its listing price, the most aggressive consolidator went through Chapter 11, one of the largest automotive services companies wrote off most of its wash investment and exited, and new express openings fell by roughly half from the 2022 peak. None of that makes a car wash a bad credit; the SBA loan tape says the opposite. What it did was change the order of the questions a lender asks, and a feasibility study written in 2026 has to answer the first six before anyone reads the seventh.

This is the reading we give lenders and sponsors on express and full-service car wash files, current as of 7 September 2026, with the SOP 50 10 8.1 acquisition rules noted where they apply from 1 October. It is built from the public operators' filings, the International Carwash Association's consumer research, subscription benchmarking data, net lease and development cost reporting, and SBA's loan-level FOIA releases.

Why the questions changed

Three events reset the tape. Mister Car Wash listed in June 2021, briefly traded above $20, and was taken private by Leonard Green and Partners in 2026 at $7.00 a share, roughly $3.1 billion, after three years of derating on a growth story the public market stopped believing; the business itself was sound, with $1,051.7 million of 2025 revenue, 548 locations, nearly 2.3 million Unlimited Wash Club members and subscription sales at 79% of wash sales in the fourth quarter. ZIPS Car Wash, a leveraged consolidator, filed Chapter 11 in early 2025 carrying $653.9 million of funded debt and closed underperforming sites through the process. Driven Brands took an $851 million goodwill impairment on its U.S. wash segment and sold the roughly 380-site Take 5 Car Wash business to Whistle Express in April 2025 for $385 million, $255 million in cash and a $130 million seller note.

Supply followed. New express openings ran at about 943 in 2022 and roughly 550 in 2025, as all-in development cost rose to $4 million to $8 million a site against $3 million to $5 million at the start of the decade, with a prime one-acre corridor pad at $1.0 million to $2.5 million and a 100 to 150 foot tunnel equipment package at $1.0 million to $1.5 million. Car wash net lease cap rates decompressed from the mid-5% range at the 2022 peak to 6.2% to 6.4% in 2025 and 2026. And municipalities from Cape Coral, Florida to Birmingham, Alabama, Hemet, California and Perrysburg, Ohio adopted moratoriums or overlay limits on new tunnels.

The lender's conclusion from that sequence is not that the asset failed. It is that the underwriting habits of 2021, capture rates that assumed the competitor across the interchange would not exist, memberships priced to acquire rather than to retain, and ramp assumptions borrowed from a boom, are what failed. The six questions below are the credit committee's reconstruction of those habits.

1. The capture rate against the traffic count

The first number a reviewer checks is the one that drives everything else: what share of the vehicles passing the site will stop to wash. The convention is a percentage of average daily traffic on the fronting road, and the two errors that return a study are a traffic count with no source and a capture rate with no comparable.

On the count: state DOT counts are published as annual average daily traffic, and AADT is the two-way total on the segment unless the count is expressly directional. A study that applies a capture rate built on one-direction volume to a two-way count has doubled its washes before the first assumption is tested. The count must carry its source, its year, its segment and its direction, and where the fronting road is a divided arterial the study should state which direction the site actually serves.

On the capture: there is no published trade benchmark for washes per day as a share of ADT, and a reviewer knows it. The defensible capture rate is the one demonstrated by comparable sites: format, road class, visibility, access and competitive density matched to the subject, with the comparable's actual car counts stated. Practitioner ranges for a well-located express site sit in the low single digits of two-way ADT and rise with corner position, signalized access and the absence of a competing tunnel within the trade area; a study that sits above its comparables has to say why.

The fix. Cite the count by source, year and direction. Show the capture rate as a derived figure from named comparables, not an input. Run the coverage at the low end of the comparable range.

2. Membership penetration and its share of revenue

Subscription converted an episodic cash purchase into recurring revenue and is the reason institutional capital came to the sector. At the largest operator, Unlimited Wash Club sales were 79% of total wash sales in the fourth quarter of 2025, up from 75% a year earlier, with membership growing 7% to nearly 2.3 million. Across the industry, roughly 90% of members express intent to renew, and members wash two to three times a month in some cohorts. That is the annuity the lender wants to see.

The study's job is to show how the subject gets there. The questions are the ramp of members from opening, the share of washes and of revenue that members represent at stabilization, the price and tier at which they are acquired, and whether the acquisition price is a promotional rate the model assumes will step up. A study that projects a mature operator's membership share in year one, or a stabilized share without the 24 to 36 months it takes to build a base, is projecting the destination without the road.

The fix. Model members as a stock that accumulates from opening at a stated monthly acquisition pace and a stated churn, with promotional pricing and its step-up shown explicitly, and reconcile the stabilized membership share to comparable sites of the same format.

3. Price per wash and tier mix

Growth since 2023 has been earned through price rather than volume: tier increases, premium chemistry upcharges and ceramic and graphene packages carried revenue while transaction counts flattened. That makes the tier architecture a first-order assumption. Average revenue per member depends on the mix between base and premium tiers, and operators report trade-down before cancellation in a softer consumer environment, with members stepping from ceramic to base tiers while keeping the subscription.

The fix. State the price of each tier, the assumed mix at opening and at stabilization, the resulting average revenue per member and per retail wash, and a downside case in which the mix shifts one tier down. Show the comparable sites' posted pricing, with the date it was checked.

4. The competitive census and saturation

Saturation is a corridor-level fact, not a national one, and it concentrates in the Sun Belt. Florida, Texas, Arizona, Georgia and Tennessee absorbed a disproportionate share of the 2019 to 2024 construction wave, and the most active submarkets of Phoenix, Orlando, Jacksonville, San Antonio and the Dallas-Fort Worth suburbs now carry multiple competing tunnels within a single retail interchange, each underwritten on capture rates that assumed the others would not exist. The Northeast, the Upper Midwest and dense West Coast cores carry fewer washes per capita, higher land barriers and, in the salt belt, the single most reliable wash trigger in the country.

The reviewer wants a census, not a paragraph. Every professional wash within the trade area by format, age, tunnel length, pricing, membership offer and estimated volume; every site permitted, under construction or announced; and the ordinance layer, because a moratorium is adverse for a pipeline and protective for an incumbent. The census should conclude on washes per capita and per corridor mile against comparable trade areas, and it should name the competitor most likely to open in year two.

The fix. A dated competitive census with the pipeline included, a saturation conclusion against comparables, and an explicit answer to the question a credit officer asks first: how many tunnels can this interchange carry, and how many will it have.

5. The ramp to break-even

A leveraged new build needs a defined car count to service its debt, and the study has to show when it gets there. Practitioner benchmarks put break-even for a leveraged new express site at roughly 120 to 180 cars a day depending on price, mix and debt load, with 24 to 36 months to accumulate the membership base that defines maturity. Mature, automated express sites earn four-wall EBITDA margins of 35% to 50% on $1.5 million to $3.0 million of revenue; the years before maturity do not.

Under SBA rules the ramp is the binding test. Coverage must be demonstrated at 1.15x for a Standard 7(a) loan and 1.10x for a Small Loan, and the year-one and year-two ratios, not the stabilized one, decide the credit. A study that presents stabilized coverage and treats the ramp as a footnote is presenting the wrong year.

The fix. A monthly car-count build from opening through stabilization, the break-even count stated, the month it is reached, coverage by year at the applicable floor, and the sensitivity that pushes break-even past the interest reserve.

6. Churn, and the retail customer who is leaving

The last assumption is the one the 2021 studies never had to make. Subscription revenue is demonstrably resilient; the drive-up customer is not. Benchmarked non-member retail revenue fell 11.9% year over year in the second quarter of 2025 across connected sites as price pressure and softening confidence bit, and every new tunnel underwritten on drive-up capture is exposed to that line. Professional wash adoption has reached 79% of U.S. drivers against roughly 50% in the mid-1990s, and each five-year gain has been smaller than the last, which means the conversion runway that carried the sector for three decades is nearly spent.

The fix. Model retail and member revenue separately, with retail volume flat to declining and member churn stated monthly and stress-tested; show what share of stabilized revenue is contractual and what share is discretionary; and price the subscription management the site will use to intercept churn, because the gap between managed and unmanaged member bases is now a measurable operating difference.

What the special-purpose classification does to the capital stack

SOP 50 10 8 defines special-purpose property as "a limited market property with a unique physical design, special construction materials, or a layout that restricts its utility to the specific use for which it was built," and car washes sit on its non-exhaustive list beside hotels and gas stations. Three consequences follow. Under 504, the borrower contribution is 15% where the collateral is special-purpose and 20% where the business is also a start-up, and the structure on a start-up express build commonly runs 50% third-party lender, 30% debenture and 20% equity. The appraisal must be a full USPAP going-concern report from a Certified General appraiser with at least four equivalent going-concern assignments in the prior 36 months, ordered by the lender, allocating value across land, building, equipment and intangibles, with the loan capped at appraised value and value at completion on a ground-up build. And on any change of ownership numbered on or after 1 October 2026, SOP 50 10 8.1 requires 1.25x coverage on trailing EBITDA with no use of projections to meet it, caps total debt at the appraised business value, and requires a lender-ordered quality of earnings report where the business purchase price, net of the real estate, is $3 million or more.

For the study that means the six assumptions above are read against the equity they imply. A capture rate that does not survive the census raises the injection the lender will ask for; a ramp that does not clear the floor in year two is a request for a larger interest reserve; and on an acquisition, the buyer's growth plan is the risk case the QoE and valuation do not address.

What the loan tape says

The lender has the record. Tabulations of SBA's loan-level FOIA releases show about $6.83 billion of 7(a) and 504 financing to car washes since fiscal 2010 across roughly 5,800 loans. The 7(a) arc tracks the investment cycle: approvals peaked at about $676 million in fiscal 2021 and fell to about $272 million in fiscal 2025, a 60% decline, with the average loan retreating from $1.73 million to $1.12 million as lenders pulled back from large new-build requests, while the 504 program set its record in fiscal 2024 at about $163 million of debentures as borrowers migrated to real-estate-heavy, lower-leverage structures.

Credit performance is the contrarian finding. Across resolved 7(a) loans since fiscal 2010, car washes charged off at about 5.25% of loans against 7.45% for all industries, and at 0.62% of approved dollars against 1.50%. The asset class the trade press describes as feared by lenders has defaulted at roughly two thirds of the program baseline on sixteen years of data. The caveat belongs in the study: the 2021 and 2022 vintages, underwritten at peak land cost, peak multiples and peak capture optimism, are still seasoning, and the resolved-loan rate will drift up as they mature. A study that cites the headline rate without the vintage caveat will be corrected by the reviewer who has the same tape.

The pre-submission checklist

1. Traffic count with source, year, segment and direction, and a statement of whether the capture rate is applied to one-way or two-way volume.
2. Capture rate derived from named comparable sites with their actual car counts, coverage run at the low end of the range.
3. Membership modeled as an accumulating stock with stated acquisition pace, promotional pricing and step-up, churn, and stabilized share reconciled to comparables.
4. Tier prices, mix at opening and stabilization, average revenue per member and per retail wash, and a one-tier-down case.
5. A dated competitive census including the pipeline and the ordinance layer, with a saturation conclusion against comparable trade areas.
6. Monthly car-count build to break-even, the break-even count and month stated, coverage by year at 1.15x or 1.10x, and the sensitivity that exhausts the interest reserve.
7. Retail and member revenue modeled separately, retail flat to declining, churn stress-tested, contractual share of stabilized revenue stated.
8. Equity injection consistent with the special-purpose and start-up classification, appraisal allocation reconciled to the study's cash flows, and, on acquisitions from 1 October 2026, coverage on trailing EBITDA with the QoE figure carried through.
9. The SBA default record stated with its denominator and its vintage caveat.