AboutInsightsToolsSign in
All insightsHousing policy
Oct 15, 202618 min

Can Build-to-Rent Make Housing More Affordable?

More rental homes. More chances to own? The answer depends on what gets built, who receives the savings and who carries the risk.

Begin the story
Architectural model of detached homes and a landscaped courtyard enclosed by one blue property boundary, with a transparent shared-plan plane behind them.

The houses are new. The promise needs testing.

A house can be new without being an addition. It can become cheaper to finance without becoming cheaper to rent. And it can give a family a front door without giving them a deed.

These are not contradictions. They are different questions hiding inside the same phrase: build-to-rent.

Imagine two households looking at the same newly built street. This is a thought experiment. One wants a house with enough room for a child and a parent, but cannot assemble a down payment. The other has been saving to buy and wants a neighborhood where an ordinary house is still for sale. A rental community might open the street to the first household while leaving the second searching elsewhere.

Neither household is mistaken about what it needs. Neither experience, by itself, tells us whether the policy behind the street has worked.

On October 14, 2026, HUD opened two FHA multifamily mortgage-insurance programs to qualifying build-to-rent communities. Detached houses, duplexes and triplexes can be part of an eligible rental property. The change takes effect immediately.

The announcement offers a route to more housing. It does not yet offer evidence of how many additional homes will result, what their rents will be or which households will benefit.

That is where the interesting story begins. Not with a choice between loving or hating landlords, but with a chain of consequences that can be followed, measured and sometimes interrupted.

Four roles.
Not one cash transfer.

The insurance supports a loan.
The tenant signs a lease.

Lender

Makes the private loan.

Developer

Builds or finances the rental property.

Rental homes

One operating property, separate buildings.

Residents

Pay rent under their leases.

FHA mortgage insuranceProtects the lender against covered losses. Not a payment to the resident.
221(d)(4)

New construction

A route for qualifying new rental communities.

223(f)

Acquisition or refinancing

Recently developed communities. A financing event need not add a home.

Unweighted schematic of roles, not a claim about the size or timing of cash flows. HUD, October 14, 2026. The specific eligibility provisions remain in the underwriting letter.

The building and the business are different things.

From the sidewalk, a detached rental house can look like a detached owner-occupied house. From a lender's desk, the arrangement can be very different.

An owner-occupant borrows against a home and repays from household resources. A rental development is evaluated as a business: its ability to collect rent, pay expenses, maintain the property and service debt. Several physically separate houses can belong to one operating property.

Under the new policy, Section 221(d)(4) covers new construction. Section 223(f) covers acquisition or refinancing of recently developed rental communities. These are two different routes, not two names for building another house. HUD's announcement distinguishes them.

The FHA insures an approved lender against covered mortgage losses. It is not handing the tenant a rent voucher, buying the houses or lending a family its down payment. HUD's program descriptions explain the insurance mechanism and long-term financing.

This distinction matters twice. First, public backing can change the terms on which private capital is willing to finance a project. Second, the party receiving that financing advantage is not necessarily the party whose monthly housing bill falls.

The advantage might allow a stalled project to proceed. It might support a lower rent, improve the owner's return, absorb a new expense or influence the price offered for land. Several of these can happen at once.

Calling the policy an affordability measure identifies its ambition. Following the money tells us whether the ambition reaches a household.

More doors to rent.
Which doors would not exist otherwise?

Which homes exist because the policy exists?

Consider a hypothetical program that finances 100 homes. At the ribbon-cutting there are 100 doors to count. The harder number is the one that cannot be photographed: how many doors would have existed without the financing change?

Suppose 40 homes would have been built anyway. The policy may improve their financing, but it cannot claim all 40 as additional supply.

Of the remaining 60, suppose the project draws enough land, workers or capital away from other construction to displace 30 homes elsewhere. The net addition is then 30, not 60 and not 100. These numbers are assumptions chosen to expose the arithmetic, not estimates of the new policy.

This is the question economists call additionality. It is the difference between supporting an activity and causing more of it to happen.

There is a serious upside here. If financing is the constraint preventing a viable project from being built, removing that constraint can produce homes that otherwise would not exist. It is just as important not to count a refinancing transaction as another completed house.

The downside is subtler than a vacant building. A subsidy can succeed at attracting participants while accomplishing little beyond changing who finances a project or who owns it. A busy program is not automatically an effective one.

The study below lets the reader move both assumptions. One controls how much would happen anyway. The other controls displacement among the remaining candidate additions. It is an accounting exercise, not a model that knows the true counterfactual.

Its purpose is to make the missing comparison impossible to forget.

A financed home is not always
an additional home.

Change the missing comparison.
Keep the arithmetic visible.

100-home reference cohort

40Would be built anywayFinancing changes, not total supply.
30Replaces other constructionDisplacement of candidate additions.
30Additional homesThe net increase in this scenario.
0%100%
0%100%

The second slider applies to the 60 remaining candidate additions, not to all 100 financed homes. Each glyph is one percentage point; category boundaries are approximate when results are fractional.

Try another hypothetical volume

30 additional home equivalents from 100 financed homes under these assumptions. Not a national forecast.

Download CSV
Assumptions, not observed policy outcomes. The reference scenario assigns 40% to homes built anyway, then displaces half of the remaining 60%, leaving 30% genuinely additional. The proportional bar carries exact modeled shares; the glyphs are a rounded visual aid. This exercise does not estimate land, labor, credit or household responses.

More supply can help without solving every household's problem.

The strongest argument for rental construction is not that every new tenant receives a bargain. It is that an additional usable home gives someone another place to live and can ease competition for the alternatives.

A household moving into a new home may leave another home behind. Another household can move into that one. The effects can travel through a market rather than remain inside the new development's fence. How far they travel depends on who moves, what they leave and which households can afford the homes in the chain.

Research offers a reason to take this mechanism seriously. A Philadelphia Fed working paper by Asquith, Mast and Reed, published in February 2020, found lower nearby rents around large new apartment buildings in the low-income urban areas it studied.

That is evidence about those settings. It is not a build-to-rent rent-discount formula. A suburban collection of detached houses can have different land use, customers, transport costs and competing supply.

Nor does a market becoming less expensive than it otherwise would have been guarantee an affordable lease for a household with very little income. A small improvement in a very large bill can still leave the bill out of reach.

Both observations belong in the same account. New supply can reduce pressure. Direct assistance, accessibility, location and household income can still determine who benefits.

The useful disagreement is therefore not whether supply matters. It is how much additional supply this particular intervention produces, where it appears and how it connects to the homes people actually need.

How large is it?

Two different measures.
Neither is a forecast of this policy.

Construction

63,000

Builder-held single-family rental starts

Previous four quarters75,000
Four quarters ending Q2 202663,000

Narrow measure: excludes homes built and sold to rental owners.

NAHB analysis of Census data · Aug 21, 2026

Ownership

Institutional share of single-family rentals

Seattle4%
Jacksonville22%
Under 1% to 3%

Institutional share of all single-family homes across six studied metros

GAO · 2024 data · selected metros
Construction starts are not completions. GAO's ownership measure is broader than build-to-rent: investors with 5,000+ homes across 5+ metros. Its rental ratios compare identified investor properties with ACS renter-occupied single-family stock; percentages are published rounded values. These selected metros are not a national sample. Download the evidence CSV.

A national percentage cannot describe one street.

The market needs two separate lenses.

For construction, NAHB's August 21 analysis of Census data counted about 63,000 builder-held single-family rental starts over the four quarters ending Q2 2026, down from 75,000 in the preceding four-quarter period. This narrow segment represented just under 7% of single-family starts. It excludes homes built and sold to another rental owner.

For ownership, GAO's March 2026 study used 2024 data for six selected metros. Its identified large investors had ratios ranging from 4% of single-family rental stock in Seattle to 22% in Jacksonville. Against all single-family homes, the range across the six metros was below 1% to 3%.

Those are different denominators and different populations. Build-to-rent is a development model. Large institutional ownership is an ownership category that also includes existing houses. A small developer can build rentals; a large landlord can buy an old house.

GAO defined the investors as owners of at least 5,000 homes across at least five metros. Its selected areas are not a representative national sample. The rental ratios compare identified investor properties with Census estimates of renter-occupied single-family stock.

The practical lesson is not that a small national share makes local concerns imaginary. It is that a national number cannot describe the choices available to someone searching in one suburb, at one price, this month.

Equally, a heavily concentrated market cannot stand in for the whole country. Scale is part of the claim, not an optional footnote after it.

Neither dataset measures the effect of the policy announced on October 14, 2026.

A lower borrowing cost is the start of a calculation.

Take an illustrative $250,000 of amortizing debt per home. Hold the principal and 35-year term constant. At 7%, principal and interest are about $1,597 a month. At 6%, they are about $1,425.

The difference is approximately $172 a month. No lender has promised that discount here. The comparison is an educational example, not an FHA rate quote.

Now ask what is missing. Insurance premiums, fees, required reserves, transaction costs and operating changes can reduce the usable advantage. The study allows an extra annual cost to be entered as a monthly equivalent. It does not pretend that every kind of reserve or fee is economically identical.

If there are no extra costs and half the saving reaches the tenant, the modeled monthly rent reduction is about $86. If none reaches the tenant, the direct reduction in this exercise is zero. If the extra cost exceeds the financing saving, there is nothing positive to divide.

The slider does not estimate how landlords will behave. It reveals how much the claimed benefit depends on an assumption about that behavior.

This is also why the comparison holds the loan term fixed. Extending amortization can lower a payment while changing the timing of principal repayment and total interest. A smaller monthly number is not, by itself, proof that financing has become cheaper in every sense.

The experiment isolates one mechanism. It leaves market rents, vacancy, construction timing, tax treatment and loan eligibility outside the calculation.

What it gives the reader is not a forecast, but a place to ask a precise question: how much of the advantage survives, and who receives it?

Cheaper finance.
Cheaper rent?

Move the savings.
See who receives them.

Net monthly financing saving$171.67
Tenant saving$85.8350% of net saving
Retained saving$85.8350% of net saving

Flow widths represent the split of positive net savings, not a measured effect on rents.

0%100%
More assumptions

Use the extra-cost field for a simple annual equivalent of premiums or other differences. It is not a complete fee or reserve valuation.

$250,000.00 debt · 35 years · $0.00 extra annual cost

Potential monthly rent reduction$85.83

Assumed pass-through, not an observed landlord decision.

Download CSV
Illustrative calculation, not a loan quote or rent forecast. Same principal and amortization; the rates and extra costs are assumptions. No vacancy, debt-service coverage, eligibility or market-rent response is modeled. The initial comparison is $250,000 at 7% versus 6% over 35 years, with 50% of positive net savings reaching tenants. Displayed values are rounded.

Sometimes the saving reaches the land before the tenant.

A development budget does not begin with a tenant. It often begins with a bid for a site.

Imagine two developers competing for the same parcel. If both expect access to better financing and both expect the same market rent, each may be able to offer more for the land. The financing improvement can support construction, but some of its value can also be absorbed before a house is built.

This is a possible economic mechanism, not a measured result of the new policy. Its importance depends on how many suitable sites exist, how quickly approvals can expand supply and how intensely developers compete for them.

Where land and approvals are scarce, cheaper capital cannot manufacture another serviced parcel overnight. Where sites are available but financing is binding, the same change may make more projects feasible.

Rental competition matters at the other end of the project. If households have credible alternatives, an owner has less freedom to keep every saving while charging the same rent. Where alternatives are scarce, a favorable loan does not require the owner to price a lease as cost plus a fixed margin.

There is nothing mysterious about this. A family's rent is negotiated in a housing market, not calculated by dividing its landlord's mortgage payment by the number of bedrooms.

For policy evaluation, the consequence is demanding. We need to examine land prices, new construction and tenant costs together. Looking at only one can mistake a transfer between businesses for a household benefit, or miss a supply benefit that emerges outside the financed development.

A better balance sheet is not the same as a better home.

For a tenant, the relevant product is not simply a building. It is a home under a particular set of terms.

Professional management can bring a staffed maintenance operation, clearer processes and the resources to replace a roof without waiting for one small owner's cash position to improve. It can also make poor practices repeatable across many households. Organization is a capacity, not a guarantee of how that capacity will be used.

There is concrete evidence that tenant costs deserve scrutiny. The FTC's Invitation Homes refund page reports sending 444,131 checks totaling more than $47.2 million for undisclosed fees and deceptive or unfair charges. The underlying conduct concerned payments between January 2021 and September 2024.

That is an enforcement case involving one company and a historical period. It is not evidence that every large landlord behaves the same way or that the new financing policy caused those practices.

GAO's May 2024 evidence review also warns that research on lease terms and tenant outcomes is limited. We should not turn that uncertainty into either a clean bill of health or a universal accusation.

The practical questions remain sharp. What is the all-in monthly bill? Which charges are mandatory? Who handles repairs, and how quickly? What happens at renewal? What alternatives exist if the relationship fails?

Our illustrative ledger starts with $2,400 of rent and adds $125 in monthly mandatory charges. The household must find $2,525, not $2,400. That arithmetic is not a claim about any company's current lease.

A policy evaluated on advertised rent alone could overlook precisely the amount that changes whether a household can stay.

The promise
needs a ledger.

A lower headline rent is not enough if the full recurring bill tells another story.

Compare like with like: all mandatory charges, responsibilities and renewal terms.

Illustrative monthly lease
Advertised rent
$2,400
Mandatory charges
$125
All-in monthly cost
$2,525
Monthly charges beyond rent:
$1,500 over a year
Illustrative amounts, not a quoted lease or a company-specific rent claim. For documented historical conduct rather than a hypothetical, see the FTC's Invitation Homes refund information. One case is not a prevalence estimate for the sector.

Access to a house and access to ownership are not the same promise.

Renting can be a rational choice. It can let a family move for work, preserve liquid savings and avoid concentrating its wealth in one building. A household should not have to buy before it can live in a house.

But a better rental option does not erase the question of ownership.

The new underwriting letter requires a cohesive rental property and prohibits parcel or building sales and partial collateral releases during the insured loan term. This is not a rent-to-own program. That restriction is not the same as saying the homes can never be sold under a different future financing arrangement.

The important comparison is still the missing alternative. Would this site otherwise have become homes for individual purchase, another rental development, a denser project or nothing at all?

If the alternative was no construction, the rental project can add access without removing an existing opportunity to buy. If the same constrained site would have supported attainable for-sale homes, the ownership trade-off deserves attention.

Counting every new rental as a house stolen from a buyer assumes the answer before examining the site. Counting every rental as an unqualified addition does the same thing in the other direction.

The household balance sheet also needs a fair comparison. An owner's mortgage payment contains principal repayment as well as interest, but ownership brings maintenance, transaction costs and price risk. A renter can invest cash not used for a down payment, but only if that cash exists and is actually saved.

The question is not whether a deed is always better than a lease. It is whether public policy expands the set of workable choices, including a realistic path between them.

A front door.
A different set of rights.

The lease

A route into a home without a purchase down payment. Rights and responsibilities depend on the lease and applicable law.

The deed

Ownership, equity exposure and the costs and risks attached to the asset. Not supplied by this rental financing program.

The letter prohibits individual parcel or building sales and partial collateral releases during the insured loan term. This is not a perpetual ban on a different future arrangement and not a rent-to-own promise.

The public promise sits behind the private loan.

Insurance changes who bears a loss if a borrower cannot repay under the contract. It does not make the underlying risk disappear.

An empty home collects no rent. Debt service, grounds maintenance and some other costs continue. If several homes lease more slowly than expected, an otherwise plausible plan can run short of cash before it reaches stable occupancy.

There are several numbers that should never be collapsed into one. The insured loan balance is the amount of debt supported by insurance. Delinquent balances identify loans experiencing payment trouble. Claims are another stage. Ultimate fiscal losses depend on premiums, recoveries, timing and other cash flows.

HUD's August 2026 commercial mortgage report recorded approximately $143.1 billion of outstanding FHA multifamily insured principal. Its 60-plus-day default balance rate was 0.22%.

Those figures describe the existing whole multifamily portfolio before this build-to-rent change. They are not a new spending authorization, a build-to-rent loss estimate or a probability that a new project will default.

The distinction protects both sides of the argument. Calling all insured principal a taxpayer loss exaggerates the cost. Treating a low historical default measure as a guarantee about a new segment understates uncertainty.

There is a legitimate public-finance case for insurance that unlocks worthwhile construction at an acceptable risk-adjusted cost. There is also a legitimate concern about public exposure supporting returns that private investors retain.

To judge the bargain, we would need the new segment's loan volume, underwriting, premiums, performance and recoveries, alongside its housing benefits. The size of the existing insurance book cannot answer those questions for us.

Insured principal
is not taxpayer loss.

Existing FHA multifamily book$143.1bn

Outstanding insured principal, August 2026. Not new build-to-rent spending.

60+ day default balance rate0.22%

Of the existing portfolio's balance. Not a BTR default probability or realized loss rate.

Insured exposurePossible claimsNet fiscal result

Premiums, recoveries and timing belong in the account. No new-program loss forecast is available here.

HUD's August 2026 Credit Risk Report, tables 1 and 3. The underlying outstanding balance is $143,051.6 million. These whole-portfolio observations predate the policy and do not measure the new segment's risk.

An approval is not an occupied home.

A financing rule can change quickly. A development cannot.

The route from an application to an occupied home still passes through land, approvals, infrastructure, construction and leasing. An application can be withdrawn. A financed project can be delayed. A completed home can wait for a tenant. Each stage measures something different.

This is why an early report of strong interest would be worth knowing, but insufficient to establish success. The public needs to see the pipeline, the attrition and the time taken between stages.

The underwriting letter contains useful signs of the risk HUD is trying to address: a market-rate vacancy assumption of at least 7% or the comparable-market rate if higher, and an initial operating-deficit escrow of at least the greater of 3% of the loan or four months of debt service. It also anticipates higher replacement reserves for the many roofs and other components. These are underwriting provisions, not forecasts of actual vacancy.

An escrow can give a project time to absorb a problem. It cannot make an uneconomic location desirable or guarantee that replacement reserves will always be enough.

The same care applies to accessibility and household fit. A home can be available at the right headline rent and still be unusable for the person who needs it, too far from work or too costly to heat and maintain.

The financing advantage is one input in this longer production process. It deserves neither the entire credit for every completed house nor immunity from scrutiny when the promised household benefit fails to appear.

A cushion.
Not a guarantee.

7%Market-rate vacancy assumption floor, or the comparable-market rate if higher.
3% / 4 monthsOperating-deficit escrow minimum: the greater of 3% of the loan or four months of debt service.
Many roofsReplacement reserves must account for the components of separate buildings.
Selected provisions from Mortgagee Letter 2026-12. These are underwriting requirements, not expected vacancy, a complete eligibility guide or a guarantee against loss.

Judge the policy by what changes outside the financing office.

The balanced conclusion is not a compromise halfway between every claim. It is a set of tests capable of proving some claims wrong.

First, count net additional occupied homes, not just applications or financed units. Separate new construction from acquisitions and refinancings. State how the alternative development path was estimated, and show the uncertainty around it.

Second, measure household costs. Track all-in rents, concessions, mandatory fees, maintenance responsibilities and renewal terms. Compare similar homes and locations rather than treating a brand-new detached house as interchangeable with an older apartment.

Third, examine choice. Does the policy widen access to suitable rental homes? Does it reduce realistic opportunities to buy in constrained locations? How concentrated is ownership in the markets and price bands where affected households search?

Fourth, publish the public-finance account. Show new insurance exposure, premiums, troubled loans, claims and recoveries over time. Keep those measures separate from each other and from the existing portfolio.

The best outcome is possible: capital makes genuinely additional homes viable, competition or enforceable commitments helps benefits reach tenants, service is good and public risk is properly priced.

A weaker outcome is also possible: much of the activity would have happened anyway, savings are absorbed elsewhere, household costs barely improve and the public carries exposure without a commensurate gain.

The policy is too new to tell us which result will dominate. That is a reason to specify the scoreboard now, before enthusiasm or disappointment selects the convenient numbers.

Return to the two households looking at the street. The family seeking a lease needs somewhere it can afford to stay. The family seeking a deed needs a route to ownership it can realistically complete.

A good housing policy does not make one family's need disappear from the picture. It asks what can be added, at what cost, and how many more people can build a life there.

The street is the beginning of the evidence. The lives it makes possible are the point.

Four promises.
Four things to measure.

01

Homes added

Net additional occupied homes, not all financed units.

02

Tenant costs

All-in rents, fees, service and renewal terms.

03

Ownership access

Workable choices to rent and buy in the same local market.

04

Public exposure

Premiums, troubled loans, claims and recoveries over time.

Sources, methods and what we do not know

Research checked October 15, 2026. The policy announcement is dated October 14. Existing-market observations predate it and cannot establish its effects. This is an explanatory investigation of a new policy, not a completed causal evaluation.

The financing study uses the standard fixed-payment annuity formula, equal loan principal and amortization, an explicitly chosen rate comparison, extra annual costs converted to a monthly equivalent and an assumed tenant share of positive net savings. It does not assess eligibility, debt-service coverage, vacancy or market-rent response. These are cash-payment comparisons, not a complete present-value or tax analysis. Rounded displayed parts can differ by a cent from their displayed total.

The additionality study subtracts homes assumed to be built anyway, then applies an assumed displacement percentage to the remaining candidate additions. The result is an accounting scenario, not an estimated behavioral model. Applying it to another volume does not create a national forecast. Outputs can be fractional home equivalents.

The two households, 100-home cohort and monthly rent ledger are illustrative. No interview, actual development, company-specific lease or measured Oftu policy outcome is invented. No independent rate quote or project-level underwriting review is claimed.

Primary sources

Read as MarkdownEvidence CSV
More insights