Ramp-Up and Absorption: How Long a New Project Takes to Stabilise, and How a Study Proves It

A new project does not open at its stabilised numbers. A new hotel has to win its first guests, a storage facility rents its first units one at a time, a car wash has to build a customer base, and a new apartment building leases against the buildings that delivered before it. The months between opening and stabilisation are when a loan is most exposed, because debt service starts on its schedule and revenue does not. Anyone searching for the ramp-up period and absorption schedule for a new project is really asking two questions at once: how long the climb will take, and what evidence would persuade a lender that the answer is more than the sponsor's hope.

This article explains how a feasibility study builds that schedule and carries it into the loan. It separates the ramp of an operating business from the lease-up of a building, sets out what the SBA, USDA and bank rules say about the months before stabilisation, ranks the evidence a study can lean on, reads the Census Bureau's apartment absorption survey in detail, and works a quarter-by-quarter model that sizes the reserve a ramp needs. The occupancy a lender will accept once a project has stabilised is a different question, discussed in the archived post Stabilized occupancy benchmarks for SBA and USDA loan underwriting. This article is about the road to that level and the time the road takes.

Two kinds of climb: an operating business and a building

Two different things travel under the name of a ramp, and a study should say which one it is modelling. The first is the ramp of an operating business. A fitness centre, a car wash, an assisted living community, a hotel or a glamping resort sells a service, and its revenue rises as it gathers customers, residents or guests and as its prices settle. The second is the lease-up of a building, in which units or space are offered to tenants and taken off the market one lease at a time. Many projects carry both. An assisted living community leases apartments and also sells care; a hotel sells room nights and may run a restaurant beside them. The distinction matters because the evidence for each is different, and so is the way each one fails.

For a building, the measure has a formal definition. The Census Bureau's Survey of Market Absorption (SOMA) treats a unit as absorbed once a unit in a building of five or more units has been rented or sold for the first time after construction and is no longer on the market ^[1]. Its rates reflect the first rental of an apartment, or the first sale of a condominium or cooperative unit, after completion ^[2]. Absorption is therefore a count of first leases, not a measure of occupancy. A unit let in the second month and vacated in the ninth stays absorbed in the survey's terms while the owner's rent roll shows it empty. A study that borrows an absorption rate has to convert it into the occupancy path its operating statement needs, and it should show how it did that.

An operating business ramps in its own units: occupied room nights, enrolled members, residents in place, vehicles washed, covers served. The curve has to be built from the operator's own history and from dated comparables, which is harder evidence to gather and easier to overstate. Federal rules recognise the idea even where they do not measure it. In the USDA rule for Business and Industry guarantees, the definition of a new business includes, besides a business younger than one full year, an older one that, in the Administrator's judgement, has not yet attained "full operational capacity or stable operations" ^[3]. Under that definition a business stays new until it has finished its ramp, whatever its age.

The two climbs also go wrong in different ways. A building that leases slowly usually ends with a tenant base, and the question is how long the owner must carry the gap. A business that ramps slowly can settle at a lower level than the loan was sized on, because a customer base that forms below expectation may simply stay there. A study should say which risk it is testing, and its stress cases should differ to match: a lease-up test stretches the calendar, while a business ramp test should also lower the level at which the curve flattens.

Where the clock starts and what counts as stabilised

Every ramp schedule rests on two dates, and both are easy to blur. The first is the start of the clock. For a business it is usually opening day, but the debt may have started long before: a construction loan draws for months before the first customer arrives. SOMA starts its clock at completion, and it treats a multifamily building as complete when 50 percent of its units are occupied or available for occupancy ^[4]. Its three-month rate therefore describes the quarter after that point, not the quarter after the first unit was handed over. A study that sets its own schedule beside a published rate should line the two clocks up before it compares anything.

The second date is stabilisation. A study should define it in measurable terms before it models it, for example as the first four consecutive quarters in which occupancy holds within a stated margin of the stabilised level and revenue per occupied unit holds at the stabilised rate. A definition written after the model has run tends to drift toward whatever the model produced. The site's own methodology page describes revenue as "phased over the absorption schedule the demand analysis supported, so that the revenue line reflects a realistic lease-up rather than instant stabilization" ^[5]. A fixed end point is what keeps that phasing honest, because without one a slow ramp can always be relabelled as a long one.

Program rules turn the start of the clock from a drafting choice into a credit question. The site's SBA program page sets a 1.15x floor of business coverage for a standard 7(a) loan, "and startups must reach it within two years of funding" ^[6]. Read literally, a window that runs from funding includes any construction period, so a project that takes a year to build has only a year of operations inside it, and the model later in this article makes that arithmetic explicit. For 504 loans the site's SBA program page gives the same 1.15x, measured "on history or two-year projections" ^[7]. Neither statement appears in the text of 13 CFR 120, from which this series takes SBA's rules, and the regulation's own references to a young business, discussed in the next section, concern credit elsewhere and the 504 contribution rather than coverage.

Conventions. This article follows the conventions of the Loan Analytics worked examples. Coverage is tested against the program floor the site's program page states, where the program has one, and against a stated 1.25x lender assumption, and the higher of the two binds. Equity is measured at the loan the cash flow supports, with fees recomputed on that loan. An acquisition is tested on the seller's historical earnings, not on projections.

What the program rules say about the months before stabilisation

13 CFR 120 does not write a ramp schedule, but several of its provisions decide how a ramp is judged and paid for. SBA's lending criteria start from a creditworthy applicant, and the rule's next sentence reads "Loans must be so sound as to reasonably assure repayment" ^[8]. Among the criteria a lender, a CDC or SBA may weigh, as applicable, are credit history and the applicant's earnings or cash flow ^[9]. For a project with no history, those earnings are projected, and the ramp is the stretch of the projection that carries the most doubt. The credit elsewhere test names the young business directly: one of the conventional lending factors the lender or CDC takes into account is whether the applicant has been in operation two years or less ^[10], and another is how long a term the loan needs if the business's actual or projected cash flow is to repay it ^[11]. On the 504 side, a borrower with two years or less of operation brings at least 15 percent of the project cost, administrative costs excluded ^[12]. In those provisions the regulation treats youth as a risk, and none of them sets a coverage level for a start-up.

The term rules give a build some room but not a ramp. A 7(a) loan runs for the shortest appropriate term, depending on the borrower's ability to repay ^[13]. Where the loan pays for equipment or leasehold improvements, its term may be lengthened by a reasonable period of up to 12 months if that time is needed to finish installing the equipment or completing the improvements ^[14]. For the real property share, the regulation allows a term of "25 years plus an additional period needed to complete the construction or improvements" ^[15]. Those extensions cover the time to finish the asset. Nothing in them postpones debt service until the business has stabilised. What the regulation does provide is a way to fund the gap: 7(a) proceeds may also pay for working capital, as well as inventory, supplies and raw materials ^[16], which is the program's route for carrying an operating shortfall inside the loan.

The 504 structure handles the build through interim financing. Interim financing, from a source that meets the conditions in 120.890, may fund every cost of a 504 project apart from the borrower's own contribution ^[17]. The regulation lists repayment of that interim loan, "including points, fees and interest", among the costs a 504 loan may pay ^[18]. The debenture then waits on two certifications: the interim lender reports what it advanced, and the CDC confirms the work was finished to the final plans and specifications, except where 120.961 provides otherwise ^[19]. Interest during construction can therefore become a project cost. Those provisions say nothing about the operating ramp after completion, so a 504 study has to show where the post-opening shortfall will be funded.

The USDA rule is the most explicit of the programs about the ramp itself. Among the uses B&I loan funds may cover is "Interest (including interest on interim financing) during the period before the first principal payment becomes due or when the facility becomes income producing, whichever is earlier" ^[20]. Another is "Startup costs, working capital, inventory, and supplies in the form of a permanent working capital term loan" ^[21]. Under the same rule a note may carry an interest-only period, but interest still has to be paid at least once a year, beginning no later than a year after the note is dated ^[22]. The rule then ties the payment schedule to the ramp in plain words: the timing of the first principal and interest payment takes account of "whether the facility is operational and generates adequate income" ^[23], and "the scheduling of the first full principal and interest payment must commence not more than 3 years from the date of the promissory note" ^[24]. A ramp that runs past that outer limit has to be carried by something other than the payment schedule.

The USDA rule also asks for the schedule in a particular form. When a lender asks for the guarantee to be issued while the project is still under construction, its credit evaluation has to carry a quarterly operational cash flow analysis running from the current financial statements "through start-up or occupancy" ^[25]. The same request needs evidence of cash flow sufficient to complete the build, including contingencies for cost overruns, and also "working capital during the business start-up period" ^[26]. A quarterly projection is also part of the evaluation for borrowers with seasonal cyclical cash flow ^[27], which can apply to a resort, a campground or a marina whose revenue follows the season. Where projections depart from what the business has done before, the rule requires them to be substantiated and documented ^[28]. It also tells the lender to make all efforts to see that the borrower holds adequate working capital, operating capital and reserves for capital expenditures, debt service and maintenance, as applicable ^[29]. The study that answers these provisions is, in the rule's definition, a report by independent qualified consultants, and the rule names what it evaluates: "the economic, market, technical, financial, and management feasibility of the proposed project or operation" ^[30]. When the Agency may insist on one is set out in the series article USDA B&I Guarantees Under 7 CFR 5001: What the Agency Tests and How a Study Answers.

For a bank loan without a guarantee, the rules speak to the lender's policy rather than to the project. The interagency real estate lending guidelines give a list of market supply and demand factors that an institution should weigh as it watches its lending area ^[31]. One item on that list reads "Current and projected vacancy, construction, and absorption rates" ^[32]; the next reads "Current and projected lease terms, rental rates, and sales prices, including concessions" ^[33]. For construction and development projects, and for completed commercial property, the guidelines ask a lending policy for further standards scaled to the project's size and type ^[34]. Three of them bear on a ramp: limits on interest reserves and when one is acceptable ^[35], pre-leasing and pre-sale requirements for income-producing property ^[36], and requirements for feasibility studies and for testing how sensitive projected income is to changes in rates, vacancy and expenses ^[37]. The 2006 concentration guidance issued by the three agencies adds that disbursement controls on development and construction loans should include tracking pre-sold units and pre-leasing activity ^[38]. Under the FDIC's appraisal rule, an appraisal for a federally related transaction must analyse and report appropriate deductions and discounts for, among other things, partially leased buildings ^[39]. The appraisal accounts for the lease-up in its value; the study tests whether the cash flow survives the lease-up. The line between the two documents is drawn in A Feasibility Study Is Not an Appraisal: What Each Document Answers, and What Each Cannot.

The evidence for the curve, ranked by weight

A ramp schedule is a forecast, and the evidence behind it falls into a clear order of weight. A study should use the strongest evidence available, say which tier each ramp assumption came from, and treat the sponsor's plan as the hypothesis under test rather than as a source.

The project's own record comes first where one exists. An expansion, a second location or the purchase of an operation running below its potential has real months to read: occupancy by month, revenue per occupied unit, the rate at which inquiries turned into customers. The assisted living case, Worked Example: SBA 7(a) Assisted Living Feasibility Study in Utah, is a purchase of an operating facility below its stabilised occupancy, which is why its ramp begins at the occupancy the buyer inherits rather than at zero, and why its decisive coverage test runs on the seller's history rather than on the climb. Under the USDA rule a projection that departs from that history must be substantiated and documented ^[28], and the same discipline is sound whatever the program.

Dated openings of comparable projects come next, and for a new project they should carry the schedule. The test of a comparable here is not only that it competes for the same customers but that its opening date is known, so that its occupancy can be read at the same months after opening the subject's schedule uses. For each comparable a study should record the opening date, the capacity, occupancy at three, six, twelve and twenty-four months where the operator or a public record discloses it, any concessions or introductory rates, and what else opened nearby in the same period. The market factors in the interagency guidelines, which include current and projected lease terms, rental rates and sales prices with concessions ^[33], make a fair checklist for what a credit officer will expect to see. How the comparable set is chosen belongs to the series article on Supply and Comparable Analysis: Counting Competitors, the Pipeline and the Properties a Study Leans On; this article takes the set as given and asks what its dated history proves.

Published series rank third. They are broad, independent and dated, but they describe a population, not the subject, and they arrive with a lag. For apartments a free federal series exists, and the next section reads it closely. SOMA counts only units in new residential buildings of five or more units ^[40], so for a hotel, a storage facility, a senior living community or an operating business the curve has to come from the tiers above.

The operator's record at its other locations ranks fourth, below independent evidence, because an operator decides which openings to share. It earns weight where the operator runs the same format in similar markets and discloses every opening, the slow ones included. The sponsor's projection comes last. It sets the hypothesis the study tests and nothing more, and a ramp that appears in the sponsor's plan and in no comparable is one a study should decline to underwrite.

Reading the Census absorption survey: what it shows and where it stops

SOMA is a federal series that measures lease-up directly. It is sponsored by HUD, draws its sample from the Census Bureau's Survey of Construction, and each month selects residential buildings of five or more units ^[40]. Its first interview, three months after completion, records rents or prices, the number of units, the building type and the units absorbed, with later interviews at 6, 9 and 12 months where needed ^[41]. The estimates are limited to units that are privately financed and nonsubsidized, in buildings with at least five units ^[42]. Time-share units, continuing-care retirement units and turnkey public housing units fall outside its scope ^[43], and buildings completed in areas that issue no permits are left out, which the Census Bureau puts at less than two percent of new construction ^[44]. Once a year it estimates, for a calendar year of completions, the share absorbed after three, six, nine and twelve months ^[45].

That annual curve is the shape a ramp schedule is trying to capture. Of the unfurnished rental apartments completed in 2016, the annual report counts 55 percent rented within three months of completion and 76 percent after six ^[46], then 88 percent after nine months and 94 percent after twelve ^[47]. Table 1 shows the same curve by region and by location inside or outside principal cities.

Three features of that table bear on a schedule. The curve is steep early and flat late: more than half of a year's leasing happens in the first quarter after completion, and the last few points take as long to arrive as the first half did. The regions differ most at three months and converge by twelve, so a regional figure says more about the speed of the climb than about where it ends. And the survey stops at a year, so it is silent about a second year of lease-up, which is exactly where the model below still finds a shortfall.

The second lesson is that the speed of absorption moves with the market, and moves a long way. The Census Bureau's source and accuracy statement for 2024 completions prints the three-month absorption rate for unfurnished apartments for every completion quarter from 2016 onward, each figure revised in later quarterly releases and fixed after the fourth ^[55]. Table 2 takes seven of those quarters.

Apartments completed in the third quarter of 2021 were 75 percent absorbed within three months ^[58], while those completed in the fourth quarter of 2023 were 41 percent absorbed ^[60]. The two delivery dates sit about two years apart, and the speed of leasing nearly halved between them. A first-year assumption carried over from a fast vintage into a project that will deliver into a slow one can be wrong by that margin, and the series makes the size of the error visible. Revision matters too. In each of the seven quarters shown the final estimate sits below the initial one: the second quarter of 2024 moved from 55 to 50 percent ^[61], the second quarter of 2022 from 77 to 72 percent ^[59]. Revision does not always run downward; the fourth quarter of 2019 rose from 54 to 55 percent ^[63]. A study that cites a recent quarter should say whether its figure is initial or final.

The limits matter as much as the numbers. The series is national and regional, not a submarket count, and a project's own competitive set can lease far faster or slower than its region. It measures first rentals, not occupancy, as its definition of absorption makes plain ^[1]. It is a sample survey, and its rates are relatively more reliable than the counts behind them, particularly when the rates are 50 percent or more ^[64]. Since second quarter 2018 completions the Census Bureau has published a margin of error for every estimate in its Table Creator ^[65], and a study that quotes a rate should quote that margin beside it. The series is best used as an outside check on the shape and speed a study derives from its comparables, not as a substitute for them.

The multifamily case, Worked Example: Conventional Bank Construction-to-Permanent Multifamily Feasibility Study in Nebraska, shows where the national series could earn its place. It is set in a suburban submarket for which the example could not source a publishable count of units leased, and a study in that position can still set its assumed first-year occupancy beside the regional curve and the leasing speed of the delivery quarter, rather than leave the assumption standing alone.

Building the schedule: a quarter-by-quarter ramp and the carry it needs

The model in this article is an illustration with assumed inputs, not a project. It describes a new facility with 120 rentable units, which a reader can take as hotel rooms, senior apartments, storage units or campsites, opening after a twelve-month build financed with a loan at the 7(a) maximum. Every input is in Table 3, labelled as an assumption or tied to its source.

Rounded as shown; lines may not foot by one unit in the last place.

The curve follows one rule: each quarter, the project closes a fixed share of the gap between its current occupancy and its stabilised occupancy. In the base case it averages 35.0% occupancy in its first quarter open and closes 30% of the remaining gap each quarter on its way to 90.0%. That shape, quick at first and slow at the end, is the one the absorption survey shows for apartments in Table 1, and it gives a study a practical advantage: a single parameter sets the speed, so a stress case can slow the ramp without quietly changing anything else. A study fits the parameter to the dated comparables, checks that the opening occupancy matches what the comparables did in their own first quarter, and reports both.

Table 4 runs the schedule for the first eight quarters after opening.

Three features of the schedule carry the argument. First, the opening quarter loses money before a dollar of debt is paid. Fixed costs arrive in full on opening day while revenue arrives with occupancy, so net operating income is negative and coverage in that quarter is below zero. Second, the shortfall shrinks quickly and then ends: from the sixth quarter, net operating income exceeds the quarterly debt service, and the cumulative shortfall stops at $414,754. Third, even at the end of the second year the project has not reached its stabilised occupancy. It stands at 85.5% against 90.0%, and coverage in the eighth quarter is 1.15x. In this illustration the project first reaches the coverage the site's SBA program page states for startups ^[6] about two years after opening, which is three years after funding.

The carry a lender should see funded is the sum of two pieces. Interest during the twelve-month build, with half the loan drawn on average, is $237,500. The operating shortfall across the eight quarters after opening is $414,754. Together they come to $652,254, which is 14.9 months of amortising debt service. A study should present the two pieces separately, because each program funds them, if at all, through a different provision.

Coverage during the ramp against coverage at stabilisation

A lender reads two kinds of coverage for a new project, and they answer different questions. Coverage at stabilisation asks whether the project, once it has finished its climb, earns enough to carry the loan with a margin. Coverage during the ramp asks whether the borrower can survive the climb at all. Under the conventions stated earlier, the first is tested against the program floor and the 1.25x lender assumption, whichever is higher. The second has no floor of its own in this method; it is carried by reserves, by an interest-only period where the program permits one, or by the sponsor's own money.

Table 5 runs both tests under seven changes to the ramp and its financing.

The base case passes at stabilisation: its 1.29x clears the binding minimum, which is the 1.25x lender assumption, and stabilised income would support a loan of $5,155,107, more than the loan in the model. On a stabilised view alone the loan looks sound. The ramp tells a different story. Coverage is 0.24x in the first year after opening and 1.04x in the second, so the project misses the 1.15x the site's SBA program page states for startups ^[6] in both of its first two years, and inside two years from funding it has only 12 months of operation at all.

The shock columns show what the reserve depends on. A slower ramp, closing a fifth of the gap each quarter, leaves second-year coverage at 0.79x and raises the carry to $841,032. A weak opening, fifteen points of occupancy below the base, costs nearly as much, at $825,719, even though it ends the second year close to the base case. A project that stabilises five points lower fails at stabilisation itself, at 1.14x, and its stabilised income supports only $4,550,779: no reserve cures that case, because the loan is too large for the business. A two-point rise in the rate fails the same way, at 1.11x, and also lifts the carry to $840,210. A build six months longer raises construction interest to $356,250 and leaves only 6 months of operation inside two years from funding. Interest only in the first year cuts the carry to $603,036 and changes nothing at stabilisation. The faster ramp is the only column in which second-year coverage clears 1.15x, at 1.21x.

The site's own methodology page names among its stresses "a slower lease-up that delays stabilization, a softer achievable rent if competing supply arrives, and a higher vacancy than the base case assumed" ^[69]. The table adds the ones a ramp brings with it: the opening level, the length of the build and the shape of the payment schedule. Two conclusions follow for a credit officer. A stabilised coverage figure, however comfortable, says nothing about the first two years, so a file that prints only that figure has not tested the ramp. And the reserve should be sized on a stressed ramp rather than the base case, since every adverse column in the table needs more carry than the base.

Reserves that carry the ramp, program by program

The carry has two parts, interest before opening and operating shortfall after it, and each program reaches them through different provisions. A study should name the provision it relies on for each part, or say that the part falls to the borrower.

For a 7(a) loan, proceeds may be used for working capital ^[16], so a study can show the post-opening shortfall financed as a working capital portion of the loan, or funded by the borrower, and can recommend how it should be held and released. On the real property portion, the term may run past 25 years by the time needed to finish building ^[15], so the build need not shorten the repayment period, and the study should state how it assumes interest is paid while construction lasts. Neither provision removes the need to fund the shortfall; they decide only whether the loan can be the source.

For a 504 project, interest on interim financing during the build can be repaid as a project cost ^[18], but that provision ends where the interim loan ends. The post-opening shortfall is not interim financing, and a 504 study should show it carried by the borrower's working capital or by another identified source.

For a B&I guarantee, the interest clause stops at whichever of two events comes first: the first principal payment falling due, or the facility starting to produce income ^[20]. After that point, interest is no longer a loan use under that clause, so the study should show how the shortfall after opening is carried: a permanent working capital term loan of the kind the rule allows for startup costs and working capital ^[21], equity, or a lower payment during an interest-only period, with interest still paid at least annually from no later than a year after the note ^[22]. The glamping case, Worked Example: USDA Business and Industry Loan Glamping Resort Feasibility Study in Ohio, works this through: its first two years fall short even of interest-only debt service, and the example treats the gap as a funded reserve held outside the loan rather than as a use of loan funds.

For a bank loan, the guidelines leave the acceptability of an interest reserve, and its limits, to the bank's own policy ^[35]. The study's part is to size the reserve against the stressed schedule and to state how many months of interest it carries, because an interest reserve that runs out before the ramp is over converts a slow lease-up into a missed payment. Whatever the program, the reserve should appear in the study as a number, with its source, the stress case it was sized on and the quarter in which it would run out. A reserve sized on the base case is only the base case restated.

Approvals that never fund: the cancellation record as a hint

The loan file adds a caution of its own. An SBA approval is not a funded loan, and in every cohort a share of approvals is cancelled before any money is disbursed. How that ending differs from a charge-off is the subject of Cancellation Is Not Charge-Off: The Two Ways an SBA Approval Can End Without Repayment; here the cancellation record serves only as a reminder that a project can fail before its ramp begins.

In the FY2017 to FY2024 approval cohort, 11.24 percent ^[70] of 7(a) approvals ended cancelled, and 13.60 percent ^[71] of 504 approvals. Both are approval-cohort floors, because approvals still waiting to disburse may yet cancel; the ceilings that count every waiting approval as cancelled are 12.24 percent ^[72] and 16.90 percent ^[73]. For four asset classes whose new projects have to fill rooms, units, beds or sites after opening, the cohort floors are 13.13 percent ^[74] for 504 hotels and motels, 19.15 percent ^[75] for 504 self storage, 10.21 percent ^[76] for 7(a) assisted living without nursing care and 13.50 percent ^[77] for 7(a) RV parks and campgrounds. Each rate covers every approval in its class, whatever the project, purchases of operating businesses included. The file records that each cancelled approval ended without disbursement. It does not record why, and nothing in it links a cancellation to a ramp assumption, so these rates cannot be read as a measure of ramp risk.

The 504 file holds a second figure that matters to a project that has to be built. Because the debenture waits on the two certifications described earlier ^[19], a 504 approval for a project under construction can stay undisbursed in the file until the building is finished. The FY2024 504 cohort still had 17.30 percent ^[78] of its approvals undisbursed at the as-of date, and the file does not say which of them are waiting on construction and which will end cancelled. Charge-off is the other ending. Within five years of approval, 3.72 percent ^[79] of disbursed 7(a) loans in the FY2017 to FY2021 cohort had been charged off, a fixed-horizon rate that counts write-off dates rather than default dates, and the file cannot say whether any given charge-off followed a slow ramp. Table 6 sets out each figure with its basis and key.

Loan Data Behind This Article

Every figure in this table comes from the Loan Analytics loan dataset, analytics-loan-data freeze-3 (commit 360c856), and carries its key. SBA rates are approval-cohort floors and are withheld when fewer than 30 loans stand behind them.

Testing a ramp schedule, line by line

A ramp schedule can be checked line by line, and most of its failures are visible on the page. The questions below follow the order of this article.

• Which climb is modelled. The study should say whether it is modelling an operating business, a building lease-up or both, and its stresses should match: a longer calendar for a lease-up, a lower plateau as well for a business.
• Where the clock starts. Funding, completion and opening are three different dates. If the coverage window is counted from funding, which is how the site's SBA program page words it for startups ^[6], the construction period is inside it, and the study should show how many months of operation are left.
• What stabilised means. The definition should be written before the model runs and should be measurable from reported results.
• Where each ramp assumption came from. Every assumption should name its tier: the project's own record, dated comparables, a published series, the operator's other locations, or the sponsor.
• Whether the comparables are dated. Each one should show its opening date and its occupancy at fixed months after opening, with concessions noted, since lease terms and concessions are among the market factors the interagency guidelines list ^[33].
• Whether a published rate is quoted properly. A SOMA figure should carry its completion quarter, whether it is an initial or final estimate, and its margin of error ^[65], and it should be converted from first rentals into occupancy.
• Whether the reserve was sized on a stress. The carry should be split into construction interest and operating shortfall, sized on an adverse case, and tied to a provision that lets the loan fund it, or else shown as the borrower's money.
• Whether the payment schedule fits the ramp. A B&I note must put its first full principal and interest installment no later than 3 years after the note's date ^[24]; under any program, the quarters in which coverage sits below the minimum should be visible in the schedule rather than averaged away.
• Whether both coverage tests are printed. Stabilised coverage against the binding minimum, and coverage year by year through the ramp, should both appear, on a quarterly schedule where the program asks for one ^[25].

A study that passes those checks has not removed the risk of a slow ramp. It has measured it, put a price on carrying it, and shown the lender where the money to carry it comes from.

Related Pages on Loan Analytics

• The Loan Analytics methodology series (all articles)
• How a feasibility study is done: the Loan Analytics methodology
• SBA feasibility studies (7(a) and 504)
• USDA feasibility studies
• Bankable feasibility studies for conventional loans
• Loan data by state
• Loan data by asset class
• Worked examples (case studies)
• Worked Example: USDA Business and Industry Loan Glamping Resort Feasibility Study in Ohio
• Worked Example: SBA 7(a) Assisted Living Feasibility Study in Utah
• Worked Example: Conventional Bank Construction-to-Permanent Multifamily Feasibility Study in Nebraska
• Stabilized occupancy benchmarks for SBA and USDA loan underwriting
• Supply and Comparable Analysis: Counting Competitors, the Pipeline and the Properties a Study Leans On
• Cancellation Is Not Charge-Off: The Two Ways an SBA Approval Can End Without Repayment

Commission a Feasibility Study

Loan Analytics prepares lender-grade feasibility studies for SBA 7(a), SBA 504, USDA Business and Industry, conventional and EB-5 financing, built on the methods this series sets out. A study is not an appraisal: it tests whether a project's market, costs and cash flow can carry the debt a lender is asked to approve. To discuss a study for your project, contact Loan Analytics.

Daniel Smith, MAI, Loan Analytics