Supply and Comparable Analysis: Counting Competitors, the Pipeline and the Properties a Study Leans On

What the supply chapter of a study has to establish

A feasibility study that measures demand and stops there has answered half of a lender's question. Demand says how much of a service a trade area will buy. Supply says how many sellers will be offering it on the day the subject opens and on the day it is meant to stabilize, and on what terms. The supply chapter of a study therefore has three jobs. It counts the properties that compete with the subject today. It estimates what will be added before the subject reaches stabilized operation. And it selects a small set of comparable properties whose rents, occupancy, concessions and lease-up history the financial model can lean on. This article sets out how each job is done, why each step is there, and where each one tends to go wrong.

Bank supervisors put supply beside demand in the same list. The interagency real estate lending guidelines, printed as Appendix A to subpart A of the FDIC's 12 CFR part 365, name market factors a bank should consider, among them population and employment trends, zoning, and the vacancy, construction and absorption rates of the market, both as they stand and as projected ^[1], together with lease terms, rents and sales prices, again current and projected and with concessions taken into account ^[2]. The series article on conventional bank lending reads those lists against a bank's own policy; here they serve as the agenda for the supply count. Appraisers work under a parallel expectation, since the interagency appraisal guidelines say an appraisal should discuss market conditions, demand and supply factors among them ^[3]. The appraiser asks that question to support a value at a date. The study asks it to support a cash flow over a loan term, which is one reason a study is not an appraisal even where the two documents read the same market.

USDA's Business and Industry program writes the point into eligibility. The rule in question reaches B&I projects in excess of $1 million ^[4]. Among those, a project that would lift direct employment by more than 50 employees, and is likely to add output, services or facilities in an area that lacks the demand for them, is excluded ^[5]. The yardstick is the efficient capacity of the area's existing competitive enterprises, and the exclusion falls away where the financing would do those competitors no harm ^[6]. The series article on 7 CFR 5001 sets out the whole test. For the supply count the consequence is direct: on a project the rule reaches, the inventory of competitors and their utilization is evidence of eligibility, not background. The same regulation names competition among the conditions to which a B&I lender's credit evaluation gives consideration ^[7].

The angle of this article is the method of counting and choosing. How a trade area is drawn and how demand is estimated inside it belong to the series article on capture rates, penetration rates and gravity models; how long a new project takes to fill belongs to the article on ramp-up and absorption. Supply sits between the two, and both lean on it. The worked illustration below is a hypothetical self-storage facility, chosen because storage capacity is measured in a single unit, net rentable square feet, and because the New Jersey RV and boat storage worked example turns on exactly this count. Every figure in the illustration is a labelled assumption, and none describes a real property.

What counts as a competitor

The first decision is a definition, and it is the step most easily skipped. A competitor is a property that a customer of the subject would seriously consider in its place. The site's methodology page puts the test as a choice of properties that are, in its words, "genuinely substitutable for the subject rather than the merely nearby" ^[8]. Substitutability has several dimensions, and a study should test each one instead of relying on a name or a category.

An industry code is the weakest test, because codes are built to count businesses, not to draw competitive sets. The NAICS definition for self-storage lessors covers establishments renting space for self-storage, and the secure space it lists runs from rooms and lockers to containers and outdoor space ^[9]. A climate-controlled multistory building, a gravel lot leased for boat trailers and a container yard can all carry that code, which is why the series' asset map cannot separate RV and boat storage from self storage; the article on how codes become classes explains that limit. The RV park code is just as wide. Its definition takes in any site for a camper's gear, whether the guest pitches a tent or tows a travel trailer ^[10], so a tent campground and a full hookup resort share it. Lodging splits the other way. Hotels, motor hotels, resort hotels and motels fall in one industry ^[11], while housekeeping cabins and cottages, tourist homes and bed-and-breakfast inns are classified elsewhere, under other traveler accommodation ^[12]. A cabin project like the Catskills short-term rental example competes with whole houses rented by the night, which no hotel count would find.

Inside a code, a study tests substitutability on the dimensions a customer weighs. Location comes first, measured in drive time from where customers live or travel, within the trade area the demand model uses. Product comes next: unit sizes, climate control, door width and ceiling height for vehicles, security and access hours for storage; site type, hookups and length of stay rules for an RV park; room type, brand tier and meeting space for a hotel. Format and operating model follow. A staffed facility with a rental counter and an unstaffed one rented through a kiosk may sell the same space to different people, and a park that takes monthly workforce tenants is in a different market from one that sells weekend nights to vacationers. Age and condition close the list, since a new building and a thirty-year-old one seldom command the same rent.

The result is three groups, not two. Direct competitors match the subject on every dimension that matters. Partial competitors match on some of them: an older drive-up facility without climate control competes for part of the subject's market but not all of it. Non-competitors share a code or a street and nothing else. A study should count partial competitors at a stated weight instead of either including them in full or leaving them out, print that weight, and explain why it was chosen. In Table 3 the partial competitors enter at the weight given in Table 2, and a reader who disagrees can change it and recompute every line.

Each competitor counted should carry a record a reviewer can check: its location and drive time from the subject; its size in the unit the market trades in (net rentable square feet, sites, rooms or keys); its unit or site mix; the year it opened or was last renovated; its operating model; its quoted rates, with the date and the way they were collected, whether a posted rate card, a telephone or web quote or an interview with the manager; and its occupancy, with the source of that figure. A rate quoted by a leasing agent and a rate observed on signed leases are different kinds of evidence, and the record should say which one it holds.

Counting what exists: the field census and the public records behind it

The core of the existing inventory is a field census: every property inside the trade area that passes the substitutability test, visited or otherwise verified, recorded on one template. Public statistics cannot replace that census. They can check it, and a study should know exactly what each public source counts before treating it as a check.

County Business Patterns, the Census Bureau's annual series of business statistics by area, provides figures for establishments with paid employees ^[13], published as finely as the ZIP code ^[14]. Its unit is the establishment, a single physical location where business is conducted, which is not necessarily the same thing as a company ^[15]. That makes it a natural cross-check on a list of competitors: if the field census finds eight facilities in a county where the series counts fifteen establishments in the industry, the difference needs an explanation before the study goes further. Two features limit the check. The first is coverage. Businesses with an EIN but no employees, and businesses operating without an EIN, are excluded from the series ^[16], so an unstaffed storage facility, a family campground run without payroll or a cabin rented out by its owner may be invisible to it. The second is measurement. Its data items are establishment counts, employment in a single March week, and payroll ^[17]; the series says how many competitors there are and how many people they employ, never how many square feet, sites or rooms they offer. Its geography has limits too, because an establishment is sometimes placed by its mailing address and a relocation can reach the Census Bureau only after a significant delay, particularly for small businesses ^[18].

Nonemployer Statistics covers part of the gap. It is the only annual source of detailed data on businesses with no paid employees that are subject to federal income tax ^[19], and it covers those with receipts of $1,000 or more ($1 or more in construction) ^[20]. Its annual data are the number of establishments and their total receipts ^[21]: a count to set beside the employer count, not a list of addresses a study can visit.

The next step in the method used here turns the count of competitors into a measure of capacity, and for that the study goes to the property record. The agencies' 2006 guidance for banks with heavy exposure to commercial property, written about a bank's own market analysis, says sources of market information may include published research data, appraisers and agents, information maintained by the property taxing authority, local contractors, builders, investors and community development groups ^[22]. For a supply census a study should look to the taxing authority's parcel record first for building area, year built and land area, where the jurisdiction records them, and should still reconcile it with the rentable area an operator advertises, since gross building area includes corridors, offices and walls that no customer rents. The same guidance recognizes that an institution in nonmetropolitan markets may have fewer sources of detailed market data than one in a large metropolitan market ^[23]. The point carries over to a study of a rural project such as the North Dakota RV park example: where published series are thin, the field census carries more of the weight, and the study should say how complete it believes its census to be.

The pipeline: from a permit record to a stage-weighted count

Existing supply is what the subject faces on opening day. The pipeline is what it faces by the time it is meant to be stabilized, and a study that counts only the first has measured the wrong market. The pipeline is also harder to count, because it lives in records that were never designed for counting.

For housing, a federal series exists. The Building Permits Survey provides national, state and local statistics on the number and valuation of new privately owned housing units authorized by building permits ^[24], at the state, metropolitan, county and permit-issuing place levels ^[25], and because permits are public records the Census Bureau can publish local data without confidentiality limits ^[35]. For an apartment project the survey is the natural starting point of a pipeline count, and for assisted living as well, since the definitions treat assisted living units as housing units while nursing home units are not ^[36]. Its estimates are published by structure size, from single units to buildings of five or more ^[37], not by use, so the survey can show that a large building was permitted in a county but not whether it holds apartments or assisted living; the permit itself has to be read for that. The survey has other edges a study should state. Not every area requires a permit ^[38]. Nationally less than 1 percent of privately owned housing units are built where none is required, but the share varies greatly from state to state and among metropolitan areas ^[39]. Permit valuations may frequently differ from the true cost of construction ^[40], so the valuation column is not cost evidence; the article on construction cost evidence covers what is.

For every other asset class in this series the federal record is silent. The survey collected nonresidential permits as well until those data were discontinued in 1995 ^[26], and its housing unit statistics exclude transient accommodations such as transient hotels, motels and tourist courts ^[27]. A self-storage building, a hotel, an RV park or a car wash appears in no current Census permit series. Their pipeline has to be assembled from local records, and a study should name each record it searched and the date of the search: the planning or zoning board's agendas and minutes, site plan and special use applications, the building department's permit log, utility service applications where a utility publishes them, and the developer and operator announcements that come before any of these. The site's archived post on hotel markets where new supply outruns rate growth reads hotel pipelines market by market; the method here is the count a single project needs.

A permit, or an approval, is not a building. The Census Bureau's own construction statistics treat authorization and start as separate phases, covering units authorized by permit, units authorized but not yet started, units started, units under construction and units completed ^[41]. Its Survey of Construction follows sampled permits month after month until the work actually starts ^[42], and it publishes annual data on the length of time from authorization to start and from start to completion ^[28], which a study of an apartment project can use as a published measure of the lag, checked against the schedules of the local projects themselves. The survey makes no adjustment for units built without a permit in areas that require one ^[29], one more reason a drive through the trade area belongs beside the record search. The appraisal guidelines make the expectation explicit in one passage: writing about tract developments with unsold units, they say a reader of the appraisal should be able to understand the risks of the subject and its market, including the anticipated supply of competing properties ^[43]. A study of an operating property owes its reader no less.

Commercial pipelines have no published rate of decay from approval to opening, so the study has to state its own. The method used here sorts every known project into stages (under construction, approved but not started, application filed) and counts each stage at a weight, the share of that stage's space the study expects to be built and open before the subject stabilizes. Projects announced with no application on file are listed and described but carry no weight until a filing exists. The weights are judgments. They belong in the inputs table, where a reader can see and change them, not inside a total. A study should support them where it can: by tracing earlier years of the same planning records to see how many approved projects were in fact built, by checking whether a project under construction has its financing in place, and by noting which approvals expire, and when.

The worked illustration: a supply count carried through to occupancy

The illustration follows one hypothetical facility through the count. The trade area is a ten-minute drive. Inside it the field census finds the direct and partial competitors listed below, and the planning records show three projects at three stages. The demand side enters as a single figure, the occupied storage space each resident supports; in a real study that figure is derived and defended in the demand chapter, not borrowed from elsewhere.

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

Read Table 3 across its three columns. With existing supply alone, demand of 400,200 square feet against 420,000 square feet of supply, the subject included, implies market occupancy of 95.3%, comfortably above the 88% the sponsor's pro forma assumes. Weighting the pipeline by stage raises the competitive supply counted from 350,000 to 430,500 square feet, and implied occupancy falls to 80.0%, eight points short of the pro forma. Counting every pipeline project in full takes it to 69.6%. The last two rows turn the shortfall into a quantity a lender can test against the population forecast. At the demand model's space per resident, the weighted case needs 5,832 more residents in the trade area before the market as a whole reaches the pro forma occupancy, and the full pipeline case needs 15,333.

The subject's fair share of supply, 14.0% in the weighted case, is the starting point for the capture analysis in the demand chapter: a subject that must capture far more than its fair share needs a reason, such as a better product or location, that the comparable evidence supports. The pipeline also reshapes the ramp-up. Each project under construction will be leasing up while the subject does, drawing on the same new demand, which is why the absorption schedule in the ramp-up article should be built against the weighted pipeline, not against existing supply alone.

Occupancy is where supply reaches the loan. A cut from the pro forma occupancy to the occupancy the market implies flows through revenue to net operating income and from there to the coverage test.

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.

The stage weights are therefore natural variables for the sensitivity work set out in the article on stress testing. A lender should see coverage with the pipeline weighted by stage and again with every project counted in full, and the study should name the case it is prepared to stand behind.

Supply per capita: a screening ratio, and why it is no more than that

Supply per capita divides the competitive supply of a trade area by its resident population: square feet of storage per resident, hotel rooms or RV sites per thousand residents. It is quick to compute and easy to compare across places, and the site's archived self-storage saturation index ranks metropolitan markets on it. For a single project it is a screen and not evidence, for four reasons a study should state whenever it prints one.

The first reason is the denominator. The Census Bureau's Population Estimates Program produces estimates for the nation, states, counties, cities and towns ^[30], and its county series is total resident population ^[31]. That suits a use sold to the people who live in the trade area. It does not suit a hotel, an RV park or a cabin rental, whose guests come from somewhere else; a per capita figure for those classes divides by the wrong population. Each new vintage of the estimates covers every year since the last census and supersedes the earlier figures for those dates ^[44], so a ratio should be built from one vintage throughout and should name it.

The second reason is the boundary, and it is where the ratio is most easily abused. Table 4 keeps the market and the subject fixed and changes only the line drawn around them.

Drawn at a ten-minute drive, existing supply comes to 6.03 square feet per resident, and demand equals 95.3% of existing supply plus the subject. Widen the line to a fifteen-minute drive and count every competitor inside it, and the ratio rises to 6.98 while demand falls to 91.8% of supply: the wider market is better supplied than the inner one. The third column is the trap. It takes the population of the wider area but leaves out the competitors in its outer band, and the same market now appears short of space, at 5.80 square feet per resident with demand at 108.9% of supply. A study, and anyone reviewing one, should check that the population and the competitor count were drawn on the same map.

The third reason is that the ratio is silent on product and price. A market can be well supplied with old drive-up units and short of modern climate-controlled space, and a ratio built on total square feet will hide the very shortage the subject is meant to fill. The fourth is that a ratio means nothing without a reference, and a national average is a weak one. This illustration uses no published benchmark: the demand model's own space per resident, derived from the trade area's households and businesses and checked against the occupancy the comparables actually achieve, is the only reference it accepts. A study that cites an outside figure should name its source, date and method, and should print its own market's ratio beside it.

Choosing comparables: the properties a study leans on

Counting tells the study how much competes. Comparables tell it on what terms. From the competitive set a study picks a handful of properties whose operating evidence (rents customers actually pay, occupancy, concessions, the pace at which they leased up) will set the subject's revenue assumptions. A study's comparables do a different job from an appraiser's. An appraiser's comparables support an opinion of value; a study's support a cash flow, and the study never converts that evidence into a value for the subject. The selection follows from the substitutability test above, applied more strictly, because each comparable will carry weight in the rent conclusion.

• Location: inside the trade area, with similar access and visibility. A property outside it can serve only if the study shows that its market behaves like the subject's.
• Size: within a band around the subject's capacity, since a small facility and a large one can price and fill differently.
• Age and condition: close to the subject's at opening, or adjustable on evidence.
• Format and product: the same building type, unit or site mix and amenities, or a difference that can be measured.
• Operating model: the same way of selling, whether staffed or unstaffed, nightly or monthly, seasonal or year-round, branded or independent.
• Market position: stabilized, dealing at arm's length with unrelated customers, and not in a promotional or distressed phase.
• Data quality: rates and occupancy that can be verified for a known date.

Each comparable chosen should carry the same record as a counted competitor, plus the evidence the model will use: quoted rates by unit type with the date collected, the concessions on offer, occupancy and how it was learned, the year the property opened and, where it can be known, how long it took to stabilize. The interagency lending guidelines put concessions beside rents in their list of market factors ^[2], and the reason is practical. A quoted rent with a free month attached is not the rent the customer pays. Where a study cannot learn a comparable's occupancy, it should say so plainly instead of inferring a figure from a manager's passing remark.

Adjusting the comparables to the subject

No comparable matches the subject in every respect, so the study adjusts each comparable's evidence toward what it would show if the comparable were the subject. Table 5 applies the method to three rent comparables.

The order of the steps matters. Concessions come first, because they change what the comparable's customers actually pay: Comparable A's quoted $1.32 becomes $1.27 after its concession, and Comparable C's $1.45 becomes $1.33. The effective rent is then the base for every other adjustment. The property adjustments move each comparable toward the subject, upward where the comparable is inferior (Comparable B's older building, weaker location and smaller climate-controlled share) and downward where it is superior (Comparable C's stronger location). Each adjustment needs its own evidence, and the strongest evidence is paired: the same operator's rate for the same unit at two locations, or one facility's rate card for climate-controlled and standard units of the same size. An adjustment with nothing behind it is an assumption, and the study should label it as one.

Two summary rows test the result. The net adjustment says how far a comparable's rent moved; the gross adjustment, the sum of the individual adjustments without regard to sign, says how different the comparable was to begin with. Comparable A moves only 2%, Comparable B moves 15% on both measures, and Comparable C carries 9% of gross adjustment, all of it downward. A comparable whose gross adjustment passes the limit the study set, 25% in this illustration, is set aside however convenient its rent. Fixing the limit before the survey keeps it from being chosen to fit the answer.

Reconciliation weighs the adjusted rents by their reliability, and the reliability measure used here is the size of the adjustment: Comparable A, adjusted least, carries the most weight and Comparable B the least, as Table 6 shows. The concluded rent is $1.28 per square foot per month. A reader who prefers equal weights, or weights by distance, can apply them to the printed adjusted rents and see how far the conclusion moves.

The comparable that cannot be used

Some properties fail the tests outright and cannot be repaired by adjustment. The common cases are worth naming, because each arrives with a plausible reason to keep it.

• A property in lease-up. Its occupancy describes a building that has not stabilized, and its rents may be introductory. Federal appraisal rules single the condition out: an appraisal has to account for partially leased buildings, and for leases on non-market terms, through deductions and discounts it analyzes and reports, a list that also covers proposed construction and unsold tract units ^[45]. A study can still use a lease-up property, but as evidence of absorption pace for the ramp-up analysis, not as evidence of stabilized occupancy or rent.
• A related party lease or an owner's own use. A rent charged to an affiliate, or space the operator occupies itself, is not a market transaction.
• A different market. A property outside the trade area, or in a place with different incomes, density or land costs, belongs on the list only if the study shows the two markets behave alike.
• A different operating model. A resort campground priced by the night says little about a park leasing sites by the month, and a branded hotel with a restaurant says little about an independent motel.
• Unverifiable data. A rate that appears only in a sponsor's spreadsheet or a marketing brochure, with no date and no way to confirm it.
• A property in distress or repositioning. Rents cut to fill a building before a sale, or raised ahead of a renovation, describe the transaction and not the market.

The study should not drop such a property in silence. It should list the property, say why it was set aside and, where the sponsor offered it as a comparable or another report relies on it, show what it would do to the conclusion. Comparable D in Table 6 is that case: an urban multistory facility outside the trade area, quoting $2.05. Averaged in, with all four weights renormalised to 100 percent as shown, it lifts the concluded rent from $1.28 to $1.43 and the annual rent at the pro forma occupancy by $114,144. That figure is the cost of a lax comparable, and printing it lets a credit officer see that the conclusion does not depend on the property the study excluded.

Comparables also age. The interagency appraisal guidelines list limited or over supply of competing properties, and improvements to the subject or to competing properties, among the changes that can move an originally reported market value enough that a new appraisal or evaluation is necessary ^[46]. A study's comparable set carries a date for the same reason. If a large competitor opens, or a comparable renovates, between the survey and the credit decision, the survey should be refreshed rather than footnoted.

What the SBA loan record says about supply, and what it cannot

The loan dataset behind this series records SBA approvals, and it is tempting to read a count of approvals in a class and a place as a count of new competitors. It is not one, for reasons the program texts and the file's own dictionary make plain. 13 CFR 120.802 defines the 504 Project broadly, as buying or leasing long-term fixed assets, or improving or renovating them, for the small business to use ^[34], so a 504 approval can finance the purchase of an existing facility, not only the construction of a new one. SBA's data dictionary lists refinancing programs among the 504 processing methods ^[47]. The file carries a field describing the age of the business ^[33], but the cells this series publishes do not split approvals by it, and the field that says where the project occurs names a county, not an address ^[32]. It also holds only SBA-guaranteed loans. A facility financed by a bank on its own terms, by another lender or with cash leaves no trace in it.

With that understood, the record can still frame a supply chapter. Nationally, the self-storage class holds 946 ^[48] approvals under 504 and 1,584 ^[49] under 7(a) over fiscal years 2017 to 2026. The yearly series is uneven. The class's 504 approvals came to 181 ^[48] in FY2021 and 59 ^[48] in FY2026, and its 7(a) approvals to 296 ^[49] and 109 ^[49] in the same two years. Nothing in the file explains the swing, and no explanation is offered here. For a study the series serves only as a prompt to check the local record for the facilities such approvals may have financed, never as a count of what was built.

Cancellations make the same point as a planning pipeline. Of the self-storage 504 approvals in the FY2017 to FY2024 cohort, 19.15 percent ^[48] were cancelled, against 13.60 percent ^[50] for all 504 approvals in that cohort, and the 7(a) figure for the class is 11.14 percent ^[49]. Each is an approval-cohort floor: it can rise, but not fall, as approvals still undisbursed are cancelled. In the 7(a) file SBA's dictionary codes a cancelled loan CANCLD and keeps an undisbursed one apart as COMMIT ^[51], so an approval counted as pipeline can vanish just as a planning approval can lapse, and the file cannot say whether a cancelled project was later built with other money. How that floor is computed, and why it only understates, is the business of the series piece on cancellation and charge-off. The RV park and campground class, with 645 ^[52] 7(a) approvals, and the hotel class, with 8,942 ^[53], carry their own records, and the block below prints the figures this section uses with their keys.

Local cells are thinner. New Jersey self-storage approvals across both programs number 16 ^[54] over the ten years, spread across 10 ^[54] counties, and the cell's cancellation and charge-off rates are withheld because fewer than 30 loans stand behind them. Ocean County, where the New Jersey worked example is set, is not among the counties the cell lists. That absence says nothing about how much storage Ocean County has. It says only that no approval coded to the class names that county as the project's location. The worked example reaches the same place by its own road: it puts a drive-by count of competitors ahead of its other field work. The article on what the loan file cannot tell you sets out the file's other limits, and the article on the thirty-loan rule explains the withholding.

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.

What a lender or borrower should check

A supply chapter can be read quickly against a short list. Each item is a question the study should already have answered on its own pages.

• The definition. Does the study say what counts as a competitor, on which dimensions, and why each partial competitor carries the weight it does?
• The census. Is every counted property listed with its size, product, operating model, rates and occupancy, each with a date and a source? Does the count reconcile with County Business Patterns for the area, allowing for the businesses without employees that the series leaves out ^[16], and is any remaining gap explained?
• The boundary. Were the population and the competitor count drawn on the same map, and does the per capita ratio, if one is printed, divide by residents of that same area from a single vintage?
• The pipeline. Which records were searched, and on what date? How was each project staged, are the stage weights printed, and what do occupancy and coverage look like with every project counted in full?
• The comparables. Are they stabilized, at arm's length and verified? Were concessions taken out before the other adjustments, and is each comparable's gross adjustment shown against a limit set before the survey?
• The exclusions. Which properties were set aside, why, and what would including them do to the rent conclusion?
• The loan record. Is any SBA approval count presented as a count of new supply? It should not be.
• The date. Has anything opened, closed, been approved or been renovated since the survey?

A supply chapter that passes these checks does not guarantee the project. What it does is let a credit officer see the market the cash flow assumes, test each judgment in it, and change any one of them without rebuilding the study.

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: SBA 504 RV and Boat Storage Feasibility Study in New Jersey
• Worked Example: Conventional Community Bank Construction-to-Permanent Short-Term Rental Feasibility Study in New York
• Worked Example: USDA Business and Industry Loan RV Park Feasibility Study in North Dakota
• Hospitality Feasibility: Where ADR Growth Can’t Outrun New Supply
• Self‑Storage Saturation Index (Supply per Capita vs Rent Growth)
• Trade Area and Demand Methods: Capture Rates, Penetration Rates and Gravity Models
• Ramp-Up and Absorption: How Long a New Project Takes to Stabilise, and How a Study Proves It

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