A 120-Slot Infant, Toddler and Preschool Center in Middlesex County, Massachusetts: An SBA 504 Model Study

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

Project summary

All figures in this table are model results or assumptions, not market facts.

The credit question

A lender reviewing a file like this does not need to be convinced that Middlesex County has young children or that its parents work. The census data settle both. The question is narrower and harder: can a center with no operating history, no state operating grant and no subsidy contract fill about 110 of 120 slots at premium prices, pay educators enough to keep its classrooms open, and do so fast enough that the second year of operations covers debt service of $412,608?

Five things have to be true.

First, the project must be eligible and correctly classed. Only for-profit businesses are eligible for SBA business loans [2]. The operating company in this model is new, so its owner must contribute at least 15% of project cost, and 20% if the building is also a limited or special purpose property [1], [2]. Child care centers do not appear on the SOP's list of example limited or special purpose properties, but the list is not all-inclusive and the CDC must address the question in its credit memorandum and explain its conclusion [2]. The base case funds 20% so the structure works under either answer.

Second, enrollment must reach about 95 full-time-equivalent children by Year 4 for coverage of 1.00x, and about 99 for 1.15x. The modelled base is 110. That 15-child margin is the core of the file.

Third, tuition must hold. The modelled infant price of $34,200 a year in 2027 is more than twice the 2025 national average price of infant care in a center, $15,636 [3], and it sits below the state cost study's 2024 estimate of the full cost of infant care in Metro-region centers [4]. A 10% tuition shortfall at base occupancy takes Year 4 coverage to 0.84x.

Fourth, wages must stay within the plan. Personnel is 59.4% of stabilized revenue. Every 1% of wage cost moves Year 4 coverage by about 0.045x, and a classroom cannot run short of staff without breaking the licensing ratios the staffing plan assumes.

Fifth, cash must last through the ramp. Year 1 loses $84,431 before debt service and $497,040 after it. SBA allows a projection-based 504 file to show repayment in Year 2 rather than Year 1 only if the borrower shows liquidity to cover the Year 1 shortfall [2].

Location and county economy

Middlesex is part of the Boston labor market north and west of the city.

Three readings matter for this asset. First, the county is growing slowly, and the growth comes from international arrivals and births, not domestic movers. In 2025 the county lost 11,269 residents to domestic migration and gained 13,193 from abroad [5]. Births have been flat at about 15,600 to 16,200 a year since 2021 [5]. A child care center therefore cannot count on a rising tide of children; it has to win share of a stable cohort.

Second, income is high but uneven by family type. The median married couple with children earns $235,971, while the median single mother with children earns $61,725 [6]. At the modelled infant price, care costs 14.5% of the first figure and 55.4% of the second. The model's private-pay price only works for the first group. The second group depends on state financial assistance, whose availability to new families this model study could not confirm (see the market analysis).

Third, the county's job base is weighted to high-paying sectors (below). That supports demand from two-earner households but makes educators harder to hire.

Employment base

Shares computed from [7].

Professional services, health care and education account for 39.2% of private employment [7]. They produce the parent profile a full-day center depends on: two working parents, a predictable weekday schedule and the income to pay private tuition. The same sectors also explain the wage problem. Average payroll per private employee in the county is $101,547 [7], while child day care payroll per employee is $36,568 (next section). A center hires from a labor pool that has better-paid options nearby.

Market analysis

Demand drivers

The labor force figure is the one that matters. In 78.1% of Middlesex homes with a child under 6, every parent present is in the labor force [6]. In two-parent homes, both parents work in 78.4% of cases. These households need non-parental care for about 50 hours a week, 50 weeks a year. Children under 3 make up 60.1% of the under-5 population in households [6], which is relevant because infant and toddler care is both the scarcest and the most expensive to provide.

Supply

County Business Patterns counts employer establishments, so it includes centers but excludes most family child care homes, which usually have no employees. Between 2019 and 2023 the number of employer child care establishments fell slightly while employment rose and payroll per employee rose 30.4% [8], [7]. The industry consolidated into fewer, larger operations and paid more for the same labor. That supports the modelled wage assumptions and warns that a new entrant will face the same cost curve.

Nationally, Child Care Aware of America found that the supply of licensed centers fell in 26 of the 43 states with complete data from 2024 to 2025, about 1% overall, while Massachusetts was one of seven states where the supply of family child care homes grew [3]. A new center in 2027 opens into a state where competing providers also draw on a large state operating grant (see below), which is a caution for a lender.

Trade-area licensed capacity could not be retrieved for this model, so the supply figures below are assumptions to be replaced with an EEC count for the actual site.

Public pre-K and the preschool book

Preschool is 79 of the 120 slots and 58.6% of gross potential tuition in the model, so public pre-K is a direct revenue risk. Cambridge already offers free preschool for all of its 4-year-olds and prioritized 3-year-olds through public and community-based providers, covering the school-day, school-year portion, with possible additional costs for full-day, full-year care depending on family income [9]. The model's site is assumed to be in a town without universal pre-K. Scenario C in the stress tests removes 20 points of preschool enrollment from Year 4 onward to show what happens if that changes.

Demand method: penetration and required capture

The appropriate method for a child care center is a penetration test: estimate the number of working-parent children in the trade area who use center care, compare it with existing licensed capacity, and measure the share the new center must capture to stabilize and to break even.

The center needs about 8.4% of local center-care demand to cover debt service and 9.7% to reach the base case, a share a new facility can win within three years. But the assumed unmet demand of 138 children barely exceeds the center's 120 slots, so it must also draw children from existing programs. Price, quality and staffing stability are therefore competitive variables, not just capacity.

Turnover matters as much as capture. About a third of enrollment, 35 to 40 children, must be replaced each year as preschoolers leave for kindergarten, against about 627 births a year in the trade area (4.0% of the county's 15,687 births [5]).

Price position

The cost-of-care study that the Massachusetts Department of Early Education and Care (EEC) contracted for is the most useful public input for a lender reviewing a Massachusetts center [4]. It estimated that the daily cost of current care in Metro-region centers was $154 for infants, $120 for toddlers and $64 for preschoolers in 2024, and that per-child costs for centers rose 14% to 26% from 2022 to 2024, with facility costs up 28% to 29% and staff compensation up 7% to 9% [4]. The modelled infant price is below that cost estimate, the toddler price is near it and the preschool price is well above it. That is the normal economics of the sector: the study notes that programs often use preschool revenue to cross-subsidize the age groups that are more expensive to serve [4]. The preschool rooms carry the margin. The study's figures are 2024 estimates, and the 250-day annualization is this model's own assumption.

Subsidy and state operating grants

Two state programs could support revenue, and the model gives limited credit to one and none to the other.

Child Care Financial Assistance (CCFA) pays providers a daily rate for eligible children. In 2024 the center CCFA rate covered only 69% to 78% of the estimated cost of current infant care, depending on region, but 125% of preschool cost in the Metro region [4]. Whether income-eligible families in the site's region can obtain new vouchers should be confirmed with EEC; this model study could not reach EEC's own pages. The model assumes 12% of enrolled children are CCFA-funded, mostly agency referrals, at rates 7% below private tuition, a deduction of 0.84% of gross tuition.

Commonwealth Cares for Children (C3) is the state's operational grant to licensed providers. The FY 2027 General Appropriations Act provides $475 million for operational grants to child care providers, distributed under the statutory formula [10], and the Senate Ways and Means summary describes the C3 program at $475 million, $375 million matched by $100 million from the Early Education and Care Operational Grant Fund [11]. Whether a center that opens in 2027 can receive C3 grants was not confirmed in any source this study could fetch. The base case therefore includes no C3 revenue. If EEC makes the grant available to the new center, it is upside to coverage, not a component of it.

Forecast

The model forecasts average enrollment of 55% of licensed capacity in Year 1, 84% in Year 2, 91% in Year 3 and 92% thereafter. A 55% first year implies opening with 35 to 40 pre-enrolled children and adding five to six a month. It is the most important assumption in the model.

Site and technical

Site requirements

The design standards in this table, other than the SBA row, are model assumptions drawn from the draft's reading of EEC's licensing regulations (606 CMR 7.00), whose official text this study could not fetch. Each must be confirmed with EEC and the local building official before design.

The site also needs a safe drop-off loop for about 120 morning arrivals, land for 9,000 sq ft of play yards, a location on parents' commuting path and a clean Phase I. A former office or retail building on about 1.3 acres meets those needs better than a downtown storefront.

Classroom plan

Ratios and group sizes are staffing assumptions taken from the draft's reading of 606 CMR 7.10; confirm them against the official regulation and with EEC.

The ratio table drives the business. Each infant room of seven needs two educators, so an infant room earns at most seven tuitions to pay two salaries. A preschool room of 20 needs the same two educators and earns 20 tuitions. That is why the room mix in the plan is weighted to preschool, and why a lender should ask whether the mix fits local demand rather than just the operator's preference. The staffing plan also assumes each room is led by an educator certified for its age group and that a second trained adult is always available.

Ratio positions cover only the core of the day. A center open from 7 a.m. to 6 p.m. runs 11 hours, while an educator works eight. The six assistant and floater positions in the staffing plan cover opening and closing shifts, breaks, planning time and absences. The model assumes rooms can be combined in the first and last hours of the day, subject to EEC's rules.

Zoning and permits

Massachusetts law limits local control over child care. Under the zoning act, no city or town zoning ordinance or bylaw may prohibit, or require a special permit for, the use of land or structures for a child care facility, defined as a child care center or school-age program; land and structures may still be subject to reasonable regulations on bulk and height, yard sizes, lot area, setbacks, open space, parking and building coverage [12]. For a lender, this removes the largest entitlement risk, a use denial, but not the practical ones. Parking and drop-off rules, site plan review where a town applies it to protected uses, and building and fire code requirements for the change of use still govern cost and schedule. EEC licensing comes last, and its steps and timing should be confirmed with EEC.

Construction approach and schedule

The conversion is interior work on an existing shell: classroom walls, child-height fixtures and diapering stations, a kitchen, HVAC upgrades, a fire alarm and sprinklers if the building official requires them.

The schedule is a model assumption. The model capitalizes nine months of interim interest in the project budget. Any delay in licensing extends that carry and pushes opening into a less favorable enrollment month. A September opening fits the enrollment cycle best.

Development budget

Development budget

All budget lines are assumptions; the land, building and conversion prices need an appraisal of the actual site and a contractor's guaranteed maximum price. The $175 per sq ft conversion cost assumes a sound shell and roof, and the 10% contingency reflects the hidden conditions typical of older buildings.

Sources and uses

All rows are model results.

The 504 program finances fixed assets, so working capital and start-up losses sit outside the project. The $525,000 reserve is sized to the Year 1 shortfall after debt service of $497,040 plus the $12,420 needed to lift Year 2 coverage from 1.12x to 1.15x, rounded up. The owner therefore needs about $1.89 million of cash before opening. That figure, not the 20% injection, is the real equity test in this file.

The 504 debenture's administrative costs are financed: the $1,769,625 net debenture grosses up to $1,822,000 after the CDC processing fee, SBA guaranty and funding fees, closing and legal costs, and the fee for selling the debenture, rounded up under the SOP's procedure [2]. Exact fee amounts should be confirmed against the CDC's fee schedule.

Revenue build and operating expenses

Assumptions

Revenue build

All rows are model results.

Gross potential tuition at 100% occupancy is $3,193,500 in Year 1. In Year 4 the center earns $28,509 of net revenue per enrolled child. By age, Year 4 enrollment is about 12.9 infants, 24.8 toddlers and 72.7 preschoolers.

Staffing and wages

All rows are model assumptions.

Massachusetts law treats a wage of less than $15.00 an hour as oppressive and unreasonable [14]. The modelled lead teacher rate of $24.00 is an illustrative assumption, set high enough for a new center to hire certified lead teachers away from established programs. The weighted classroom rate is $21.81 an hour.

The owner is modelled as the director at a market salary of $92,000, so the projections show what a replacement director would cost.

Wages are the margin. Personnel costs fall from 74.4% of revenue in Year 1 to 59.4% in Year 4, and drift back up to 61.2% by Year 10 because wages grow at 4.0% against tuition at 3.5%. A market-wide wage step would be faster; that is Scenario B.

Stabilized expense structure, Year 4

All rows are model results.

Debt service is 13.1% of stabilized revenue. For an owner-occupied center that is the equivalent of rent, and it sits in the range a leased center would pay for occupancy.

10-year pro forma

All rows are model results in dollars. EBITDA is before owner distributions and income taxes; a global cash flow analysis must also cover the owner's personal taxes and debts.

Debt structure

All rows are model results on assumed rates.

The structure follows the 504 pattern for a project that is a new business and may be a limited or special purpose property: at least 20% from the borrower, no more than 30% from the debenture and at least 50% from the third-party lender [2], [1]. If the CDC concludes that the building is not limited or special purpose, the new-business rule alone sets the floor at 15%, and the debenture may rise to 35% [2], [1]. Under that alternative, modelled equity falls to $886,762, the gross debenture rises to $2,130,000, debt service rises to $438,721 and coverage falls to 1.05x in Year 2 and 1.49x in Year 4. The 20% version is the better credit, and the difference in owner cash is about $293,000.

The bank loan needs a term of at least 10 years when paired with a 20- or 25-year debenture [2]. The 504 rate is set at debenture sale, most commonly benchmarked to the 10-year Treasury [2], which stood at 4.88% on November 9, 2026 [13]. Neither modelled rate is a quote. During construction the bank's interim loan carries the whole debt; the debenture takes out the 504 share after completion. By Year 10 the 504 balance falls to $1,422,565.

At Year 10 the bank balance is $2,351,168, which must be refinanced or reset. With Year 10 EBITDA of $761,192, the refinancing risk is modest, but it is real if rates are higher then.

DSCR and break-even

All rows are model results.

SBA's 504 standard is a debt service coverage ratio of at least 1.15, measured as EBITDA over combined debt service after the transaction [2]. For a projection-based project the CDC must analyze at least two years of projections, justify the assumptions, and, if repayment appears in Year 2 but not Year 1, show liquidity sufficient to cover the Year 1 shortfall [2]. This model passes the liquidity test with the reserve, reaches repayment in Year 2 at 1.12x, and clears 1.15x from Year 3. A credit officer will read the Year 2 figure as the weak point. The $525,000 reserve covers the Year 1 shortfall plus the $12,420 gap to 1.15x in Year 2.

All rows are model results.

The break-even figures include the staffing curve: as enrollment falls, the model reduces classroom staff, but not in proportion, because rooms cannot run below ratio. Each point of occupancy is worth about 0.05x of coverage. At stabilization, debt service is covered at 1.00x with 95 children, which leaves a cushion of about 15 enrolled children. For a center that loses a third of its preschoolers to kindergarten every September, that cushion has to be rebuilt each autumn.

Returns

All rows are model results.

The levered return counts the Year 1 shortfall of $497,040 as owner cash. A 9.5% unlevered return is thin for a start-up; the levered return works because 504 financing is long and fixed, and the owner's market salary is already in expenses.

The exit value, about 8.7 times EBITDA for business and real estate together, is the most uncertain line; a lender should treat it only as evidence that collateral holds value if the business performs.

Collateral is the other lens. At an income value of Year 4 EBITDA capitalized at 11.5%, the stabilized going-concern value is about $5.70 million, which puts total debt at about 84% of that value. An appraisal of the real estate alone, before the business is proven, may come in below cost. Child care conversions add value for child care use only, and where the collateral is special purpose property the appraiser must be experienced in the industry [2].

Sensitivity and stress tests

Grid A: Year 4 DSCR, tuition level against stabilized occupancy

Model results.

Grid B: Year 4 DSCR, wage level against bank rate

Model results.

The grids rank the risks. Price moves coverage fastest: 5% of tuition is worth about 0.37x, because tuition falls straight through to EBITDA. Occupancy is next, at about 0.05x a point. Wages follow: 10% higher wages cut Year 4 coverage from 1.59x to 1.14x. The bank rate matters least; a full two points of rate costs about 0.17x, because only half of the capital stack floats with the bank's pricing and the 504 rate is fixed at sale.

Downside scenarios

All rows are model results.

Scenario A is the most likely way the file goes wrong. A slower ramp consumes $914,538 in three years, $389,538 more than the reserve, yet the center still reaches 1.31x at 86%. That is a liquidity failure, not a viability failure, so owner liquidity beyond the reserve matters.

Scenario B shows the wage risk. An 8% market-wide wage step from Year 3, with no tuition response, takes coverage to 1.16x in Year 3 and 1.23x in Year 4. The center still pays its debt, but with little room.

Scenario C models a town-level public pre-K program, of the kind Cambridge already runs [9], removing 20 points of preschool enrollment from Year 4. Coverage falls to 1.09x. Because preschool carries the margin, losing 15 preschoolers does more damage than losing the same number of toddlers.

Scenario D is the one the structure cannot absorb. If the market will not pay the modelled premium, and tuition is 7% lower and grows only 2.5% a year, coverage never reaches 1.00x and declines over time as wages outgrow tuition. The tuition study for the real trade area is therefore the single most important piece of due diligence.

Scenario E combines three moderate risks and lands at 1.01x: the project has room for one problem at a time.

Program compliance

Risks and mitigants

Model determination

The modelled project is financeable under SBA 504 as structured, but only with conditions, and the conditions are the substance of the credit. At stabilization the center covers debt service 1.59x, earns a 20.8% EBITDA margin and needs only 95 enrolled children to cover its debt. The demand base is deep: 75,935 Middlesex children under 6 live in homes where every parent works [6], and the 504 structure keeps debt service near 13% of revenue. The weaknesses are timing and price. Year 2 coverage of 1.12x sits below the SBA 1.15 standard [2], the model gives the new center none of the state operating grant its competitors may receive, and the base case depends on tuition more than twice the national average for center infant care [3].

A lender would likely set these conditions:

1. An owner cash injection of at least 20% of project cost, regardless of the special purpose conclusion, plus $185,000 of pre-opening funds.
2. A pledged operating and debt service reserve of at least $525,000, released only after two consecutive quarters at 1.15x coverage.
3. Evidence of pre-opening enrollment, such as deposits from at least 40 families, before the interim loan funds construction.
4. A third-party tuition and capacity study of licensed centers within the trade area, confirming prices within 5% of the model and center capacity for children under 5.
5. A guaranteed maximum price construction contract, a completed Phase I, and lead testing where required, before closing.
6. Unlimited personal guaranties from every 20% owner [2], with life and disability insurance assigned.
7. Quarterly enrollment, staffing and financial reporting, with a 1.15x coverage covenant tested annually from Year 3.
8. A written plan from the operator for converting rooms between age groups if preschool demand falls.

If the tuition study shows prices 5% or more below the model, the file should be restructured with more equity or fewer infant rooms before it goes forward.