Credit for residential construction loans has been tightening for four and a half years straight, and loan officers have been managing that reality deal by deal. NAHB’s second-quarter 2026 AD&C Financing Survey puts real numbers behind it, and for the first time, lenders and builders are describing two different markets.
Eighteen quarters and counting
The number itself isn’t a surprise to anyone who’s been underwriting these loans. NAHB’s quarterly AD&C Financing Survey has now shown tightening credit conditions for eighteen straight quarters, with the net easing index landing at -12.0 for the second quarter of 2026. That streak goes back to early 2022, which means most loan officers active in this market have never worked a quarter where conditions were actually loosening.
What’s new is who agrees with whom. The Fed’s Senior Loan Officer Survey asks lenders the same question NAHB asks builders, and this quarter lenders reported a net easing index of +3.7, essentially the opposite read. According to Eye on Housing’s writeup of the survey data, that’s the first time since NAHB started tracking both surveys side by side in 2013 that lenders and builders have landed on opposite sides.
For a loan officer at a bank, that gap is worth knowing. It means the caution builders are running into at the closing table isn’t necessarily showing up in how their own institution describes its posture. The tightening is real, even where the official read says otherwise, and that’s useful to know the next time a builder pushes back on terms.
What “tightening” actually looked like this quarter
The survey doesn’t stop at whether credit got harder. It asks how. Among builders who said conditions worsened in Q2, 53% pointed to lenders requiring personal guarantees or collateral outside the project itself, the single most common answer. Raising interest rates, cutting loan-to-value or loan-to-cost ratios, and turning down relationship loans tied for second, each cited by 47%.
That first number is worth knowing for anyone structuring these deals. A personal guarantee or outside collateral is a reasonable tool to reach for under this much pressure, and it does protect the institution’s paper. But it protects the paper more than it protects the project. It shifts where a loss lands if something goes wrong. It doesn’t make the build itself any less likely to go over budget or miss a milestone.
Pricing moved too, in ways that are easy to miss if the only thing being tracked is the headline rate. Contract rates rose on two of the four loan categories the survey tracks, land acquisition went from 7.42% to 7.77%, land development from 7.27% to 8.09%. But effective rates, which fold in points, climbed on all four categories, more than 0.6 percentage points above where they sat at the end of 2025. Land acquisition’s effective rate moved from 9.36% to 10.43%. Land development jumped from 10.15% to 12.59%. A lender can hold the contract rate flat and still tighten hard through points, and it shows up on the closing statement more than the term sheet.
Where these tools fall short
Every one of those four levers has a place, and none of them is unreasonable given how long this tightening has run. But none of them touch the actual construction risk on the loan. A guarantee doesn’t catch a draw going out over-funded. A wider point spread doesn’t flag a missing lien waiver before the money moves. The risk that actually sinks a construction loan lives inside the build, and that’s a harder thing to manage from a term sheet alone.
There’s also a real cost to leaning too hard on the borrower right now, and it’s one that shows up on a pipeline report more than a risk report. Builders who get asked for outside collateral, or turned down for a relationship loan, have somewhere else to go. Private lenders have spent the past few years building their businesses around exactly those builders. For a bank or credit union working to grow this book, that’s worth knowing before the next renewal conversation, not after a good builder has already moved on.
Where risk reduction actually comes from
There’s a way to bring the real risk down without asking more of the builder, and it doesn’t touch a single term sheet. It comes from how closely the draw itself gets managed once the loan is already funded.
Checking every draw against a line-item budget instead of a loan-level total catches an overrun while it’s still small, not two reconciliation cycles later. Verifying lien waivers as part of draw approval, rather than chasing them down after the fact, closes that exposure before it ever opens. None of this changes what a loan officer negotiates at closing. It changes what happens to the loan after that, which is usually where the actual losses start.
That’s the model CoFi Blueprint runs. Blueprint pairs purpose-built technology with hands-on construction expertise to run the program with you, backed by more than 100 years of combined construction lending experience across the team reviewing every draw.
Red Canoe Credit Union runs its construction portfolio this way. They’ve funded more than $63 million across 110-plus projects with zero foreclosures and a one-day median draw approval, without adding a single person to keep up with the growth. For the loan officers originating those loans, that means the deals they bring in are backed by a process built to catch problems early, not just built to close fast.
Growing the book without adding more risk
At CoFi Blueprint, the position is simple—you shouldn’t have to choose growth over risk. None of the trends in this survey are likely to reverse soon, not after eighteen straight quarters of tightening, and that’s exactly when the temptation to keep pushing more of the risk onto builders is highest. The alternative is putting the discipline somewhere it actually reduces risk, inside the draw process, where a loan officer’s origination work gets protected instead of second-guessed six months into the build.
Talk to the Blueprint team about what a fully managed draw process could catch, and how much more room it gives a loan officer to grow their book with confidence.
