Worked Example: SBA 7(a) Change of Ownership Gas Station Feasibility Study in Alaska

This is a worked example, not a client engagement. It follows an invented purchase of an operating, unbranded highway fuel station and convenience store with ten fueling positions on the Parks Highway in the Matanuska-Susitna Borough, financed with an SBA 7(a) loan. No buyer, seller, lender, address or business exists behind it. It shows a lender, a would-be buyer and an owner thinking of selling how a feasibility study tests a change of ownership, and it carries one lesson that a passing case rarely teaches: a deal can clear both coverage tests and still ask the buyer for more cash than a lender would like to see.

At a glance. A buyer would acquire an operating ten-position station with its real estate at an assumed 5.0 times the seller's last-year earnings, spend an assumed sum on the forecourt, and ask for an SBA 7(a) loan. The case passes both coverage tests. The loan, however, is held to what the seller's books carry, so the buyer's own cash is over the assumed equity cap of 35 percent.

• Total project cost: $1,509,900
• SBA 7(a) loan requested: $850,000
• Sponsor's equity share of total project cost: 43.7%
• Stabilized net operating income, base case: $238,181
• Debt service coverage on stabilized earnings, base case (lender minimum 1.25x): 2.07x
• Debt service coverage on the seller's last fiscal year, base case (program floor 1.25x): 1.25x
• Break-even volume at the binding minimum (percent of base case): 90.0%

What a Feasibility Consultant Tests First on an Alaska Station Purchase

A lender looking at a change of ownership starts from a plain question: do the earnings this business has already produced pay the debt the buyer wants to take on? The buyer's forecast comes second. For a buyer in the page's Initial Acquisition, Owner Buyout or ESOP category, the site's own SBA program page sets a floor of 1.25x coverage on historical earnings, against 1.15x for a Business Expansion ^[1]; for this acquisition that equals the lender minimum assumed here. The page asks for the last fiscal year or a two-year average to show the figure and lets no forecast stand in for it ^[2]. Its FAQ speaks of "projections unable to meet the floor except for qualifying special-purpose property" and does not say what qualifies ^[3]; this example holds the historical test. The SBA's own standard operating procedure was not read for this example; the historical-earnings rule is the site's statement ^[2], and every other program rule below rests on federal regulation.

The regulation itself leaves the number to the lender. 13 CFR 120.150 asks that a loan give reasonable assurance of repayment ^[4], and lists credit history, what the borrower earns or collects in cash, and any equity or collateral as matters a lender may weigh, with no numeric coverage minimum ^[5]. The lender minimum used here is a stated assumption, and the program floor is the site's own statement of the historical-earnings rule. They land on one number, so the debt table in Table 8 prints them on separate rows with a binding-minimum row beneath.

The two tests run side by side. The historical test takes the seller's last fiscal year, restated with owner pay replaced by a hired manager's cost and a replacement reserve taken out, and holds it against the floor. The stabilized test takes the buyer's third year, after the forecourt work and the change of hands have had their effect, and holds it against the assumed lender minimum. On the site's page the first test governs an acquisition, so here it also sizes the request: the loan the seller's books carry at the floor is $850,759, rounded down to $850,000.

The result is a pass with a catch. Base coverage is 2.07x on stabilized earnings and 1.25x on the seller's last fiscal year, with the assumed lender minimum and the floor at the same figure. Because the request was sized on the seller's year, that year clears the floor by a few hundred dollars and no more, and a reader should not take it as independent evidence of strength. The catch is the cash. The seller's year supports an $850,000 loan, the whole project costs $1,509,900, and the buyer pays the rest in cash: 43.7 percent of the cost, against the assumed cap of 35 percent that this example applies to sponsor equity. The case says so plainly, and a later section says what would break the structure.

Decision. The study would size the loan on the seller's books first, report the buyer's stabilized year beside it, and state the buyer's equity as the first finding, not as a footnote to a passing ratio.

The Market in Alaska and in the Matanuska-Susitna Borough

Alaska is a road-system market only in part. The state transportation department says Alaska has about 17,000 miles of public roads over 663,000 square miles, and that not all communities are on a contiguous road network ^[6], so a highway station draws from a thin linear catchment. Federal highway data for 2024 show 210,748 thousand gallons of gasoline and 162,148 thousand gallons of special fuels (diesel and alternative fuels) used on Alaska highways ^[7]. Diesel is a large share of the mix, so the model carries it as its own line, at an assumed 25 percent of gallons, a judgement for which no figure for a single store was found.

The project sits in the part of the state that keeps adding people. Census Vintage 2025 estimates put the Matanuska-Susitna Borough at 118,666 residents on July 1, 2025, up about 10.8 percent from the April 1, 2020 base of 107,073 ^[8]; Alaska as a whole stood at 737,270 against 733,383 ^[9], and the Anchorage Municipality at 287,155, down from 291,242 ^[10].

The state labor department reports that the Borough grew over 2 percent a year on average over the last 20 years, against about half a percent a year for Alaska ^[11], and that 51 percent of all new homes built in Alaska in 2024 were in the Borough ^[12]. For a fuel retailer the commuting figure counts for more than the head count: in 2024 only 61 percent of the Borough's 49,000 employed residents worked within the borough ^[13], so much of the morning and evening traffic travels the highway to jobs elsewhere.

Pump prices are the other half of the market picture, and they swing. AAA put the Alaska average price of regular gasoline at $4.5322 per gallon on November 11, 2026 ^[14] and the Alaska average diesel price at $6.2992 on the same date ^[15]. AAA's averages a month earlier were $5.0223 and $6.5691, and a year earlier $3.8062 and $4.0767, so late 2026 prices were well above a year before but falling ^[16]. Those retail prices include all taxes ^[17].

Decision. The study would treat the Borough as a growing, price-sensitive catchment, model fuel on margin per gallon, and refuse to read pump-price swings as earnings.

Supply, Competition and Public Comparables

The fuel margin is the first supply-side fact. OPIS data cited by NACS put the retail gross margin on gasoline at 39.7 cents per gallon in 2025, and put retailer gross margins before expenses at 38.3 cents per gallon on average over the past five years, the figure the model assumes for the sponsor and for the study alike ^[18]. By NACS's estimate, selling fuel costs a retailer roughly 20 to 25 cents per gallon, about 8 cents of it in credit card fees ^[19]. No Alaska margin was found, so the case assumes the national figure, with a pump price near the West Coast less California average ^[20] and drivers who shop on price: in a NACS consumer fuels survey, 72 percent named price as the leading consideration when picking a fuel stop and 16 percent named location ^[21].

On format, NACS reports that stores selling fuel averaged 9.8 fueling positions each ^[22]. The invented station has the whole number nearest that average, an assumption chosen so that the case does not rest on an unusually large or small forecourt.

Two listed operators give the study something to hold the model against. Neither is a party to this example.

ARKO Corp. reported a retail fuel margin of 42.8 cents per gallon in 2025, but says its fuel costs exclude a fixed margin or fee paid to its own supply segment ^{[26] [32]}, so the figure is not a clean like-for-like with an independent's. Casey's General Stores reported 42.57 cents per gallon in fiscal 2026 ^[30]. Both chains earn more per gallon than the model's assumed margin and buy fuel at a scale a single Alaska station does not, so the model leaves the margin at the five-year national figure. ARKO Corp. reported a merchandise margin of 33.7 percent in 2025 ^[27]; the model's assumed inside margin of 31.5 percent sits below that, at the centre of the 28 to 35 percent range that STAX Real Estate, a broker, gives as typical for a blended inside margin ^[33].

Costs are where this Alaska station will look unlike the chains: ARKO Corp. says labor is one of its largest site operating expenses ^[34], and the seller's books here carry an Alaska wage, an Alaska electricity price and a paid manager, which is the cautious way round for a lender testing a purchase.

The competitive set in the Borough has a chain in it. Alaska Business magazine reported in August 2024 that the company website of a Matanuska-Susitna Borough based regional retailer listed thirteen stores in Alaska and eleven gas stations, not counting new Fairbanks locations ^[35], and that the longtime owners of a Parks Highway fuel and grocery operation at Houston had sold the business the year before, splitting it into two parts ^[36]. The report carries no price. No Alaska station sale price, capitalization rate or earnings multiple was found, which is why the price here rests on a Florida broker's bands adjusted for rural sites.

Decision. The study would carry the national NACS margin and format averages as the supply anchor, mark every use of them as an assumption, hold the model's costs up against the listed operators' filings, and treat the lack of an Alaska sale comparable as a limit on the price.

Demand: Traffic, Seasonality and the Parks Highway Segment

Demand for a highway station is traffic, and Alaska traffic is seasonal. The state transportation department describes highly seasonal variation and a short count season of May through September, so summer counts must be factored down to annual averages ^[37], and it states that the counting season falls in the highest travel periods of the year ^[38].

The corridor counts found for the Parks Highway are old, and the case picks its segment from them. The department's Parks Highway Visioning Document, published in 2006, reports an average 2000 AADT of 15,100 vehicles for the segment from Knik-Goose Bay Road to Big Lake Road, about four times the 3,800 reported for the segment from Big Lake Road to Willow ^[39], and projects the busier segment at 24,800 in 2020 and 31,700 in 2030 on a recommended 2.5 percent growth rate ^[40]. Volumes ranged from 1,400 vehicles a day in rural areas to 32,400 through the urban section of Wasilla in 2000, and generally decrease from the Wasilla area north ^[41]. The station sits on the busier segment, west of Wasilla and outside any city limits; no milepost or address is part of the example.

Seasonality is steep on the quieter segments: at the Willow permanent recorder the 2000 highest monthly average daily traffic was 4,495 against an AADT of 2,667 ^[42], and the department says peak-month traffic runs 17 to 112 percent above the annual average along the corridor ^[43]. The counts are more than twenty years old, no current count was found, and the Borough has grown since, so a study would obtain the current count and the seller's pump-meter records before any lender relied on a volume.

The volume the model uses is an anchor and no more. In the NACS 2025 State of the Industry data the typical convenience store handled 322 fuel transactions daily at an average 8.4 gallons, or about 2,705 gallons a day ^[44]. NACS puts the average fuel-selling store at 9.8 positions ^[22], which gives a throughput per position, and the case carries that throughput onto the station's ten positions. No count since 2000 supports a figure above or below the national throughput, so none is used. The revenue table in Table 5 shows 2,070 gallons of gasoline and 690 of diesel a day at the assumed diesel share. The ramp table in Table 7 starts the buyer at an assumed 90 percent of stabilized volume, which is also the level of the seller's last fiscal year.

Inside the store, NACS reported U.S. convenience in-store sales of $341.2 billion in 2025 ^[45]. The model spreads the NACS figure over the 151,975 stores in the NIQ TDLinx count ^[46] and then over fueling positions, which gives the inside sales in the revenue table. On a highway the store's customers are mostly its fuel customers, so tying the store to the forecourt is a fair simplification. In 2025 fuel supplied 65.0 percent of sales dollars at a convenience store and only 38.8 percent of gross profit dollars, NACS found ^[47], which is why a fuel-heavy store still leans on its counter for margin.

Decision. The study would call gallons per day the deciding assumption, take the national throughput per position instead of inventing a site figure, and ask for a current count and the seller's meter records before any volume is relied on.

The Site: Tanks, Taxes and Power

The site is invented and described only by what the model needs: an unbranded station with ten fueling positions, a modest store with a food counter, a canopy, dispensers, tanks and piping as installed, on the Parks Highway in the unincorporated Borough. The renovation is the forecourt work alone; the store work the first concept carried is dropped.

The site's regulatory weight sits in its tank system. Alaska is not on the EPA list of states with approved underground storage tank programs ^[48], so the state is the primary implementing agency under an agreement with EPA while the federal regulations also apply ^[49]. EPA lists Alaska among the states and territories that, as of September 2025, had updated their regulations to incorporate the revised 2015 federal requirements ^[50]. The state rule text and fee schedule could not be fetched, so the tank-fee line is an assumption.

Several federal rules fall on a buyer. Under 40 CFR 280.22(b), anyone who assumes ownership of a regulated underground storage tank system must notify the implementing agency within 30 days of the acquisition ^[51]. Under 40 CFR 280.93(a)(1), petroleum marketing facilities must demonstrate at least $1 million per occurrence of financial responsibility ^[52], and under 280.93(b)(1) owners of 1 to 100 petroleum tanks must demonstrate at least $1 million in annual aggregate ^[53]. The insurance line, an assumed $48,000, is there because the coverage must meet those minimums; no premium quote was found.

Equipment age matters at a change of ownership. Under 40 CFR 280.20(f), any new dispenser system installed after April 11, 2016 must have under-dispenser containment ^[54], so a buyer would want the installation dates before closing, because replacing dispensers in the forecourt work could trigger it. Overfill prevention equipment must be inspected at least once every three years ^[55]. An environmental review of the tank system belongs in the closing costs, and the model assumes $110,000 for third-party reports, closing and financing costs, the review included.

Property tax follows the Borough. Its 2025 tax brochure computes the average bill at a mill rate of 8.856, made up of 8.485 areawide and 0.371 nonareawide ^[56], and each service area carries a separate levy set annually by the Borough Assembly ^[57]. The levies for a specific site were not sourced, so the model assumes $2,500 per $1,000,000 of assessed value for them. The assessor finds the full and true value of each property ^[58], and the case takes the real estate share of the price as that value. The Borough levies its business inventory tax only on inventory over $1,000,000 ^[59], so none is modelled. Power is a large cost in a cold climate: EIA shows Alaska's average commercial electricity price at 23.90 cents per kilowatthour in August 2026 ^[60], and the model assumes 230,000 kilowatthours a year.

Decision. The study would require a Phase I environmental review of the tank system, the dispenser installation dates and the state's current fee schedule before it relied on the forecourt budget.

The 7(a) Program Rules That Shape This Purchase

A change of ownership fits the 7(a) program because the purchase mixes real estate, equipment, going-concern value, inventory and working capital. Under 13 CFR 120.202, 7(a) proceeds can pay for an existing business outright, or for some or all of an owner's ownership stake in one ^[61]. Under 13 CFR 120.120(a), proceeds may buy existing buildings and renovate them ^[62], and 120.120(b) lets them pay for stock, supplies and operating cash as well ^[63]. The buyer runs the station, so the passive-business bar in 13 CFR 120.110(c) does not reach it ^[64].

On a 7(a) loan above $150,000 the SBA guaranty tops out at 75 percent ^[65]. The longest term is 25 years, and the real property part can run 25 years plus the construction period ^[66]; otherwise the term is ten years or less unless the loan finances real estate or equipment with a longer useful life ^[67]. The model therefore splits the loan between a share for the real estate and a share for the equipment, going-concern value, inventory and working capital. The maximum 7(a) loan is $5,000,000 ^[68], far above this request.

The rate is prime plus a spread. Daily readings of the prime rate for the week of November 2, 2026 stood at 7.00 percent throughout ^[69], and for a 7(a) loan above $350,000 the cap is 3.0 percentage points over the base rate ^[70]. This example assumes a 2.75 point spread and so a 9.75 percent note rate.

The guaranty fee is the cost most often forgotten. 13 CFR 120.220(a)(1)(iii) and (iv) cap it at 3.5 percent of the guaranteed portion on loans above $700,000, adding 0.25 percent when the loan passes $1,000,000 ^[71]; the statute applies the extra amount only to the part of the guaranteed share above $1,000,000 ^[72]. The borrower may finance the fee from proceeds ^[73]. The fiscal 2027 fee notice could not be fetched, so the model uses the ceilings. At this request the guaranteed portion is below $1,000,000, so only the 3.5 percent applies.

The coverage floor decides how the loan is sized. The site's own SBA program page puts an Initial Acquisition at its floor on past earnings ^[1] and says a forecast cannot make up the gap ^[2]. The site's own methodology page says coverage is judged against whatever threshold the program and the lender require, generally between 1.15x and 1.25x, with specific floors depending on the program and loan size ^[74]. The lender minimum is an assumption set equal to the floor, so one number binds on both tests.

Rounded to the dollar; lines may not foot by $1.

Decision. The study would size the loan on the smaller of the two coverage tests, show the fee recomputed on that loan, and keep the sourced rule and the assumed lender minimum on separate rows.

Uses and Funding: What the Station Costs and Who Pays

The cost table in Table 4 builds the project from the seller's books. The seller's last fiscal year earned EBITDA of $169,018 once restated, with owner pay normalized to what a hired manager would cost, as STAX Real Estate describes a buyer doing ^[78]. The price is a multiple of that figure. STAX Real Estate, a broker, gives market-typical multiples for stations sold with their real estate: for unbranded or independent stations, 6.0x to 8.0x in its Tier 1 Florida metros and 5.5x to 7.0x in Tier 2, bands that include the real estate ^[75]. STAX Real Estate says tertiary and rural sites generally run about half a turn under its Tier 2 bands ^[76], which makes 5.0x to 6.5x for a rural unbranded station. The case takes the low end for a price of $845,088. These are Florida bands. No Alaska multiple, capitalization rate or sale price was found, and the case treats the broker's bands as the nearest published anchor, not as an Alaska market price.

Why the low end? This example holds the request to the loan the seller's books support, because the buyer holds the books and a lender sizes an acquisition on them. At the centre of the range, 5.75x, the buyer would have paid 47.8 percent of the cost in cash; at the low end it still pays 43.7 percent. The point was fixed on the sponsor's figures and the seller's books, not on a coverage ratio, and no point in the range brings the equity under the cap.

STAX Real Estate also says a station business sold without its real estate runs roughly 1.5x to 3.0x EBITDA, plus inventory bought at closing ^[77]. The model uses the centre of that range, 2.25x, only to split the price between the business (equipment, fixtures and going concern, $380,289, financed over the shorter term) and the real estate ($464,798, financed over 25 years). Applied to an EBITDA that carries no rent, it overstates the business share and raises the loan constant, which errs on the side of caution.

On top of the price the model adds an assumed $230,000 of inventory at cost, bought on top of the multiple as STAX Real Estate prices it. It adds an assumed $175,000 of forecourt work (lighting, paving repair, card readers and a canopy refresh) with an assumed contingency of ten percent on it, an assumed $110,000 for third-party reports, closing and financing costs, and an assumed $110,000 of working capital. NACS research staff advised retailers to hold three to six months of fixed costs in reserve ^[88], which is why working capital is not left at zero. The first concept also carried assumed store work, dropped before any base case was computed. Project cost before the guaranty fee is $1,487,588, and the financed fee brings the total project cost to $1,509,900.

The request is the smallest of three loans, rounded down to $10,000: what the sponsor's own stabilized pro forma would support at the lender's minimum, what the seller's books support at the program floor, and the 7(a) maximum. The seller's books are the smallest, at $850,759, so the request is $850,000. The guaranteed portion is $637,500, and the equity at that request is $659,900, which is 43.7 percent of the total project cost, with the loan at 56.3 percent of cost.

This is the lesson of the case in one table. The buyer's own cash is $659,900 of a $1,509,900 project. It pays for the forecourt work, the inventory, the working capital and the closing costs, none of which a lender will finance against earnings the seller has already booked, because those earnings do not yet include any benefit from them. This example caps sponsor equity at an assumed 35 percent of total project cost: a sponsor is expected to bring a deal needing about a third of its cost or less in equity, and to change a concept that needs more. This concept needs more, the cap is not loosened to let it through, and the case reports the overage and leaves it standing. At the supported loan of $850,759 the recomputed fee is $22,332, the total project cost is $1,509,920 and the equity is $659,161. The four figures are quoted together so that the equity at one loan is never set against the project cost at another.

Decision. The study would state the equity at the supported loan, with the fee recomputed on that loan, as the first line of its findings, and would say that it is over the cap.

The Pro Forma: Sponsor's Figures Beside the Study's Base Case

Two statements are printed side by side so that a reader sees both views. The first is the sponsor's pro forma, the statement a buyer would hand a lender, built on market figures. The second is the study's base case, built on the best estimate for this site. A reader should see where they differ and why.

The revenue table in Table 5 builds a stabilized year from volume, margin and store sales, and the two columns are identical in it. Annual gallons are 1,007,400, or 276 gallons per position per day. Gasoline is priced at the AAA Alaska average for regular on November 11, 2026, 453.2 cents per gallon ^[14], and diesel at the AAA average of 629.9 cents ^[15]. Fuel sales come to $5,010,556, and fuel gross profit at the assumed 38.3 cents per gallon is $385,834. Inside sales are $2,290,925, and at an assumed inside gross margin of 31.5 percent the inside gross profit is $721,641. Other income, an assumed $24,000 for ATM, air, ice and propane exchange, completes the revenue. Total revenue is $7,325,480 and total gross profit is $1,131,475, of which fuel is 34.1 percent, in line with the NACS split between sales dollars and gross profit dollars ^[47].

The two columns differ in one input, the store wage. The sponsor budgets the NACS national store-level hourly wage, $15.04 ^[81]. The base case budgets an assumed $17.50, which is 25 percent over the Alaska minimum wage of $14.00 from July 1, 2026 ^[82]; that minimum is itself already 93 percent of the national store average. The model adds an assumed 15 percent of wages for taxes, workers compensation and benefits, since NACS research staff reported that store wages and benefits rose 4.2 percent in 2025, with workers compensation up 9.8 percent and health insurance and other benefits up 10.5 percent ^[84]. The hours are an assumed 44 a day, enough for two people on the floor through the open day plus the food counter, stocking and deliveries.

Store payroll, taxes and benefits are therefore $277,774 in the sponsor's column and $323,208 in the base case, and the rest of the statement is identical. Card fees come to $189,838, from NACS's report that card fees reached a record $21.3 billion ^[79] on $817.5 billion of industry sales ^[80]. The owner-manager is paid an assumed $80,000, the cost of a hired manager, to which buyers normalize owner pay, and an assumed $25,000 replacement reserve sits below EBITDA. The other lines are assumptions stated with their bases in the inputs table. A NACS research report of August 2026 found direct store operating costs up sharply across five years ^[89], so no expense line is trimmed in the hope that costs will ease.

The operating statement puts net operating income at $283,615 for the sponsor and $238,181 in the base case, which is 3.9 percent and 3.3 percent of total revenue. The sponsor's figure is higher by the wage difference and nothing else.

The ramp table in Table 7 shows the first three years. Volume is an assumed 90 percent of stabilized in year 1, 96 percent in year 2 and 100 percent in year 3. Year 1 is the volume the buyer inherits, the seller's level before the forecourt work and new ownership lift it. Net operating income runs $144,018, $200,516 and $238,181, and coverage 1.25x, 1.74x and 2.07x. The seller's books are one set of figures for the price, the cap on the request and the historical test, built at the Alaska wage, so the sponsor does not get a lower wage in the books that size the loan.

Decision. The study would show both statements together, name the wage as the single difference, and size the loan on neither of them but on the seller's books.

Debt Service Coverage Under the Convention, and Sensitivity

The debt table in Table 8 puts the $850,000 request on a note rate of 9.75 percent. The real estate share of $365,577 on 25 years has annual debt service of $39,094, and the share for the equipment, going-concern value, inventory and working capital, $484,423 on the shorter term, has annual debt service of $76,018. Total annual debt service is $115,111, a loan constant of 13.54 percent.

Against that, stabilized net operating income gives coverage of 2.46x in the sponsor's pro forma and 2.07x in the base case. The base case is above the assumed lender minimum by $94,292 a year. That is the first test. The second is the historical test. Here the program floor, as that page gives it, sits level with the lender's assumed minimum ^[1]. The seller's last fiscal year, run through the same normalized statement with owner-manager pay and reserve included, at an assumed 90 percent of base volume and the Alaska wage, has net operating income of $144,018. Against debt service of $115,111 that is a coverage of 1.25x, above the floor by $128 a year. The lender assumption and the floor being one figure, the binding minimum is that figure, and the test that sizes the loan is the seller's year: the seller's earnings carry $850,759 at the floor, the stabilized earnings would carry $1,407,015 at the lender assumption, and the smaller is the supported loan.

Both tests pass, and the case should be read at its real strength. The test on the seller's year passes by construction: the request was set at the loan that year supports, so the margin of $128 shows only that the rounding fell the right way. The stabilized test is the independent one, and it holds with room, 2.07x against the assumed minimum, but that room is the buyer's forecast, and a lender weighs history over forecast. History alone limits the loan to $850,000, and the limit is what puts the buyer's equity at 43.7 percent.

The sensitivity table in Table 9 tests five shocks against the base. A fall in revenue of a tenth gives net operating income of $144,018 and coverage of 1.25x, the same as the seller's year, because that year is by construction a volume year at nine tenths of the base. A fall of a fifth gives $49,854 and 0.43x. A rise of 2 points in the rate leaves net operating income unchanged, raises debt service to $127,959 and gives 1.86x; the request is sized at the base rate, so the shock does not move it. Operating expenses up by a tenth give $148,852 and 1.29x. A fuel margin 5 cents lower gives $187,811 and 1.63x. Only the fall of a fifth fails, and the fall of a tenth stands exactly on the line.

The break-even table in Table 10 shows how far the station can slip. Volume can fall to 90.0 percent of the base case before coverage reaches the binding minimum, and to 86.9 percent before it reaches 1.00x. At the binding minimum that is 2,484 gallons a day, with inside sales of $2,061,520. Fuel margin alone could fall to 28.9 cents per gallon at base volume before reaching the binding minimum, and to 26.1 cents before reaching 1.00x, against the assumed 38.3 cents. The store is highly geared to volume because most of its costs are fixed. The cushion in the stabilized year is real, and it is a tenth of volume wide.

Decision. The study would report that both tests pass, that the seller's year sizes the loan and clears the floor only by construction, and that the stabilized year carries a cushion of a tenth of volume.

What Would Change the Answer, and What Would Break It

The case passes, so no restructure is computed. The open point is the equity share.

What would narrow the equity. The seller's year carries $850,000, and the sponsor's 43.7 percent is the gap between that loan and the cost of the project. A seller's year that earned more than the restated $169,018 would lift the supported loan. Project extras smaller than the assumed sums, or a price below the low end of the sourced range, would shrink the cash the buyer must fund. A seller who carried part of the price would change who funds it, though this case models no such note. Operating costs nearer a chain's average site would raise the seller's EBITDA; ARKO Corp. reported site operating expenses of $685,144 thousand against segment operating income of $268,365 thousand ^[24], and a real seller's books could look either way. The case adopts none of these, because each would be chosen after the results were seen.

What would break the structure. Three things would turn a pass into a failure.

Volume. The base case has a cushion of a tenth of volume and no more. Revenue down a tenth leaves coverage at exactly the minimum, and revenue down a fifth drops it to 0.43x. The corridor counts date from 2000 ^[39], the Borough has grown since ^[8], and the NACS national store average is the only volume anchor found ^[44]. Pump-meter records that showed the seller's last fiscal year well under the assumed volume would move the historical test directly, and the loan the books support would fall with it. Since the loan is sized to that test, a worse year means a smaller loan and a higher equity share before it means a failed ratio.

The wage and the fixed costs. Labor is the largest cost line and its direction is up: the minimum wage rose to $14.00 from July 1, 2026 ^[82], and NACS reports workers compensation and benefits rising faster than wages ^[84]. A rise in all operating expenses of a tenth brings coverage to 1.29x, a hair above the minimum, and a second cost shock on top would break it.

The fuel margin. A margin 5 cents lower leaves 1.63x, and the margin would have to fall to 28.9 cents at base volume to reach the minimum, so the margin alone is not the weak point. Combined with a volume shortfall it would be, because the shocks add and the cushion in either is thin.

Decision. The study would ask the seller for three years of tax returns and pump-meter records, ask for a current traffic count, and re-run the arithmetic before anyone quoted a loan size. It would put the equity share at the head of its findings and say that the structure passes both coverage tests while asking the buyer for more cash than the assumed cap allows.

Questions to Ask a Feasibility Study Company or Consultant

What does a lender look at first on a change of ownership? The seller's last fiscal year. On the site's own SBA program page, an Initial Acquisition is tested on the last fiscal year or a two-year average, and a forecast cannot stand in for it ^[2]. A study that opens with the buyer's forecast has started in the wrong place.

Why is the loan $850,000 when the stabilized earnings would carry more? Because the stabilized earnings would carry $1,407,015 at the assumed lender minimum, and the seller's last fiscal year carries only $850,759 at the floor. The smaller sizes the request. The difference is not lost; it is cash the buyer must bring.

Is 43.7 percent equity a problem if both coverage tests pass? It is a problem for the buyer and a finding for the lender, not a coverage failure. This example caps sponsor equity at an assumed 35 percent of total project cost, and this case is over it. A real lender would set its own equity requirement.

How should a buyer choose a feasibility study company or consultant? Ask whether the consultant will run the historical and the projected tests separately, show the loan the earnings support, print every assumption with its basis, and say which numbers are unsourced. Ask what the consultant does when the loan the books carry is smaller than the loan the buyer wants. A consultant who reports only the stabilized year, or who cannot say which input drives the result, is not testing the deal. Ask also whether the consultant is independent of the seller.