Owner-Occupied Medical Office Building Under SBA 504 in Hennepin County, Minnesota: A Model Study

Model study: an illustrative feasibility analysis built on public data for Hennepin County, Minnesota; not a client engagement.

Project summary

The credit question

A lender here is not lending against a building. It is lending to a physician practice that wants to own one. Repayment comes from the group's cash flow after paying its physicians market compensation. The building is collateral, and its value is partly tied to the tenant whose failure would put it on the market.

Three things must be true. The group's cash flow must cover the new combined debt service with a margin that survives realistic hits to physician income: reimbursement cuts, payer-mix drift and the loss of productive physicians. The transaction must meet SBA's rules for owner-occupied new construction, including the 60% occupancy and 20% leasing limits and the equity SBA requires for medical facilities. And if the group fails, another medical user must value the building enough to cover the first lien. This study finds the first two met with conditions, and the third met only for the first lien, which is the normal shape of a 504 medical office credit.

There is a fourth, procedural question that shapes the structure. SBA lends only where credit is not available elsewhere, and the lender must state in its credit memorandum the identifiable weakness that keeps a conventional loan out [1]. Here the weakness is plain: a new medical building is worth less than it costs, so a conventional loan at 80% to 85% of cost would exceed prudent loan-to-value limits. The 504 second lien fills that gap, and the rest of this study asks whether the cash flow justifies it.

Location and county economy

Population and age structure

Hennepin County had an estimated 1,284,784 residents in 2025, up only 3,218 (0.3%) since 2020 [2]. From April 2020 to July 2025 it recorded 22,515 more births than deaths and 35,776 net international arrivals, but lost 54,951 residents to domestic out-migration [2]. A practice here cannot rely on population growth. It can rely on aging.

Source: Minnesota DEED county profile, Census 2025 estimates [2]; ten-year bands combined; shares rounded.

The median age is 37.6, younger than the state's 38.8 [2]. That supports a commercially insured payer base today. The State Demographic Center projection, published in the same profile, changes the picture over the loan term:

Source: State Demographic Center projections in [2]; under-65 row by subtraction.

Population is flat, but residents over 75 rise by 32,095, or 34.9% [2]. They see physicians more often and carry more chronic disease, which suits a multi-specialty group. They also move revenue toward Medicare, whose rates are set federally rather than negotiated. That trade-off drives the reimbursement analysis below.

Income, employment and unemployment

Median household income was $97,653 in the 2020 to 2024 American Community Survey, against $89,062 statewide, ninth of 87 counties [2]. Median family income was $134,878 and per capita income $59,401 [2]. Higher incomes go with employer coverage, the payer category with negotiable rates, but also with a wage base above the state average: DEED estimates a single adult needs $20.33 an hour to meet basic costs in Hennepin, against $18.45 statewide [2].

Source: DEED Quarterly Census of Employment and Wages in [2].

Health care is the largest sector and added 5.5% in a year when county employment was flat [2]. DEED projects health care and social assistance in the Twin Cities region to grow 9.2% from 2022 to 2032, adding 25,326 jobs, the largest gain of any industry [2]. The group competes with the region's large health systems for staff.

Unemployment averaged 3.6% in 2025, below the state's 3.9% [2]. A looser labour market would ease hiring of support staff but put some commercially insured patients at risk of losing coverage. Some 83.8% of employed residents work inside the county and the mean commute is 21.9 minutes [2], so patients choose clinics near home or work. Site visibility on an arterial road matters more than proximity to the core; a site study would add state traffic counts.

Healthcare workforce

Source: DEED wage, vacancy and occupations-in-demand data in [2].

Support staff are about a quarter of the modelled group's revenue, and the region carries 12,600 open healthcare jobs [2]. Health care and social assistance accounted for about a quarter (26%) of all Twin Cities vacancies in 2024 [2]. A clinic that cannot staff its exam rooms cannot bill for them, so staffing is a revenue risk as well as a cost risk. That is why this study stress-tests an 8% wage step rather than assuming staff costs track inflation.

Physician supply

Hennepin is home to academic and system hospitals, so a new practice should not count on unmet demand; this model is built for an established group bringing its patient panel. A departing physician also has local options, which makes retention agreements a credit term, not an HR detail.

Taken together, the county data support an owner-occupied clinic for an established group and argue against a speculative medical building. Demand growth comes from aging inside a flat population and the workforce is contested. The credit therefore rests on the group's existing revenue and its ability to keep its physicians, not on market growth.

Market analysis

Demand drivers

Two sourced drivers support outpatient demand: growth of the 75-plus population [2] and health care job growth [2]. The counter-driver is Medicaid. Public Law 119-21, enacted July 4, 2025 [3], requires states, from the first quarter beginning after December 31, 2026 at the latest, to condition Medicaid for expansion adults on at least 80 hours a month of work or other community engagement, to redetermine their eligibility every six months, and to shorten retroactive coverage to one month for expansion adults and two months for others [3]. The county's 2025 financial report expects rising uncompensated care pressures from late 2026, and its safety-net medical center recorded a $38.1 million operating loss in 2025, excluding a pension adjustment, related to elevated uncompensated care [4]. A private group's exposure is smaller, but its Medical Assistance and MinnesotaCare patients will churn.

Demand method: age-weighted visit demand

The primary test is whether the group's existing panel fills 12 physicians and 6 advanced practice providers (APPs). The secondary test is whether county demand supports volume growth after stabilization. This study weights residents by age, with the weights as assumptions reflecting higher visit frequency in older groups.

Demand rises 2.6% over the decade, about 0.26% a year, with population flat. The pro forma holds per-provider volume growth after Year 3 to 0.25% a year, so the practice grows with its market, not faster.

Existing supply, rents and construction cost

No public source tracks medical office vacancy and rents for buildings of this size in the county, so the model's current rent, the sublease rent, the valuation rent and the re-leasing cases are assumptions, listed in the assumptions table. A survey of local medical office comparables, including smaller buildings, is a condition of the model's determination.

The model's $205 per gross sq ft shell and $180 per rsf owner improvements are illustrative assumptions to be replaced with a contractor's guaranteed maximum price.

Sublease absorption and forecast

The 3,600 rsf sublease suite is small, so the risk is tenant quality and timing more than market depth. The best candidates are complementary users that send or receive referrals, such as physical therapy or a specialty practice the group does not offer, because their patients add traffic to the group's imaging. The model assumes a lease from opening and tests a scenario where the suite stays empty for three years. The forecast holds the group at 10.5 physician FTE in Year 1, 11.5 in Year 2 and 12.0 from Year 3, with APPs rising from 5.0 to 6.0. Productivity dips to 95% in Year 1 for the move and recovers to 100% by Year 3.

Site and technical

A 19,200 gross sq ft, two-storey clinic needs about 1.75 acres with roughly 90 parking stalls (5.0 per 1,000 rsf), arterial frontage, an elevator, service access for supplies and medical waste, and structural and shielding allowances for imaging. Zoning is a city matter: Minnesota law lets each municipality regulate the location, height, width and bulk of buildings and the uses of land [5]. Clinics are commonly permitted or conditional uses in office and commercial districts. Minnesota's 60-day rule treats a zoning request as approved if the agency fails to deny it within 60 days, and an extension may not exceed 60 days unless the applicant agrees [6], which bounds the entitlement period. The Department of Labor and Industry states that the 2020 Minnesota Building Code remains in effect and that the 2024 Minnesota Commercial Energy Code took effect January 5, 2024 [7]. Sewer and water connection charges are carried in the permits line.

The group's space program drives the building size:

At 800 rsf per provider the plan is efficient but not tight. SBA counts common areas toward rentable property and excludes stairways, elevators and mechanical space [1], so the group's 80% share is measured on 18,000 rsf, not 19,200.

The model assumes a guaranteed maximum price contract, which most banks require before closing an interim loan. The sublease suite is delivered as a warm shell with a $75 per rsf allowance paid in sponsor cash, because the costs of interior finishing of space to be leased to another business are not eligible 504 project costs [8].

Development budget and sources and uses

Development budget

Contingency of 6.0% sits within the 10% of construction cost that 504 rules allow [8]. Interim interest is financed, so the model iterates it to a fixed point.

Sources and uses

The 50/35/15 split is not the standard 504 split. SOP 50 10 8.1 lists "hospitals, surgery centers, urgent care centers, and other health or medical facilities" as examples of limited or special purpose property and requires the CDC to address the question in its credit memorandum and explain its conclusion [1]. Such projects need at least 15% borrower cash, a debenture of no more than 35% and a bank share of at least 50% [8], [1]. A medical building with a conventional shell could be argued general purpose, allowing 50/40/10:

The lower-equity version saves about $507,000 of cash, adds about $46,000 of annual debt service and cuts Year 1 coverage to 0.96x. The 15% structure is the better credit either way.

Revenue build and operating expenses

Payer mix and reimbursement as operating-company risk

CMS published its CY 2027 physician fee schedule final rule on November 9, 2026, effective January 1, 2027 [9]. The model assumes Medicare rates fall 1.68% in 2027 and then rise 0.25% a year; a lender should re-run the model with the final rule's conversion factor. Minnesota adds a 1.8% MinnesotaCare provider tax on a health care provider's gross revenues [10]; payments for services under the Medicare program, including payments from Medicare Advantage organizations, are exempt, and payments for services not covered by Medicare are taxable [11]. The model taxes all non-Medicare revenue, $147,420 in the base year.

The weighted rate index rises 0.82% in 2027 and 1.5% to 1.7% a year after, against cost inflation of 3.0% to 3.25%. Physician pay moves with collections; staff and overhead do not. The base case therefore squeezes EBITDAR from 10.8% of revenue in Year 3 to 7.2% in Year 10, about 0.5 points a year. That is what happens to a practice when reimbursement grows slower than wages, and it is why coverage must be tested annually.

Revenue build, year by year

All figures are modelled.

Operating expense structure

Physician compensation is 57.5% of physician professional collections plus 20% of ancillary income, about $536,500 per physician including benefits. That is the market-rate normalization a CDC applies when SBA directs it to make prudent adjustments to cash flow for items such as distributions and ownership compensation [1]. Anything physicians draw above that is a distribution, which a covenant can hold back.

Assumptions

Ramp to stabilization

Year 1 carries the move, 95% productivity, a $150,000 relocation cost and a full year of new debt service: global cash flow of $961,967 against $958,395, a 1.00x DSCR. Added physicians and imaging lift coverage to 1.53x in Year 2 and 1.72x in Year 3.

10-year pro forma

Global cash flow is the group's EBITDAR, less the whole building's operating cost, plus sublease rent and recoveries, less one-time costs. SBA caps the EPC's rent at the loan payment plus the EPC's direct costs of holding the property [8], so the consolidated view is the right one for coverage.

Debt structure

The 504 debenture

The debenture follows the SOP pricing method: net debenture plus the guaranty fee, funding fee, CDC processing fee and closing costs, divided by 0.996 to add the 0.4% fee charged when a 20- or 25-year debenture is sold, and rounded up to the next even thousand [1]; the fee rates match the regulation's 0.5% guaranty fee, funding fee of up to 0.25% and processing fee of up to 1.5% [8]. The 6.95% rate is an assumption: debentures are priced at sale against long-term government debt, usually the 10-year Treasury [1], which yielded 4.88% on November 9, 2026 [12].

Bank loan, equity and guaranties

The bank loan's 7.00% is an assumption, not a quote; prime was 7.00% in November 2026 [12]. The loan must run at least 10 years beside a 20- or 25-year debenture [1], with no early call and no open-ended advances [8]. Because project cost is at least $5,000,000, SBA requires the bank's credit memorandum or credit analysis with the CDC's submission [1]. The 15.0% borrower contribution is measured on project cost excluding administrative costs [8]; it may be borrowed only if subordinate to the third-party loan and the debenture and, without SBA's written approval, not repaid faster than the 504 loan [1]. Any individual owning 20% or more must give an unlimited full guaranty, and if no one owns 20%, at least one owner must give a full unconditional guaranty [1]. A ten-physician group with equal shares has no 20% owner, so a lender would ask for guaranties from several physicians on an $8.8 million debt package.

The job test is easy to meet but must be documented. At one job opportunity per $95,000 guaranteed by SBA [1], the $3,671,000 gross debenture requires 39 jobs created or retained. An established group with 10 physicians, 5 APPs and the support staff its cost base implies retains well over that number, and the move adds two physicians, an APP and their staff.

DSCR and break-even

Historical test

SOP 50 10 8.1 requires historical debt service coverage of at least 1.15, measured as EBITDA over combined post-transaction debt service [1]. With eliminated rent added back, the base year covers 1.44x; after charging the net cost of the new building, 1.33x. The credit qualifies on history; the projections test the transition.

Year 1 liquidity

Year 1 is 1.00x. Where projections show repayment in Year 2 but not Year 1, the SOP requires sufficient liquidity to cover the Year 1 shortfall [1]. Lifting Year 1 to 1.15x takes $140,187. A reserve of six months of real estate debt service, $371,378, covers that more than twice and absorbs a delayed move.

Break-even

At stabilization the group can lose 8.5% of revenue before coverage hits 1.00x, or about 8.6 physician FTE out of 12 with APPs held constant. Losing three physicians without replacement consumes the margin. By Year 7 the cushion narrows to 6.3% as the base-case squeeze builds.

Returns

The EPC's rent is capped [8], so this study tests whether owning beats leasing by imputing an assumed market rent to the group's space: $37.00 triple net.

The building opens worth 84.4% of its all-in cost. Imputed income of $596,880 covers real estate debt service of $742,757 only 0.80x, so owning costs the physicians $145,877 more than leasing in Year 1, a gap that closes by Year 9. Debt at about 7% costs more than the 5.70% initial yield, so borrowing lowers the equity return (5.39%) below the unlevered return (6.65%); the 1.85x multiple arrives only on sale. For the physicians, ownership buys control and protection from rent resets, not early cash. For the lender, it confirms the building cannot be the repayment source.

The annual levered line shows what the physicians actually feel: a contribution of $145,877 in Year 1 above what market rent would cost, shrinking each year as rent grows against fixed debt service, and a single large inflow at sale. A practice that might sell the building or relocate within five years would do better leasing.

Re-leasing value as collateral

The first mortgage starts at 57.7% of value, but combined debt is 99.2%. If the group fails in Year 3 and other medical users take the space at $30.00, recovery covers the bank loan 1.03x and leaves the debenture a 4.1% recovery. Marketed as general office at the model's assumed $17.00 rent and 75% occupancy, recovery covers 0.20x of the bank loan. Design that keeps the building medical-ready is therefore collateral protection: a divisible floor plate, plumbing and electrical capacity for three suites, and reusable imaging shielding.

Sensitivity and stress tests

The grid shows modelled stabilized Year 3 coverage. A 150 basis point rise in both rates costs about 0.16x; a 5% revenue shortfall costs about 0.42x. Both loans are fixed, so rate risk ends at closing except for the bank reset in Year 10; revenue risk lasts the life of the loan.

S1 cuts Medicare and Medicare Advantage rates a further 5% in 2029 and holds commercial increases to 1.0% for 2028 to 2030. S2 loses two physicians in Year 3 and replaces them over two years. S3 removes 25% of Medical Assistance and MinnesotaCare revenue from Year 1. S4 raises staff cost 8% from Year 2. S5 leaves the sublease suite empty for three years. None includes a management response.

The single shocks are survivable. S1 and S4 hold at 1.15x or better from Year 2 through Year 9 before ending near 1.04x, because their damage compounds. S2 is the sharpest sudden hit, cutting Year 3 coverage from 1.72x to 1.29x before replacements ramp. S3 and S5 bite mainly in Year 1, which the reserve covers. The combined case S6 fails in every year. Restoring 1.15x in Year 3 would need $321,997, about $26,833 per physician or 4.9% of modelled physician pay. Physicians in a stressed practice routinely accept cuts of that size, which is why a distribution covenant is worth more to this credit than extra equity.

Program compliance

Risks and mitigants

Model determination

The modelled project is financeable as structured under SBA 504, with conditions. Historical cash flow, with physicians paid at market and eliminated rent added back, covers post-transaction debt service at 1.44x, above the 1.15x floor. Stabilized coverage is 1.72x and stays above 1.6x through Year 10. The project meets the new-construction occupancy rule and uses the 15% contribution SBA requires for medical facilities treated as special purpose.

The weaknesses are equally clear. Year 1 is 1.00x, so liquidity must be funded. The building opens below cost, leaving the debenture undersecured. And the margin of safety is counted in physicians.

A lender reviewing a file like this would likely require:

1. A reserve of $371,378, six months of real estate debt service, released after two years at or above 1.25x.
2. An annual global DSCR covenant of 1.25x, with distributions limited when it fails.
3. Physician employment agreements of at least five years and a funded buy-sell agreement.
4. Key-person life and disability insurance assigned to the lenders.
5. Personal guaranties from several physician owners, not only the minimum.
6. A guaranteed maximum price contract, a payment and performance bond and an overrun guaranty.
7. A signed sublease or letter of intent, with suite improvements funded outside loan proceeds.
8. An appraisal with as-complete and dark values, a Phase I environmental assessment, and a survey of local medical office comparables.
9. A design review confirming the floor plate can be divided into three medical suites.

With those terms the credit rests where 504 intends: on a durable operating business, with the building as secondary support and the debenture absorbing the collateral gap that keeps conventional lenders out.