The Cost of the Sticks Is the Wrong Lever

Let's do a little mind experiment.

A builder I know is about to put a spec house up on a lot she just bought outside Kingston. Sixteen hundred square feet, mid-range finish, the kind of house that lists around $600,000 when it's done. Her hard construction cost pencils to about $360,000. The rest is land, soft costs, financing, and whatever margin survives to the end.

Sitting in front of her are three people, each holding a different lever.

A prefab rep says he can shave 10 percent off the build if she goes modular. A mortgage broker across town is watching the 30-year rate tick down and thinks it might keep going. And her town's planning board is the wild card, the one nobody can price, where entitlement might take three months or it might take nine.

She has finite attention. So do you, if you build. The question I want to actually answer here, with real numbers, is which of those levers is worth chasing. Because the industry spends almost all of its breath on the first one, and I've come to think that's close to backwards.

We can run the math on all three.

Lever one: cut the cost to build

Start with the cost, because everything else gets measured against it. If she trims 10 percent off her $360,000 hard cost, she saves $36,000. Real money, banked once, permanently. A more honest prefab number, and I'll get to why, is closer to 5 percent, or $18,000.

Hold onto those two figures. $36,000 in the best case, $18,000 is the realistic one. Now watch how little either one does for the person who actually buys the house, and how impossible it is to reach that prize any other way.

Lever two: make the construction loan cheaper

The instinct, if you can't cut the cost of building, is to cut the cost of borrowing to build. In 2026 a small spec builder is paying somewhere around 8.5 percent on a construction loan (that's the current market for spec work, running a point or so over standard - Buildermuse and NewSilver both peg the 2026 average in the low-to-mid 8s).

Here's the fact almost nobody sits with: that loan barely costs anything in total interest, because it draws down as the house goes up. You're not carrying the full balance for the full year. On average you're carrying about half of it.

So the whole interest bill, over a 12-month build, is:

rate x loan x time x average-outstanding fraction = 8.5% x $360,000 x 1 year x 0.55 = $16,830

That's it. Sixteen thousand, eight hundred dollars is the entire construction-loan interest bill for the project. Which means even if her bank handed her the loan at zero percent, she'd save $16,830 - less than half of what the 10 percent cost cut delivers, and still shy of the modest 5 percent cut.

I find that genuinely clarifying. If you wanted to match the cost cut purely by lowering the construction rate, how far would the rate have to fall? The math is clean, because the loan roughly equals the cost. To match a 10 percent cut you'd need to drop the rate 18.2 points, from 8.5 percent to negative 9.7. To match a 5 percent cut, 9.1 points, to negative 0.6.

Both answers are impossible. There are only 8.5 points of rate that exist to cut, and matching even the modest case would take you past zero into negative territory. You cannot buy your way to a cheaper house with cheaper construction debt. The carry is too thin a slice to matter.

Lever three: the buyer's mortgage rate

Now flip to the person who lives with the number for thirty years: the buyer.

Say they put 20 percent down on the $600,000 home. That's a $480,000 mortgage. At today's 6.5 percent (Freddie Mac had the 30-year fixed in the mid-6s the week of July 2-8), it runs $3,034 a month and racks up $612,000 in interest over the life of the loan.

Two ways to help that buyer. The builder can cut the cost to build and pass it through to the price. Or someone can leave the price alone and move the mortgage rate - you might think that's impossible, but it's not.

What changes

Buyer's monthly payment

Saved vs. base

Base: $600k at 6.5%

$3,034

-

Pass through a 5% build-cost cut (price to $582k)

$2,943

$91/mo

Pass through a 10% build-cost cut (price to $564k)

$2,852

$182/mo

Mortgage rate down half a point (to 6.0%)

$2,878

$156/mo

Mortgage rate down a full point (to 5.5%)

$2,725

$309/mo

Read the bottom two rows against the top two. A half-point dip in the mortgage rate beats a full 5 percent build-cost cut on the monthly payment, and saves roughly three times as much over the life of the loan. A full point beats even the aggressive 10 percent cost cut by 1.7 times on the payment and about 3 times on lifetime interest.

Why so lopsided? Because the rate works on the entire price across 360 payments, while the cost cut is a one-time haircut on one slice of the price. It holds no matter what the house costs. The single biggest driver of whether that home is affordable is a number no builder, and no factory, controls.

Lever four: the calendar

Here's the build-side lever that actually pays.

Every month the project runs, the meter runs with it. Construction-loan carry. The GC's general conditions and overhead, call it $3,000 a month for a job this size. And the invisible one, rate and market risk over the hold, which you don't feel until it bites.

Shave six months off the combined entitlement-and-build clock and the construction carry drops from $16,830 to about $8,400, and six months of general conditions at $3,000 is another $18,000. Total, before you count anything hard to price: about $26,400.

Twenty-six thousand dollars. That's the equivalent of a 7.3 percent cut to her hard cost, nearly the whole aggressive prefab dream, and it's the one lever on this list a small builder can actually move. Not by working faster, but by getting to a confident go or no-go sooner and moving cleanly through zoning and feasibility instead of stalling out. And what she can build is set town by town, not statewide - the dimensional standards in Hurley don't match the rules next door in Marbletown, so pinning down what her exact parcel allows early is most of the battle. And that number leaves out the two biggest wins entirely: less exposure to rate swings over the hold, and getting her capital back to start the next deal.

So, does prefab actually save money?

I want to be straight about this, because being straight about it is the whole point.

We are not anti-modular / prefab. Our first build was prefab, and it was great - but it wasn't cheaper. Maybe at scale, but at small levels the savings are low to negative. I'd love a cheaper, faster way to build, and I look at every one that comes along. But the advertised savings deserve a real audit, so here's what the evidence actually says.

The marketing runs hot: modular vendors and some industry bodies cite 10 to 30 percent. Those are gross factory-labor numbers, the savings at the factory door, before the house gets anywhere near the site. The real gross saving on a typical single-family home, from bulk buying and less waste, is more like $15,000 to $30,000. Then reality pulls it back. Trucking the modules, five to fifteen thousand. The crane and set day, three to ten. The steel each box carries so it survives the trip, which a stick-built house never needs. The factory's own margin. And the parts that don't change at all: the foundation is still poured on site, and the well, septic, grading, and driveway on a rural Hudson Valley lot run the same $50,000 to $150,000 no matter how the walls get framed.

Net it out and the savings mostly evaporate. McKinsey, who has studied this more seriously than the brochures, is blunt about it: there's "often no track record of consistent, game-changing cost savings," and for one-off, low-volume work a 2 to 15 percent price premium is common. The labor savings only reliably beat the offsets at scale - dozens of identical modules, good transport corridors, jurisdictions with mature modular code. A custom spec home on a rural lot hits none of those. The honest net for our builder is somewhere between a small premium and a modest saving, clustering around 0 to 5 percent, and realistically a wash on a true one-off.

Which is exactly why I'm always suspicious about advertised savings, especially nearing 10%, and also why I'm not here to bury prefab. Because where prefab genuinely and consistently wins is speed. McKinsey again: 30 to 50 percent faster, on the order of 3 to 5 months off the schedule. And look back at lever four. Five months off is worth around $22,000 in carry and overhead alone.

So prefab's real gift was never the cost, but the time it saves. It's a timeline play, and we should be talking about it that way.

Where to actually spend your attention

Line the levers up by how much they move affordability and the ranking is clear: the buyer's mortgage rate, then the timeline, then the hard cost, and a distant last, the construction-loan rate. It holds harder at prefab's realistic 5 percent than at the advertised 10.

Here's the version that makes it concrete. Take our builder and give her the two levers that actually pay. Compress the timeline by six months, and help her buyer to a rate a point better than they'd have gotten on their own (a builder-paid buydown is a common tool at the closing table). Do both and the buyer's payment drops about $429 a month - roughly $154,000 over the life of the loan.

Now put her next to the builder who ignored all of that and chased the factory instead, and got the realistic 5 percent off the sticks. That builder handed his buyer $91 a month, and pocketed most of the saving himself because in a supply-constrained market the price doesn't move to match his lower cost anyway.

Same house. $429 a month versus $91. Almost five times the affordability, from the two levers the industry barely mentions.

This is why Offsite is built the way it is. We don't lead with a cheaper way to build, because a cheaper way to build, on its own, mostly just widens the developer's margin. We built Aldo to win the calendar - to read a parcel's zoning, feasibility, and build cost in a day instead of a month of consultant fees, so you reach that confident go or no-go before you've sunk real money into a maybe. Speed to conviction is not a soft benefit. It is, dollar for dollar, one of the two most valuable things you can do on the build side of a home.

Cheaper sticks are the weakest lever you can pull. The calendar and the buyer's financing are the strongest. Spend your attention accordingly.

A note on the math, because this audience checks it: $600k finished price, $360k hard cost (60 percent of price, mid-range Hudson Valley at ~$330/sqft on ~1,600 sqft), 12-month build, 0.55 average-outstanding draw fraction, 80 percent LTV, 6.5 percent buyer mortgage, 8.5 percent construction loan. The construction carry is rate x loan x time x fraction; buyer payments are standard 30-year amortization. Full sources and the reproducible model are in the research memo.