Build-to-Rent Development: Yield on Cost, Debt, and Exit

Build-to-rent development is underwritten the way an apartment deal is underwritten, not the way a spec house is. You throw land, horizontal costs, hard construction, engineering, brokerage and team into one bucket, project the stabilized rent roll against that total, and the resulting yield on cost has to clear the exit cap rate by enough to pay you for five years of risk. Banks generally want to see that yield above 7, though some project types have been financed closer to 6.

Andy McMullen of Legacy Acquisitions builds 100- to 300-home rental communities in Lafayette, Louisiana, Foley, Alabama, and across the Carolinas and Texas. He walked through the hurdle rates he underwrites to, two very different capital stacks, and what actually happens at the five-year mark when the debt comes due.

Below: the deal profile, the yield-on-cost math, a 75% bank loan versus an 88% loan-to-cost life company partner that takes 40% of the deal, the construction-to-exit timeline, and the operating and entitlement realities most first-time sponsors underestimate.

From the Real Estate Pros Show

This article draws on an interview with Andy McMullen of Legacy Acquisitions on the Real Estate Pros Show, hosted by Dylan Silver.

What a Build-to-Rent Deal Actually Looks Like

The unit of production is a community, not a house. McMullen’s projects run 100 to 300 single-family rental homes on roughly 30 acres, at about eight to nine homes per acre. The work is entitlement, approval, infrastructure, streets, roads and utilities first, then vertical construction, then lease-up and stabilization, then sale.

Geographically he concentrates in the South: two projects in Lafayette, Louisiana, one in Foley, Alabama, plus activity in the Carolinas and Texas. Lower land basis in those markets is a deliberate underwriting choice, not a preference — it buys room in the cost bucket.

The mental model he uses is “pulled apart apartments.” The homes are detached, but the business is multifamily: one owner, one management platform, one lease-up curve, one exit. That framing is why Fannie, FHA and other lenders that once treated this product as novel now underwrite it with familiar tools.

The sector is also smaller than the headlines suggest. Despite the recurring story about institutions buying up everything, McMullen puts build-to-rent at roughly three to five percent of rentals. What has changed is who is in it. American Homes 4 Rent, Lennar and D.R. Horton all run a BTR division alongside their for-sale business.

The origin of the model matters for how you read the numbers. After the financial crisis, operators were buying scattered single-family homes and trying to manage them as portfolios. Around 2015 and 2016 the obvious improvement emerged: build them next to each other, manage them like an apartment community, and stabilize faster because you can lease a home the day it is finished.

Underwriting to Yield on Cost, Not Price per Door

One number decides the deal. McMullen’s method: put every dollar in a single bucket — land, hard construction costs, engineering, brokerage commissions, the team — then project the stabilized cash stream at the end of lease-up and measure it against that total. That ratio is your yield on cost.

The threshold is straightforward. Banks want to see yield on cost somewhere above 7. On certain project types it has been financed down toward 6. What makes the deal work is not the absolute number but the gap between it and where cap rates are trading when you exit.

That gap is the development lift. If you build to a 7 and the market prices stabilized BTR communities at a cap rate meaningfully below that, the delta is the value you created by taking entitlement, construction and lease-up risk. Compress the spread and the project becomes five years of work for a bond-like return.

Two levers protect the spread:

• Land basis. Land is expensive in many markets. Southern secondary markets let McMullen buy at a lower basis, which lowers the denominator and lifts yield on cost before a single home is framed.
• Contingency. How much cushion you can carry inside the cost bucket determines whether a materials or insurance surprise eats the lift. Insurance, fuel and lumber costs have all moved against builders.

This is the real difference from build-to-sell underwriting. A for-sale builder solves for margin per unit and absorption pace. A BTR sponsor solves for a stabilized income stream against total project cost, which is an apartment developer’s problem applied to detached product.

A bank wants to see yield on cost somewhere above seven. Now you’re above where cap rates are, and you can use that delta between yield on cost and the cap rate — that’s what we call the development lift.

— Andy McMullen, Legacy Acquisitions

The Capital Stack: 75% Bank Debt vs. an 88% LTC Life-Co Partner

The conventional path is a construction lender at 75% to 80% loan to cost, with the balance raised from individual investors and funds. In the current market that equity requirement has run as high as 30% to 40% on some deals, which is the real barrier to entry in build-to-rent financing — not the complexity, the check size.

The alternative McMullen described changes the arithmetic entirely. A life insurance company will lend at around 5.5% when market rates are sitting at 6.5% to 7%, and will go up to 88% loan to cost. The price is ownership: they want 40% of the deal.

Run the trade. Your equity raise drops from roughly 30% to about 12%, your cost of debt drops a full point or more, and the capital you would have tied up in one project now supports several. You give up 40% of a bigger pile of deals instead of keeping 100% of one.

Timing determines whether that trade is smart:

• Makes sense at the start, structured as a programmatic partnership across multiple projects.
• Usually does not make sense mid-project, because by then you have collateral in the ground you can borrow against to keep building.

The obstacle for newer sponsors is access, not terms. A life company is not taking a meeting with a first-time developer. McMullen solved it the way most people do — by bringing in a senior partner who already had the lending relationships or the balance sheet to satisfy the lender. Relationships and track record are the part of this business that no AI underwriting stack can produce for you.

The Five-Year Capital Cycle and How You Exit

Budget five years from closing on land to sale. The sequence McMullen runs is roughly three years of construction, about a year of marketing, and another year of lease-up. Value creation and IRR peak in years four and five, and in a clean scenario that is exactly when you sell.

That timeline is not a coincidence — it maps to the way the debt is structured. Investors who have watched multifamily syndications hit a five-year maturity and face a refinance-or-sell decision will recognize the pattern. The difference in BTR is that a stabilized, newly built rental community at peak NOI is a cleaner thing to sell than a repositioned 1980s apartment building.

There are two exit paths:

1. Sell the whole community as one group. One buyer, one closing, institutional pricing on a stabilized income stream.
2. Peel off tranches. If the homes are individually platted — some of McMullen’s projects are, some are not — you can sell 10, 20 or 30 at a time and retain the rest.

The tranche option sounds flexible and sometimes is, but understand what you are selling. A buyer of 10 homes inside a 100-home community you still control is subject to the decisions the majority owner makes about management, capital projects and community operations. That friction is real and it shows up in pricing.

Whether to individually plat is therefore a decision to make at entitlement, years before it matters. Plat individually and you preserve optionality on the exit. Skip it and you have committed to a single-buyer sale.

Operating a Rental Community of Detached Houses

Management looks like apartment management with the walls moved. You need the same economies of scale, the same leasing and marketing teams, the same collections process. Lawn care replaces hallway cleaning. McMullen is explicit that the operating playbook is the multifamily playbook.

The economics run better in the early years for two reasons. The product is new, so maintenance is light. And residents stay longer, so you turn fewer units. Anyone who has operated apartments knows turns are what eat the bottom line — vacancy, make-ready, leasing cost, all at once. Fewer turns is a direct margin improvement.

The caveat: those lower expenses are a short-term condition. New product eventually ages, and roofs, HVAC and exteriors on 200 detached homes are a different capital plan than one building. Underwrite the out-years accordingly.

Stickiness is the operational advantage worth engineering. Residents stay because they have a backyard, because they see their neighbors, because they walk the dog trail and know people. McMullen’s communities run programming to reinforce it — events, Christmas and Halloween decorating contests, and a community resource person he describes as a chaplain, though the role is about gathering people rather than running Bible studies.

The useful insight is where that programming comes from. Outsourced community-life vendors underperform compared to empowering residents who are already natural connectors — the people who host neighbors anyway. Support them with a modest budget and pizza. The payoff, per McMullen’s data, is average tenancy of three to five years instead of one.

Investors think community programming comes out of somebody’s profit. It usually grows the pie.

Demand, Markets, and Getting a BTR Project Approved

McMullen’s site selection thesis is secondary and tertiary markets with balanced, steady growth — not the headline metros. His reasoning: within Austin and Dallas there are ascending and descending pockets, and saying you are “in Dallas” tells you nothing about whether your corner of it is working.

Lafayette is the example. The economy shifted from oil to oil and gas, then added healthcare and tech, with major space-sector investment about 45 minutes away. That job diversification is what supports rent growth over a five-year hold, and it is the variable to underwrite before you underwrite anything else.

The resident is usually a family that wants new construction with a yard and walking trails. Some projects carry Class A apartment amenities — clubhouse, pool. Others are simpler: a three-bed, two-bath plan repeated with different elevations. Notably, part of this resident base could buy and chooses not to. With mortgage rates where they are and roughly half of existing owners locked into 2s and 3s, these renters would rather put their capital somewhere else.

Entitlement is where deals die. Expect council hearings and neighborhood meetings, and expect the perception question. On a Broussard, Louisiana project, the councilman whose own home sat next to the site asked McMullen directly whether he was bringing in Section 8.

What wins those rooms is the physical product. A gated community with landscaping, green space, walking trails and people visibly out and about reads very differently to a neighboring homeowner than a dense apartment building. Bring renderings that show the open feel of the site plan, because that openness is the argument.

Frequently asked questions

What yield on cost do lenders want to see on a build-to-rent project?

Banks generally want yield on cost above 7, according to Andy McMullen of Legacy Acquisitions, though it has been financed down toward 6 on certain project types. Yield on cost is calculated by putting all project costs in one bucket — land, hard construction, engineering, brokerage, team — and measuring the projected stabilized income stream against that total.

The threshold alone is not the test. What matters is the spread between your yield on cost and the cap rate the stabilized community will trade at when you exit. That spread is the development profit.

How much equity do you need to raise for a build-to-rent development?

With a conventional construction lender at 75% to 80% loan to cost, you are raising 20% to 30% of total project cost from investors and funds, and in some current deals that has run as high as 40%.

A life insurance company lending up to 88% loan to cost can cut the raise to roughly 12%, but will typically want around 40% of the deal in exchange. That trade generally makes sense when structured programmatically at the outset across multiple projects, not partway through a single one.

How long does it take to build and stabilize a build-to-rent community?

About five years. McMullen’s cycle runs roughly three years of construction, about a year of marketing, and another year of lease-up to stabilization.

Value creation and IRR peak around years four and five, which is why that window is the typical sale point — and why it lines up with five-year debt maturities.

Can you sell build-to-rent homes individually instead of as a portfolio?

Only if the homes are individually platted, which is a decision made at entitlement. When they are, you can sell in tranches of 10, 20 or 30 homes while retaining the rest.

The tradeoff is management friction. A buyer of a small tranche inside a community you still control is subject to your decisions on management and operations, which narrows the buyer pool and affects pricing.

How is build-to-rent different from build-to-sell for a homebuilder?

The underwriting target changes. Build-to-sell solves for margin per unit and absorption pace, and increasingly requires embedding rate buydowns in the cost of construction to move product at today’s mortgage rates. Build-to-rent solves for stabilized yield on cost against total project cost, with the exit happening years later at the community level.

The operating infrastructure also differs — a sales team versus a leasing and marketing team — but the capital commitment is the bigger distinction. Many for-sale builders, including Lennar and D.R. Horton, now run both.

The bottom line

Before you model a single rent comp, price the land and get a credible read on where stabilized BTR communities are trading in that market. If you cannot see a yield on cost that clears 7 with real contingency built in and still sits meaningfully above the exit cap, the project is five years of execution risk for no development lift — and no capital stack, however cheap, will fix that.