Demographic Analysis in Feasibility Studies: How Institutional Reports Convert Population Data Into a Bankable Demand Forecast

1. Why Demographics Decide the Conclusion

Demographic analysis is the load-bearing input of a feasibility study. Every figure a lender ultimately underwrites, the demand projection, the capture or penetration rate, the absorption schedule, the stabilized revenue line, the debt service coverage ratio, is derivative of a population picture drawn around the subject. A defensible demographic foundation produces a defensible conclusion. A borrowed or stale one invalidates everything above it, no matter how polished the financial model looks.

The reason is the same multiplicative logic that governs the Primary Market Area. Demand equals a qualified population or visit base multiplied by a capture rate. The population base is demographic. Overstate it by twenty percent through a generous trade area or an out-of-date growth assumption, and aggregate demand rises roughly twenty percent, which mechanically depresses the implied capture rate the project needs against the same competitive supply, and makes a marginal deal read as comfortable. Reviewers know this. The SBA lender, the USDA reviewer, the CMBS underwriter, and the state housing finance agency all test the demographic foundation before they read the pro forma, because they know the pro forma inherits whatever error the demographics contain.

The error is rarely in the spreadsheet. It is in the inputs the spreadsheet trusts. A discounted cash flow model is only an engine for compounding assumptions, and the most consequential assumptions in any feasibility study are demographic: how many qualified households or daytime workers sit within reach, how fast that base is growing or shrinking, what share of it the subject can realistically capture against the competition. Get those right and a modest financial model is bankable. Get them wrong and the most sophisticated model in the market is precise about the wrong thing. This is why the demographic section is read first and read hardest, and why a study that treats it as boilerplate, a page of census tables lifted from a data vendor with no method declared, has already lost the reviewer before the financial exhibits begin.

This paper sets out the discipline of that foundation: the regulatory anchors that bind it, the data apparatus it draws on and the points at which that apparatus breaks, the four substantive demographic variables that carry the analysis, two worked examples that show the arithmetic institutional reviewers expect to see, the asset-class conventions, a defensibility checklist, and the failure modes that invalidate a demographic conclusion in peer review.

2. Four Regulatory Anchors

A feasibility study cites the regulatory standard that binds its loan program. The demographic section is where most of those standards actually land, because market and economic feasibility are demographic before they are financial.

SBA SOP 50 10 8 (effective June 1, 2025) (1) is the operative procedural standard for SBA 7(a) and 504 lending, superseding SOP 50 10 7.1 for all loans receiving an SBA number on or after that date. The SOP does not name an independent feasibility study as a standalone deliverable, but it functionally requires demographic substantiation through three channels. The first is the credit memorandum, which must justify the projected cash flow on which the credit decision rests; for any project where projections do real work, that justification cannot be made credibly without a demographic and market substrate. The second is a set of prescriptive projection tests for specific deal types. The clearest is the lodging eligibility test, which requires a start-up hotel applicant's projections to show that more than fifty percent of revenue will derive from transients staying thirty days or less, a demand-segmentation exercise that is demographic at its core. The third is the limited or special purpose property regime. The SOP's special-purpose list sweeps in car washes, gas stations, hotels, motels, nursing homes, assisted living, bowling alleys, theaters, golf courses, and roughly two dozen other owner-operator asset classes, and it requires the lender or CDC to address whether the project property is limited or special purpose in the credit memorandum and to state a conclusion. Designation carries real consequences: borrower equity injection escalates from a 10 percent baseline to 15 percent for special-purpose or start-up projects and to 20 percent where both conditions apply, and the appraiser must have completed no fewer than four going-concern appraisals of equivalent special-use property within the prior 36 months (1). For the entire special-purpose category, a third-party feasibility study is the practical means of supporting a projection the credit committee will otherwise reject.

USDA 7 CFR Part 5001 (2), the OneRD Guaranteed Loan Regulation governing Business & Industry, Community Facilities, Water & Waste Disposal, and REAP guaranteed loans, is the most explicit demographic mandate in federal commercial lending. Section 5001.3 defines a feasibility study verbatim as a report by an independent qualified consultant evaluating the economic, market, technical, financial, and management feasibility of a project. For B&I, an independent feasibility study is mandatory for any guaranteed loan above one million dollars to a new business under Section 5001.306, and for loans at or below that figure the Agency may require one where the lender's analysis is insufficient to determine technical feasibility or economic viability. A residual provision lets the Agency demand a study on any application, regardless of size, where it cannot otherwise determine a basis for repayment. Project eligibility itself is a demographic test: the site must lie outside any city or town of more than fifty thousand inhabitants and outside the contiguous urbanized area, with population drawn from the most recent decennial census (2). The threshold is not uniform across the OneRD programs; Community Facilities direct loans apply a 20,000-inhabitant test, and Water and Environmental direct funding a 10,000 test, so the analyst must confirm which boundary binds the specific program. Appendix A to Subpart D requires the market component to define a service area and populate it with current demographic and economic data, which in practice means current ACS five-year estimates rather than figures carried over from the prior decennial. The FY2026 fee notice raised the B&I guarantee to 90 percent for loans under five million dollars and cut the initial fee to one percent, shifting sub-five-million B&I economics materially closer to SBA 7(a), but it left the feasibility requirements unchanged in substance.

USPAP Standards Rule 1-3 (3), 2024 Edition, requires the appraiser to identify and analyze the market area trends and the economic supply and demand conditions affecting the use and value of the subject. SR 1-3 is the institutional standard for market-area identification in appraisal practice, and by extension in feasibility-adjacent appraisal work, where the demographic profile is the evidentiary basis for the trend analysis. A market-area trend conclusion unsupported by demographic data is, under SR 1-3, an unsupported opinion.

Census Bureau ACS data standards (4) function as the de facto fourth anchor. No single rule states "use the ACS," but every framework above presumes current, citable population and income data, and below the relevant population thresholds the American Community Survey five-year estimates are the only source that exists. A study that populates a service area with prior-decennial figures when current ACS data is available has, in USDA terms, failed the data-recency expectation that Appendix A implies, and in appraisal terms has failed the support requirement of SR 1-3.

The convention is that a study names the anchor that binds its loan. SBA-financed studies cite SOP 50 10 8. USDA studies cite 7 CFR Part 5001. Appraisal-adjacent work cites USPAP. A study that fails to name its anchor has not declared the standard against which it expects to be reviewed, and a reviewer who cannot find the anchor treats the methodology as the consultant's preference rather than the program's requirement.

3. The Data Apparatus and Where It Breaks

Nearly every demographic figure in a U.S. feasibility study traces to the American Community Survey, the rolling sample that replaced the decennial census long form after 2005 and now surveys roughly 3.5 million addresses a year (4). The most current vintage is the 2020-2024 five-year file, released in January 2026. Three properties of that apparatus govern how it must be used, and each is a common point of failure.

The 65,000 threshold. ACS single-year estimates are published only for geographies above 65,000 population (4). Every census tract, every block group, most ZIP Code Tabulation Areas, and the majority of U.S. counties exist only in the five-year file, which pools sixty months of collected data for stability. The geographic hierarchy nests from block to block group, which targets roughly 1,500 people, to census tract, which targets roughly 4,000, to county and beyond; the smaller the geography, the thinner the sample behind any single estimate. For any small or rural trade area, and that includes essentially every USDA-eligible market, the five-year estimate is the only honest source, and it describes a five-year average rather than the current year. A study quoting a single-year figure for a sub-65,000 market is quoting a number that does not exist, and a reviewer who knows the apparatus will catch it immediately.

Margins of error. Every ACS estimate carries a margin of error at ninety percent confidence, and that error grows sharply as geography shrinks. Tract-level estimates average roughly seventy-five percent more error than the legacy long-form estimates they replaced (5), and at the block-group level a median income estimate of $60,000 routinely carries a margin of plus or minus $20,000. The institutional discipline is to compute a coefficient of variation, the standard error expressed as a fraction of the estimate, and to treat anything under fifteen percent as reliable, fifteen to thirty-five percent as usable only with the margin disclosed, and above thirty-five percent as unfit to cite without aggregating to a larger geography (5). The discipline is not academic. A study comparing block-group medians to the dollar, with no error band, is reporting noise as signal, and an affordability or income-qualification conclusion built on such a comparison is indefensible because the true value could sit far enough from the point estimate to reverse the conclusion. Pandemic-era five-year files, including the 2020-2024 vintage, carry elevated uncertainty in some geographies and warrant additional caution.

No forecast. The Census Bureau does not project. Current-year estimates and five-year forward projections come from commercial vendors, principally Esri Updated Demographics (2025/2030 vintage, released June 2025) and Claritas Pop-Facts Premier (2026 vintage) (6). Each produces current-year and five-year figures controlled to county totals, on the logic that county-level information is more reliable than block-group extrapolation, and each is paired with a household segmentation system: Esri's ArcGIS Tapestry, rebuilt in June 2025 to 60 distinct segments, and Claritas PRIZM Premier at 68 segments (6). These products are necessary and credible, but they are not independent verifications of one another; both lean heavily on the same ACS five-year base and the same postal-delivery counts, so they agree with each other more closely than either agrees with ground truth in fast-moving markets. Where two vendors diverge by more than roughly ten percent at the block-group level, that disagreement is itself the signal to verify against ground truth, building permits, utility connections, school enrollment, rather than to adopt the more favorable figure. Segmentation systems describe modal neighborhood behavior, not individual households, and applying a segment label to a specific address overstates certainty through the ecological fallacy.

The discipline reduces to four disclosures: name the vintage, name the vendor, disclose the margin, and never mix vintages within a single study.

4. Four Substantive Variables

Beneath the data apparatus sit four demographic variables that do the analytical work. Which one is load-bearing depends entirely on the asset.

Residential versus daytime population. A trade area has two populations, and using the wrong one is among the most consequential errors in the discipline. Residential population is where people sleep, the standard census count and the basis of every ACS demographic table. Daytime population, what the Census Bureau calls commuter-adjusted population, is residents minus those who commute out plus the workers who commute in. The two diverge by an order of magnitude in employment cores, where a central business district can hold ten times its residential population on a weekday afternoon, and they invert in commuter suburbs. The data source is the Census LEHD program and its LODES Origin-Destination Employment Statistics, the current vintage of which carries employment through 2023, exposed through the free OnTheMap tool down to the census block (7), and corroborated by County Business Patterns for establishment and employment counts by industry (8). Convenience retail, fuel, quick-service, and coffee are daytime-population businesses, drawing the bulk of revenue from workers who may live miles away; grocery, big-box, and multifamily are residential; office is the inverse case, an asset that houses daytime workers rather than serving them, so the relevant input is employment density and the jobs-housing ratio. Remote work permanently shifted daytime population away from traditional job centers, and Census research has documented the move (9), which means any study citing pre-2020 worker counts for a daytime-driven asset rests on a geography that no longer exists; the current LODES vintage is the only defensible input.

Income, read as a distribution. Median household income is the right headline measure, robust to the outliers that distort the mean, but the median alone is rarely sufficient. The national median sat near $78,500 in the most recent multi-year ACS figures (10), a benchmark, not a substitute for the trade area's own distribution. The distinctions matter: median household income is the fiftieth-percentile household, mean income is pulled upward by high earners, per capita income divides total income across everyone including children, and disposable or effective buying income is after-tax and the conceptually correct basis for any spending model, since pre-tax income overstates spending capacity by twenty to thirty percent. The distribution is the analytical point. A $75,000-median trade area supports wildly different concepts depending on its upper tail: if the top decile earns $150,000 the base for a high-ticket concept is thin, while if it earns $400,000 the same median conceals a real luxury market. The ACS sixteen-band income table is where that lives, and for high-ticket retail and senior living the relevant figure is not the median at all but the count of households above an absolute threshold. Federal expenditure data show the top income quintile spending more than four times the bottom on discretionary categories (11), which is why income-elastic concepts must read the bands, not the median, and why the affordability test for income-sensitive assets runs against the relevant quartile rather than the midpoint.

Growth and migration vectors. Population change decomposes into births, deaths, net domestic migration, and net international migration, each from a distinct source: the Census Population Estimates Program for official annual totals and components, IRS Statistics of Income for county-to-county flows carrying adjusted gross income, the ACS for the characteristics of migrants, and commercial mobility panels such as Placer.ai for near-real-time inflection that the federal releases confirm only months later (12). Composition matters more than the headline rate, and four questions separate analysis from a growth-rate citation: whether the growth is domestic or international, whether it is natural increase or migration, whether it is accelerating or decelerating, and whether the income mix of in-migrants is rising or falling. The current environment makes the discipline urgent. U.S. population growth slowed to roughly half a percent in the year to mid-2025, which the Census Bureau attributed to a historic decline in net international migration from about 2.7 million to 1.3 million, with a further sharp drop projected for 2026 (13). The practical consequence is direct: markets that relied on immigration to mask weak domestic fundamentals now face a step-change in demand, while markets with genuine domestic in-migration retain a more durable thesis, and a 2026 study built on 2021-to-2023 surge-year growth rates is describing a trajectory that has already bent. The defensible move is to underwrite to the post-2024 run rate, not the peak.

Age and household cohort structure. Demand for age-sensitive assets is a function of cohort counts, not total population. Multifamily turns on the 20-to-34 prime renter cohort; senior living turns on the 75-plus income-qualified household supported by the 45-to-64 adult-child caregiver cohort; family-oriented retail and schools turn on the presence of young children. These are ACS table reads, but they are the demographic spine of the demand model for their asset classes, and a study using total population where a cohort count is required has used the wrong denominator and inflated the apparent demand pool, often by a multiple. The cohort is also where the growth analysis and the income analysis converge: the relevant question for multifamily is not how many people live in the trade area but how many 20-to-34-year-old households are forming there and whether their incomes clear the rent-qualifying line.

5. Two Worked Examples

Institutional reviewers expect to see the arithmetic, not a conclusion asserted over a hidden calculation. Two examples carry most of the demographic work in practice.

Worked Example A: Daytime Population and the Capture Base

Subject: a quick-service coffee concept on a commuter corridor in a submarket with a modest residential base.

• Residential population, 1-mile ring: 4,200

• Resident workers commuting out (LODES RAC): 1,900

• Non-resident workers commuting in (LODES WAC): 11,400

• Net daytime population: 4,200 − 1,900 + 11,400 = 13,700

• Effective demand base, weighting workers at 0.7 for a breakfast and lunch concept and residents at 1.0: (4,200 − 1,900) x 1.0 + 11,400 x 0.7 = 2,300 + 7,980 = 10,280 weighted daytime equivalents

A study that sized this site on residential population alone would have worked from 4,200 people and concluded the location was too thin to support the concept. The daytime-adjusted base of roughly 10,280 weighted equivalents tells the opposite story, and it is the correct one for a commuter-corridor coffee asset. The variable that decides the conclusion is the daytime population, sourced from LODES, not the residential count from the standard ACS table (7). The worker weighting is itself a judgment that must be disclosed and defended; a general-merchandise concept would weight workers far lower, around 0.3 to 0.5, because the workplace contributes less of a full retail occasion than it does for breakfast and lunch.

Worked Example B: The ACS Margin of Error and the Affordability Conclusion

Subject: a feasibility study quotes a block-group median household income of $60,000 to argue a trade area clears the income threshold for a Class A multifamily concept renting at $1,500 per month, which requires roughly $60,000 in annual income at the thirty percent affordability standard.

• ACS five-year estimate: $60,000

• Representative block-group coefficient of variation for an income estimate: approximately 20%

• Standard error: 0.20 x $60,000 = $12,000

• 90% margin of error: 1.645 x $12,000 = approximately ±$19,700

• Reported honestly: $60,000 ± $19,700, a true-value range of roughly $40,300 to $79,700

• Coefficient of variation: 20%, falling in the 15-to-35% "use with caution" band

The arithmetic forces the lesson. The estimate that appeared to clear the rent-qualifying threshold by exactly nothing could, within its own confidence interval, sit at $40,300, far below the line. A study that cites $60,000 as if it were precise has not earned the affordability conclusion. The fix is to aggregate to the tract or place level, where the coefficient of variation falls toward eight to thirteen percent and the band tightens enough to support a conclusion, or to disclose the margin and soften the claim accordingly, and in either case to test affordability against the lower quartile of renter income rather than the median (5).

6. Asset-Class Conventions

The demographic variable that carries the analysis is asset-class specific. The convention library is well established for the major institutional types, and the binding variable shifts by asset.

Multifamily turns on the 20-to-34 prime renter cohort, job growth, and net in-migration, bounded by the thirty percent rent-to-income affordability standard that HUD codifies as the cost-burden line. The binding test is not whether the median household income clears the rent-qualifying threshold but whether the lower quartile of renter household income does, because the qualified pool is the renter base, not all households. Roughly half of U.S. renter households are cost-burdened on the most recent ACS figures, which makes the affordability test the central demographic question for the asset (14). The Primary Market Area follows a 10-to-20 minute commute shed in suburban markets and a 1-to-2 mile radius in urban cores, with competitive supply anchored to CoStar or Yardi Matrix inventory.

Self-storage uses a three-mile ring with drive-time validation, with demand driven by renter share, household mobility, and multifamily density, measured against a national saturation benchmark that published sources place between roughly 6.3 and 9.5 net rentable square feet per capita depending on methodology and geography (15). The divergence is real and matters: a study must cite the underlying source by name and vintage rather than quoting a single industry figure, because the three principal sources differ by roughly fifty percent on methodology. Renter-heavy submarkets with rising apartment density are the structural demand pockets, since renters use storage at a higher rate than owners.

RV and boat storage inverts the storage logic. Demand is a function of registered-vehicle counts across a 20-to-50 mile catchment rather than ring population, because owners will travel for the right facility and dedicated supply is a small fraction of the owning-household base. The relevant denominator is county-level RV and boat registrations, not trade-area population, and the demographic profile of the owner, skewing toward higher household income, is the qualifier (16).

Convenience and fuel follow a 1-to-2 mile or five-minute drive convention, but the binding variable is commuter daytime traffic, not residential demographics. The site should sit on the going-home side of the road for evening capture, and the relevant benchmarks are the average daily traffic count on the adjacent road and the NACS State of the Industry per-store transaction figures, not the rooftop count nearby (17).

Car wash demand is the product of traffic count, vehicle ownership, and household income, increasingly mediated by subscription models that tighten the effective trade area toward the daily commute. Express siting commonly references a 15,000-to-25,000 AADT rule of thumb, a figure widely used in the industry but not codified by the International Carwash Association or any comparable authority, which is exactly why a study citing it must flag it as a convention rather than a standard (18).

Retail centers follow ICSC trade-area conventions, with the primary trade area defined as the geographic area generating sixty to eighty percent of sales. By center type the conventional radii run roughly three miles for a neighborhood center, three to six for a community center, five to ten for a power center, and five to fifteen for a regional mall, with ICSC's own disclaimer that these are typical rather than universal (19). Daytime population read alongside residential is decisive for any restaurant or quick-service tenant, and leakage analysis, comparing local sales to local demand, shows whether the trade area is exporting demand to competing centers or pulling it in.

Senior living follows NIC MAP Vision conventions, capturing seventy to eighty percent of residents inside roughly five miles for assisted living and memory care, with independent living and continuing-care communities drawing wider. The decisive discipline is the denominator. Feasibility penetration ceilings of roughly 10-to-12 percent for independent living, 6-to-8 percent for assisted living, and 1-to-2 percent for memory care apply to income-qualified 75-plus households, a far narrower base than all 75-plus households, against which NIC MAP occupied penetration runs closer to 4 percent. A penetration figure quoted without naming the denominator is indefensible, because the same property can look feasible or infeasible depending on which base the analyst silently chose, and the income-qualified threshold is not a flat dollar figure but the household income equal to or above the property's own annualized rent (20).

7. Defensibility Checklist

A defensible demographic analysis carries a specific set of disclosures. Institutional review tests against them. The following ten items capture the operative criteria, and a study that satisfies them is review-ready.

1. Population type named. Residential or daytime, chosen for the asset class and defended, not defaulted. Daytime claims sourced to LODES with the vintage stated and the worker weighting disclosed.

2. Trade-area method named. Ring, drive-time isochrone, or gravity model, defended on first principles for the asset, with construction inputs visible: radius justified, isochrone time-of-day and network source stated, gravity exponent and competitor set named.

3. Data vintage current. ACS five-year vintage cited and within the institutional 18-month recency window relative to report date. Vendor segmentation version, ArcGIS Tapestry June 2025 or Claritas PRIZM Premier, disclosed with release date.

4. Margins of error disclosed. Coefficient of variation reported for small-area estimates; estimates above the thirty-five percent threshold aggregated upward or flagged, not cited as precise.

5. The 65,000 rule respected. No single-year ACS figure quoted for a sub-65,000 geography. Five-year file used and labeled as a five-year average where applicable.

6. Income read as a distribution. Income bands reported alongside the median for any income-elastic concept. Affordability tests run against the relevant quartile, not the median alone.

7. Cohort denominator correct. Age-sensitive demand calculated against the correct cohort, 20-to-34 renters or income-qualified 75-plus seniors, with the denominator named explicitly.

8. Growth composition decomposed. Population change split into natural increase, net domestic, and net international migration, with the forward projection reconciled to the post-2024 run rate rather than the surge-year average.

9. Vendor figures reconciled. Commercial estimates cross-checked against the ACS baseline; divergences above roughly ten percent at small geography verified against ground truth, not adopted on convenience.

10. Regulatory anchor named. SOP 50 10 8, 7 CFR Part 5001, or USPAP SR 1-3 cited as the standard against which the study expects review.

A study satisfying these ten items is review-ready. A study missing three or more is exposed to material rework risk, and a reviewer who finds the gaps reads every downstream number with the suspicion the omissions earn.

8. Eight Failure Modes

Common review-stage failures in demographic analysis. Each is sufficient to invalidate a conclusion in institutional peer review.

Failure 1: Borrowed trade area. A three-mile ring lifted from a listing sheet rather than constructed for the asset's actual draw. Detected when the geometry ignores barriers, rivers, highways, jurisdictional lines, or when the radius matches no defensible convention for the asset class. Fix: construct the trade area from the method appropriate to the asset, before computing demand.

Failure 2: Residential population used for a daytime asset. A convenience, fuel, or quick-service site sized on the resident count alone, ignoring the commuter base that actually drives it. Mechanically understates demand and can kill a viable site on paper. Fix: compute net daytime population from LODES and weight workers explicitly.

Failure 3: Median income cited without margin of error. A block-group median quoted to the dollar with no confidence band, masking a true range wide enough to cross the threshold the study claims to clear. Fix: report the coefficient of variation and aggregate up where it exceeds the usable band.

Failure 4: Single-year data for a small market. A single-year ACS figure quoted for a sub-65,000 geography where only the five-year file exists. Fix: use the five-year estimate and label it as a five-year average.

Failure 5: Surge-year growth carried forward. A 2026 projection built on 2021-to-2023 growth rates that the 2024-2025 migration reset has already invalidated. Overstates forward demand and the absorption schedule that depends on it. Fix: reconcile the projection to the most recent components-of-change vintage and the post-2024 run rate.

Failure 6: Wrong cohort denominator. Senior-living penetration applied against all 75-plus households rather than income-qualified, or multifamily demand against total population rather than the prime renter cohort, inflating demand by a multiple. Fix: name and use the correct cohort denominator.

Failure 7: Vendor figure adopted without reconciliation. A Placer.ai or commercial-vendor number quoted without the ACS denominator that contextualizes it, or two vendors' divergent estimates resolved by silently picking the favorable one. Fix: reconcile every premium-panel figure to the public baseline.

Failure 8: No regulatory anchor. A study that never names the standard it expects to be reviewed against, leaving the demographic methodology as the consultant's preference rather than the program's requirement. Fix: cite the binding anchor at first reference.

9. The Demographic Triad

A defensible demographic analysis rests on three pillars: the right population, honestly measured, and correctly projected.

The right population binds the analysis to the variable the asset actually runs on: daytime versus residential, the correct age cohort, the income distribution rather than the lone median, defined within a trade area constructed for the asset rather than borrowed. Choose the wrong population and every downstream number is precise about the wrong thing.

Honest measurement binds every small-area figure to its margin of error, respects the 65,000 threshold, reconciles vendor estimates to the ACS baseline, and discloses vintage. A number without a confidence band is an assertion, not a measurement, and the reviewer treats it as such.

Correct projection binds the forward view to the components of population change and the current migration trajectory rather than a trailing average that the 2024-2025 reset has overtaken. The feasibility conclusion is a bet on the future demand base, and the bet is only as good as the projection method behind it.

A feasibility study that satisfies the triad, the right population, honestly measured, correctly projected, has earned its demand forecast. A study that does not has produced a financial model resting on a foundation the reviewer will test first and find wanting. Demographic analysis answers the question every feasibility conclusion begins with: who, exactly, is the demand, and can the data prove they are there. The rest of the study is downstream of the answer.

May 25, 2026, by Michal Mohelsky, J.D. Principal of MMCG Invest, LLC,

Reach out to discuss how our methodology supports your lending or development decision.

Michal Mohelsky, J.D. | Principal | mmcginvest.com 

Contact: michal@mmcginvest.com

Phone: (628) 225-1125

Disclaimer: This report is provided for informational purposes only and does not constitute investment advice. Data presented herein is derived from proprietary MMCG databases and third-party sources believed to be reliable; however, MMCG Invest makes no representation as to the accuracy or completeness of such information. Figures from third-party industry databases have been independently verified and, where appropriate, adjusted to reflect MMCG's proprietary analytical methodology. Past performance is not indicative of future results.

Sources

(1) SBA Information Notice, "Issuance of SOP 50 10 8 with Technical Updates," effective June 1, 2025; special-purpose property list and equity-injection provisions; lodging eligibility and transient-revenue test. sba.gov.

(2) USDA 7 CFR Part 5001 (OneRD Guaranteed Loan Regulation); §5001.3 feasibility-study definition, §5001.306 B&I feasibility trigger, rural-area population test, and Appendix A to Subpart D five-component framework. ecfr.gov/current/title-7/subtitle-B/chapter-L/part-5001.

(3) Appraisal Standards Board, USPAP 2024 Edition, Standards Rule 1-3 (market-area and supply/demand analysis), effective January 1, 2024.

(4) U.S. Census Bureau, American Community Survey methodology; sample-size definitions, geographic hierarchy, and the 65,000-population threshold for 1-year estimates; 2020-2024 5-year release (January 2026). census.gov/programs-surveys/acs.

(5) Spielman, Folch, and Nagle, "Patterns and Causes of Uncertainty in the American Community Survey," Applied Geography (2014); U.S. Census Bureau, "Understanding Error and Determining Statistical Significance," ACS General Handbook, Chapter 7.

(6) Esri, "Methodology Statement: 2025/2030 Esri Updated Demographics" and ArcGIS Tapestry June 2025 release notes (60 segments); Claritas Pop-Facts Premier 2026 release notes and PRIZM Premier documentation (68 segments). esri.com; claritas.com.

(7) U.S. Census Bureau, Longitudinal Employer-Household Dynamics (LEHD) and LODES Origin-Destination Employment Statistics; OnTheMap tool; current vintage through reference year 2023. lehd.ces.census.gov.

(8) U.S. Census Bureau, County Business Patterns (2023 release). census.gov/programs-surveys/cbp.

(9) U.S. Census Bureau, "Working From Home Caused a Spatial Shift in Daytime Population Away From Traditional Job Centers" (2023). census.gov.

(10) U.S. Census Bureau, American Community Survey 5-year median household income estimates.

(11) U.S. Bureau of Labor Statistics, Consumer Expenditures 2024, average annual expenditure by income quintile. bls.gov/cex.

(12) U.S. Census Bureau Population Estimates Program (components of change); IRS Statistics of Income county-to-county migration data; Placer.ai domestic migration analysis. census.gov; irs.gov/statistics/soi-tax-stats-migration-data; placer.ai.

(13) U.S. Census Bureau, "U.S. Population Growth Slows Due to Historic Decline in Net International Migration," Vintage 2025 estimates. census.gov.

(14) HUD cost-burden standard (30% of income to housing); U.S. Census Bureau renter cost-burden estimates (2019-2023 ACS); Fannie Mae Multifamily Economic and Market Commentary; Harvard Joint Center for Housing Studies, State of the Nation's Housing 2025.

(15) Yardi Matrix Self-Storage Report (January 2026); Self Storage Association Industry Data Report 2025; Self-Storage Almanac 2024. Saturation figures range approximately 6.3 to 9.5 NRSF per capita by source and methodology.

(16) RV Industry Association, 2025 Go RVing Owner Demographic Profile; Yardi Matrix self-storage and RV/boat facility tracking.

(17) NACS, State of the Industry (2026); per-store transaction and fuel-volume benchmarks. convenience.org.

(18) International Carwash Association industry data; Retail Petroleum Consultants capture-rate analysis. The 15,000-25,000 AADT siting figure is an industry rule of thumb, not codified by ICA.

(19) ICSC Research with CoStar, U.S. Shopping-Center Classification and Characteristics (January 2017); primary-trade-area definition as 60-80% of sales. icsc.com.

(20) NIC MAP Vision metro coverage and senior-housing penetration analysis; income-qualified versus all-75-plus denominator distinction. nicmap.com.