Case Study · Arizona · Assisted Living & Memory Care · HUD 232
Assisted Living Feasibility Study, Arizona — A HUD 232 Worked Case
This is how our independent feasibility study company and senior housing feasibility consultant analyzed a new-build assisted living and memory care community underwritten to a HUD Section 232 permanent takeout, from age-75+ penetration and capture through the debt-service coverage the program requires. It is a representative, anonymized worked example of the methodology — not a specific client deal — set in a fast-growing retiree submarket in Arizona, a non-Certificate-of-Need state where senior-housing supply is market-driven.
A new 90-unit community in an Arizona retiree submarket.
A sponsor and an experienced regional operator came to our feasibility study company with a ground-up assisted living and memory care community and a two-stage capital plan that needed the projected cash flow independently tested before either lender would commit. The subject is a roughly 6-acre parcel in a high-growth retiree submarket of a major Arizona metro, on a medical-adjacent corridor within a short drive of a hospital and a concentration of adult-child households. The build program is 90 licensed units — 64 assisted living and 26 memory care — in a single-story, roughly 86,000-square-foot building, with the memory-care households arranged as a secured neighborhood.
Because assisted living is a going concern rather than a passive real-estate play, the lender's question is not “what is the dirt worth” but “can this specific community fill to a supportable census and generate the acuity-driven revenue and margin to service this specific debt.”15 The plan is the standard bridge-to-HUD path: a bank construction loan funds the build and lease-up, and a HUD Section 232 permanent loan takes it out at stabilization, which requires a 1.45x debt-service coverage ratio the file must document.4 Our scope was the independent demand, penetration, capture, competition, and coverage analysis that supports both credits.
Age-75+ penetration and capture, not a headline occupancy number.
The demand read starts with a qualified elderly population, not a metro occupancy figure applied to a unit count. We defined a drive-time primary market area (PMA) around the site and screened the 75-and-older population by age and by the income needed to pay privately.
The most common demand error in senior-housing studies is applying a national or metro penetration rate to a market area without qualifying the elderly population by both age and income — households that can afford roughly $5,000 to $7,500 a month of private pay.5 Penetration averaged about 10.2 percent of the 80-plus population across the top 140 markets in 2023, and the national 80-plus cohort is projected to roughly double between 2020 and 2040, so the age band and income screen — not an aspirational capture — drive the number.5 Revenue is acuity-driven rather than a flat rent: the 2025 Genworth and CareScout survey put the national median assisted-living cost at about $6,200 a month, and memory care commands a 20-to-30-percent premium on top of that.37 Applying a defensible penetration rate to the age- and income-qualified cohort, then netting the competitive supply, the PMA supports roughly 400 units of net unmet assisted living and memory care demand — and the subject's 90 units capture about 23 percent of it, leaving real cushion behind the forecast.
| Demand driver | Basis | Supported figure |
|---|---|---|
| PMA population age 75+ | ~22,000 residents (2026), growing ~4%/yr on retiree in-migration12 | Rising qualified base |
| Age + income-qualified cohort | Screened for ~$5,000–$7,500/mo private pay3 | Core need population |
| Penetration-supported demand | ~10% penetration, near the ~10.2% top-140-market benchmark5 | ≈ 900 units |
| Competitive supply (standing + permitted) | Surveyed AL/MC set, plus announced pipeline11 | ≈ 500 units |
| Net unmet demand | Penetration-supported less competitive supply | ≈ 400 units |
| Subject capture | 90 units against net unmet demand | ≈ 23% of net unmet |
Penetration and capture logic grounded in the ~10.2% top-140-market penetration benchmark and the qualified-population method; see sources 5, 3, and 12. Figures are illustrative of the engagement type.
An undersupplied submarket, with no CON brake on supply.
Six competing communities sit within the PMA, but only part of that inventory is truly comparable, and Arizona has no Certificate of Need law — so the study, not a regulator, carries the full supply burden. A rigorous survey scans permitted and announced supply, not just the standing set.
| Community | Segment | Units | Distance | Read |
|---|---|---|---|---|
| Community A | AL + MC | 110 | 3.2 mi | Stabilized ~92%; MC waitlist |
| Community B | AL only | 84 | 4.5 mi | Dated, no dedicated memory care |
| Community C | AL + MC | 128 | 5.1 mi | Newer, leasing toward ~85% |
| Community D | MC standalone | 48 | 6.0 mi | Memory-care only, ~90% |
| Community E | CCRC (IL-led) | 160 | 7.4 mi | Entrance-fee; different buyer |
| Community F | AL only | 72 | 8.2 mi | Tertiary edge of the PMA |
Competitive set surveyed for the engagement; anonymized. The entrance-fee CCRC serves an independent-living-first buyer and competes only partly; announced and permitted supply was scanned, not just the standing set.
The nearest genuine assisted living and memory care competitor is stabilized and carries a memory-care waitlist — a demand signal, not a saturation signal — while the dated assisted-living-only communities cannot match a new secured memory-care neighborhood. The entrance-fee CCRC is filtered out as a partial competitor because it serves an independent-living-first, buy-in buyer. Because Arizona operates no Certificate of Need program — it is among only three states with none, and imposes no bed cap on senior housing or skilled nursing — supply is market-driven and oversupply risk is genuinely elevated, so the capture forecast has to be earned from the qualified population rather than assumed away by a permit gate.11 Set against that, the read is a genuinely undersupplied submarket: the qualified 75-plus base is growing faster than new licensed beds are arriving, and the subject fills the gap rather than splitting a saturated market.
Arizona senior housing: structural tailwind, submarket discipline.
The national backdrop is a strong tailwind for needs-based senior housing, but Arizona is a Sun Belt state where metro averages hide oversupplied submarkets, so the study is built to the submarket, not the state.
Nationally the asset class is riding a structural tailwind while supply has collapsed. NIC MAP Vision reported Primary Market senior housing occupancy of 89.5 percent in the first quarter of 2026, the nineteenth consecutive quarterly gain, with assisted living at 87.9 percent and independent living above 91 percent.1 Units under construction fell to roughly 17,000 by the third quarter of 2025, the fewest since 2012, with record-low 0.4 percent inventory growth, and NIC MAP estimates roughly 806,000 additional units are needed by 2030 to hold current penetration just as the first baby boomers turned 80 in 2026.2 That is the demand engine a new Arizona community needs.
The offsetting reality is that Arizona is a Sun Belt market that added senior-housing supply into the last cycle, and it has no Certificate of Need supply brake. Phoenix active-adult occupancy sat at just 85.1 percent in the first quarter of 2026, tied among the lowest of the fifteen largest active-adult markets — a reflection of Sun Belt supply rather than weak demand, and precisely why a defensible Arizona study is built submarket-by-submarket against the current pipeline.10 The demographic and fiscal pull is real: Arizona levies a 2.5 percent flat personal income tax, the lowest flat rate in the country, metro Phoenix is projected to hold about 73.5 percent of the state population by 2060, and more than 630,000 Californians relocated to Arizona over a decade — though a 2025 Sun Belt migration slowdown is a caution the model carries rather than ignores.12
Why the parcel captures the submarket.
Retiree in-migration, adult-child proximity, and medical adjacency all point the same direction — and in Arizona, one greenfield variable most studies skip has to be cleared before any of it counts.
The primary market area carries a growing, income-qualified 75-plus base, inflated by retiree in-migration into a fast-growing metro whose county is among the most populous and fastest-growing in the country.12 The medical-adjacent site matters to needs-based demand in a way it does not for market-rate apartments: proximity to a hospital and physician offices shortens the referral path that fills assisted living and memory care, and a concentration of adult-child households in the PMA supplies the decision-makers who actually choose the community. Trailing population counts understate the qualified base in a submarket growing near 4 percent a year, a distortion the study corrects for rather than extrapolates.
One Arizona-specific gate is cleared before the demand work is credited. Under the 1980 Groundwater Management Act, development inside the Phoenix, Tucson, and Prescott Active Management Areas must demonstrate a 100-year Assured Water Supply, and after the Arizona Department of Water Resources found a roughly 4.86-million-acre-foot 100-year groundwater shortfall in the Phoenix AMA it paused new groundwater-reliant determinations. A defensible Arizona study verifies the specific water provider's Assured Water Supply or Alternative Designation status before underwriting any greenfield site — here the parcel is served on a designated provider, which removes the fringe-subdivision risk that stalls water-gated land.13
Bank construction, then a HUD 232 permanent takeout.
Total development cost lands at $28.80 million, near $320,000 per unit.9 Because HUD Section 232 new-construction processing is slow and the community must be licensed and stabilized to underwrite, the deal runs on a two-stage structure: a bank construction loan builds and fills it, and a HUD 232 permanent refinances the bank at stabilization.
| Cost component | Amount |
|---|---|
| Land (~6-acre medical-adjacent parcel) | $2.30M |
| Site work & utilities | $2.10M |
| Building construction (~86,000 sf) | $16.60M |
| FF&E (units, dining, care, memory-care) | $2.40M |
| Soft costs (A&E, permits, financing) | $2.60M |
| Contingency | $1.30M |
| Working capital & operating/DSC reserve | $1.50M |
| Total development cost | $28.80M |
Cost basis near $320,000 per unit ($28.80M ÷ 90 units) and roughly $335 per square foot, consistent with the ~$317,400/unit ($333/sf) ASHA development-cost benchmark and the $280–$452/sf assisted-living construction range. See source 9.
| Item | Figure |
|---|---|
| Bank construction loan (65% LTC) | $18.72M |
| Sponsor equity injection (35%) | $10.08M |
| Term / structure | ~24–36 months, interest-only w/ interest reserve |
| Recourse / rate | Recourse, floating during construction & fill-up |
| Item | Figure |
|---|---|
| Stabilized NOI (Year 3) | $2.52M |
| Sized at 1.45x DSCR (binding constraint) | max debt service ≈ $1.74M |
| Term / rate | 35-year fully amortizing, non-recourse, ~6% |
| HUD 232 permanent loan | ≈ $25.40M |
| Implied LTV (value ~$33.6M @ ~7.5% cap) | ~76% (under the ~80% ceiling) |
| Construction-loan payoff / capital returned | $18.72M paid off; ~$6.7M returned to sponsor |
HUD Section 232/223(f) parameters — 1.45x DSCR, ~80% LTV, 35-year, non-recourse — per HUD Handbook 4232.1 and Section 232 program materials; see source 4. The ~6.84% annual mortgage constant on a 35-year, ~6% loan implies a $25.40M loan at $1.74M of debt service.
The equity injection sits at 35 percent, well above a stabilized-asset norm, and that is deliberate: this is ground-up construction of a going concern with lease-up risk, so the construction bank sizes to 65 percent of cost and holds recourse until the community is built, licensed, and filling.14 The permanent loan is where the coverage test bites. Sized to a 1.45x debt-service coverage ratio on stabilized net operating income, the HUD 232 loan solves to about $25.40 million — below the roughly 80 percent loan-to-value ceiling, so coverage binds before leverage, exactly as it does across agency and HUD senior-housing credits.4 At $25.40 million the permanent comfortably repays the $18.72 million construction loan and returns roughly $6.7 million of capital to the sponsor at refinance, because the community is worth more stabilized than it cost to build. The study exists to support exactly that chain: the coverage the HUD lender must document, tested against an independent read of penetration and capture rather than the sponsor's own projection.
Feasible and financeable, on coverage HUD can document.
The stabilized model builds acuity-driven revenue from two segments, nets a position-by-position operating expense load, and grades coverage up a fill-up ramp to the HUD 232 floor and beyond.
| Line | Basis | Amount |
|---|---|---|
| Assisted-living revenue | 64 units × ~90% occ × ~$6,450/mo all-in3 | ≈ $4.46M |
| Memory-care revenue | 26 units × ~90% occ × ~$8,400/mo all-in7 | ≈ $2.36M |
| Total resident revenue | Assisted living + memory care | ≈ $6.82M |
| Operating expenses | ~63% of revenue; labor ~55% of opex67 | ≈ ($4.30M) |
| Net operating income (NOI) | Revenue less operating expense | ≈ $2.52M |
Operating-expense ratio near 63% sits inside the 55–70% band for stabilized AL/MC (versus 35–45% for multifamily), with labor the largest line at ~55% of opex, built position by position at documented local wages. See sources 6 and 7.
| Year | Stage | NOI | Debt-service basis | DSCR |
|---|---|---|---|---|
| Year 1 | Fill-up (pre-stabilized) | ~$1.30M | Perm-equivalent ~$1.74M | 0.75 |
| Year 2 | Approaching stabilization | ~$2.00M | HUD 232 perm ~$1.74M | 1.15 |
| Year 3 | Stabilized | ~$2.52M | HUD 232 perm ~$1.74M | 1.45 |
DSCR computed as NOI divided by the period debt-service obligation. The Year 1 shortfall is carried by the construction-loan interest reserve during lease-up; HUD permanent coverage is measured once the community reaches its supportable census. See source 4 for the 1.45x convention.
The stabilized 1.45x coverage is the figure the HUD lender documents, and it is reached the way senior housing actually fills — up a graded ramp, not on day one. Assisted living and memory care lease-up is slower and more needs-driven than multifamily; industry practice models 18 to 30 months to stabilization, and short length of stay compounds the challenge, with average assisted-living length of stay near 22 months and nearly 40 percent of residents leaving within the first year, so a community must re-lease a large share of its units every year just to hold census.8 Modeling a 12-month fill of a 90-unit community, or best-in-class census on day one, is one of the most common ways these pro formas fail review; the ramp here is deliberately benchmarked to the surveyed comparable communities, which is why Year 1 sits below coverage and the interest reserve, not resident revenue, carries the gap.
On the equity side, the $10.08 million injection funds a going concern with no distributions during construction and lease-up, then two things happen at stabilization. The HUD 232 refinance returns roughly $6.7 million of capital, leaving about $3.4 million of equity in the deal against stabilized levered cash flow near $0.78 million a year — NOI of $2.52 million less $1.74 million of permanent debt service — a strong post-refinance cash yield. The exit is valued on a going-concern basis, not a leased-fee cap rate: a licensed community transfers real estate plus FF&E plus the operating business, and going-concern value can far exceed the underlying real estate.15 Capitalizing a grown Year-10 stabilized NOI at a going-concern rate a touch wide of the ~6.2 percent average senior-housing cap rate, net of the amortized HUD balance and selling costs, and blending the lease-up drag, the mid-cycle cash-out, and durable stabilized distributions, the illustrative result is a levered equity IRR of about 16 percent over a 10-year hold.14
Verdict: financially feasible and financeable. On independently derived penetration and capture, a stabilized 1.45x DSCR that meets the HUD 232 floor, and a ~16% levered equity IRR, the projections support the bank construction loan and its HUD Section 232 permanent takeout.
Independent penetration, competition, labor, and coverage stress.
The engagement was scoped the way a HUD underwriter and a credit committee read it. As an independent feasibility consultant, our role is to test the sponsor's projection against the market, not to restate it — the value of the deliverable is precisely that it carries no stake in the outcome. We derived demand from a qualified 75-plus population screened by age and income, applied a defensible penetration rate rather than a metro occupancy figure, and netted an independently surveyed competitive set that scanned permitted and announced supply because Arizona's non-CON market has no regulatory supply brake. The labor model was built position by position at documented local wages, since wages, benefits, and contract staffing run roughly 55 percent of operating expense and are the single most common underwriting error in the asset class.6
The coverage analysis was then stress-tested. We ran the debt-service coverage against a slower lease-up, a lighter stabilized census, and an agency-labor re-escalation — the three variables a needs-based community is most exposed to — to confirm both credits still hold when the ramp lengthens or labor re-prices, and we checked breakeven occupancy against the stabilized census to confirm a cushion. Two scope boundaries are worth stating plainly: the going-concern appraisal, the Project Capital Needs Assessment, and the Phase I environmental site assessment that the HUD 232 package additionally requires are separate third-party engagements we reference but do not perform.4 That combination — independent penetration and capture, competition, a position-level labor model, and a stressed DSCR on a going-concern basis — is what lets both lenders rely on the file.
Underwriting an Arizona assisted living project for HUD 232? Start with the feasibility study.
Feasibility Study Company prepares independent assisted living and memory care feasibility studies for HUD Section 232, SBA, agency, and bank construction credits, built to the coverage standard your lender must document. A methodology briefing walks through the penetration, capture, competition, labor, and DSCR analysis behind a case like this one, calibrated to your submarket and unit mix.
Request a methodology briefingData sources and dates.
The deal figures are illustrative of the engagement type; the market data that grounds each dimension is real and sourced, drawn from our standing Arizona, Assisted Living & Memory Care, and Conventional & Institutional analyses and the primary authorities they cite.
- NIC and NIC MAP Vision press release (April 23, 2026): Q1 2026 Primary Market senior housing occupancy 89.5 percent, 19th consecutive quarterly gain; independent living above 91 percent, assisted living 87.9 percent; average asking rent above $5,650 per month; occupied units a record 637,000.
- NIC MAP Vision construction and inventory data (2025–2026) and Senior Housing Outlook (June 26, 2024): units under construction ~17,000 by Q3 2025 (fewest since 2012); year-over-year inventory growth a record-low 0.4 percent in Q1 2026; roughly 806,000 additional units needed by 2030 to hold current penetration; the first baby boomers turned 80 in 2026.
- Genworth and CareScout Cost of Care Survey: 2025 national median assisted living cost $6,200 per month ($74,400 per year); 2024 survey $5,900 per month, up 10 percent; base room rate plus acuity-based care and level-of-care fees, so revenue is acuity-driven rather than a flat rent.
- U.S. Department of Housing and Urban Development, Handbook 4232.1 and Section 232 program materials (Office of Residential Care Facilities, LEAN): minimum 1.45x DSCR for market-rate assisted living and skilled care; Section 232/223(f) up to roughly 80 percent LTV, 35-year fully amortizing, non-recourse; state licensure with continuous protective oversight; third-party report set (market study, going-concern appraisal, PCNA, Phase I ESA); bridge-to-HUD takeout of bank construction at stabilization.
- Plante Moran (2024), citing Census projections: penetration rate averaged roughly 10.2 percent of the 80+ population across the top 140 markets in 2023; the U.S. 80+ population projected to roughly double from about 13 million (2020) to about 27 million (2040); qualified-population method requires an age band and an income screen.
- PHI, Direct Care Workers in the United States: Key Facts 2025 (June 2025; BLS OEWS analysis): direct-care median wage $17.36 per hour in 2024, with turnover routinely cited at 40 to 80 percent; labor built position by position is the largest operating line.
- Janover Pro and MMCG (2025): labor roughly 55 percent of assisted-living and memory-care operating expense; stabilized AL/MC operating-expense ratios 55–70 percent of revenue versus 35–45 percent for stabilized multifamily; memory care runs 20–30 percent higher rates with tighter staffing ratios and compressed margin.
- NCAL and Senior Housing News operator survey (2025): average assisted living length of stay near 22 months; roughly 18–28 months in AL and 18 months to three years in memory care; nearly 40 percent of AL residents leave within the first year; industry practice models 18 to 30 months to stabilization.
- ASHA / State of Seniors Housing (via McKnight's) and The Weitz Company ASHA construction-cost briefs (early 2026): total development cost roughly $317,400 per unit ($333 per square foot); assisted living construction $280–$452 per square foot by tier.
- NIC MAP Vision, senior housing and active-adult occupancy (Q1 2026, released April 23, 2026): Phoenix active-adult occupancy 85.1 percent, tied among the lowest of the 15 largest active-adult markets, reflecting Sun Belt supply rather than weak demand.
- National Conference of State Legislatures (via the Arizona HB2197 bill summary): Arizona among only three states operating no Certificate of Need program; no CON for hospitals, nursing or skilled-nursing beds, or ambulatory surgery, so senior-housing and skilled-nursing supply is market-driven; the only CON-like requirement is the ground-ambulance Certificate of Necessity under A.R.S. §36-2233.
- U.S. Census Bureau, Vintage 2024 Population Estimates, and Arizona Office of Economic Opportunity: Phoenix–Mesa–Chandler MSA 5,186,958 and Maricopa County 4,673,096 as of July 1, 2024; Arizona approximately 7.6 million; metro Phoenix projected to hold about 73.5 percent of state population by 2060; StorageCafe Migration Trends (2024): more than 630,000 California-to-Arizona movers over a decade, with a 2025 Sun Belt migration slowdown; Arizona 2.5 percent flat personal income tax (Proposition 132).
- Arizona Department of Water Resources, Phoenix AMA groundwater model (June 2023): 100-year Assured Water Supply requirement in the Phoenix, Tucson, and Prescott Active Management Areas; roughly 4.86 million acre-foot 100-year shortfall and a pause on new groundwater-reliant determinations; Alternative Designation of Assured Water Supply (ADAWS) pathway.
- U.S. Small Business Administration SOP 50 10 8 (effective June 1, 2025); Fannie Mae and Freddie Mac Seniors Housing program guides; CBRE 17th and JLL 2026 Seniors Housing & Care Investor Surveys: licensed assisted living eligible for SBA with a special-purpose going-concern appraisal; agency segment DSCR minimums 1.30x (IL) / 1.40x (AL) / 1.45x (MC); average Q4 2025 senior-housing cap rate 6.2 percent (JLL), with assisted living slightly wider.
- Appraisal Institute, The Appraisal of Real Estate: going-concern and business-enterprise-value method allocating the total assets of the business among real property, tangible personal property (FF&E), and intangible or business value; going-concern value can far exceed the underlying real estate, and if a facility goes dark the intangible value evaporates.