Non-QM Is Back (and It’s Not 2008)
- Ren Reed
- Feb 11
- 5 min read
Updated: Feb 12
Risk & Roll Podcast — Episode 3
Non-QM is having a moment again. If your inbox looks anything like Nathan’s, it’s basically “NON-QM!!” in 48-point font, three times a day.
So Episode 3 is a grounded conversation about what that actually means in 2026: what non-QM is (and isn’t), why it’s growing, what the data says, and where the real risk shows up—credit, legal exposure, or operational sloppiness.
Listen to the full episode or skim the summary below:
Non-QM in plain English: two layers of rules, one big misconception
Greg framed this cleanly:
There are two layers in Reg Z land:
Ability-to-Repay (ATR): You must determine the borrower can repay. Always. No exceptions.
QM vs. Non-QM: If you’re inside QM definitions, you’ve got the “shield.” If you’re outside, you’re taking on more legal risk—even if the loan is still sound.
That distinction matters because the industry sometimes treats non-QM like it’s a different universe. It’s not. It’s still lending. It’s still risk. It’s still your file, your process, and your exposure.
“This feels too easy” — the originator brain on non-QM (Nathan)
Nathan shared the most honest non-QM reaction you’ll hear from a QM-native originator:
I did one non-QM in the last three years and it scared me to death.
Not because the borrower was sketchy—because the borrower was reputable and known in the community. The discomfort was the flexibility itself.
If you’ve lived inside Fannie/Freddie/FHA/VA boxes long enough, the minute something feels “open,” your brain starts searching for the hidden trap door.
Nathan compared it to old-school relationship lending—“community credit” vibes—where character and context matter. That’s not inherently bad. But it does require discipline to avoid drifting into “we’re just vibes-based underwriting now.”
Bob’s take: non-QM makes sense… but how will people cheat it?
Bob’s perspective is sympathetic to the spirit of non-QM. He told an originator story that’s basically the perfect use-case: high-wealth borrower, real collateral, strong ability to pay… but the file doesn’t fit the neat boxes.
That’s the argument for non-QM.
The argument against complacency is this:
If it’s flexible, someone will try to work the system.
Bob called out the kinds of games that show up when products allow more angles:
DSCR loans labeled as “rental” when someone intends to live there
misrepresentations that are hard to spot unless you’re looking for them
“it technically qualifies” files that don’t feel operationally honest
Non-QM isn’t the problem. Sloppy origination practices are the problem. And those tend to surface when volume grows and people get hungry.
Ray’s framing: non-QM was born out of Dodd-Frank — and legal risk is the catch
Ray zoomed out to the post-2008 origin story:
Non-QM evolved out of the Dodd-Frank era shift where ATR became a legal liability, not just a common-sense underwriting principle.
Before, the market lived through “no doc / low doc / don’t verify much.” Then overnight, ATR turned into something that could get you sued if you got it wrong.
Ray’s point: the industry didn’t jump into non-QM quickly because nobody wanted to be a pioneer in a product category with uncertain legal edges.
Over time, it’s gotten more settled. Lenders learned what “good” looks like in practice (bank statements, time windows, documentation standards), and the market got more comfortable.
Still, Ray dropped an important caution flag:
Non-QM is not a universal solution for homeownership access. It may help specific segments (especially successful self-employed borrowers who don’t fit W-2 documentation). But it’s not a magical replacement for broader affordability and credit access tools.
In other words: don’t confuse product flexibility with systemic affordability.
Dana’s stance: creativity doesn’t automatically mean “riskier”
Dana was the strongest “pro non-QM” voice in this episode—and her argument wasn’t hype. It was practical.
Non-QM is growing because it serves a population that often looks strong on credit performance but fails QM documentation logic.
Her framing was sharp:
Either you’re an exceptional business person and can’t get a loan… or you’re a terrible business person and you can.
Non-QM helps correct that mismatch.
Dana also made a useful comparison that most lenders don’t like to say out loud:
If you’re choosing between:
a 580 FHA loan stretched to a high DTI, and
an 80% non-QM loan with a 720+ borrower, qualified via bank statements…
…from a pure credit profile standpoint, that non-QM borrower can look stronger.
The point isn’t to dunk on FHA. The point is: risk isn’t always where people think it is.
Dana also called out what we’re all seeing in real time: the market is moving faster on this than mortgage usually does. Non-QM isn’t just an add-on anymore. In many shops, it’s becoming a real growth lane.
The data: non-QM has climbed back to ~9% and may go higher (Greg)
Here’s where Greg brought the receipts.
Using Polygon’s modeling on HMDA loan-level data (and the evolving definition mechanics), Greg shared a high-level trend:
Non-QM share fell to a low point around the 2021 transition (DTI emphasis to rate-spread/pricing emphasis)
Then it climbed back:
~6.9% (2022)
~8.5% (2023)
~9% (2024)
And everyone is now watching what the 2025 HMDA release shows
He also explained that non-QM classification isn’t just “one thing.” In the data, you see different pathways:
private securitizer / “other purchaser” routes
feature-driven reasons (balloon, interest-only, etc.)
pricing/rate-spread dynamics
“multiple reasons” buckets that stack traits together
The bigger point: you can quantify this category, but you have to respect what’s under the hood—or you’ll misread the trend.
The bet: how big is non-QM going to be in 2025?
This is where the episode gets fun: everyone calls their shot.
Nathan: 16% (based on pure on-the-ground “non-QM everywhere” vibe)
Bob: 18% (beat the spread, why not)
Dana: 14–14.5% (and expects bigger growth into 2026)
Greg: 11% (data guy did data guy things)
Ray: lawyer’d around the number… then landed on “over” with Greg
When the new HMDA drops, somebody’s going home with a weird shelf artifact. That’s called “risk management.”
Who’s doing non-QM? It’s not just niche players
One of the more telling late-episode moments: when the conversation turned to the leaderboard.
Non-QM isn’t only “specialty lenders” anymore. You’re seeing traditional lenders expanding into it as a product lane—because the revenue opportunity is real and the capital markets appetite is there.
That shift matters. When mainstream lenders embrace a category, it’s a signal that the product is becoming normalized (for better and for worse).
Quick recap: what to do with this episode
If you’re a lender deciding how to think about non-QM right now:
Separate credit risk from legal risk. Non-QM can be good credit and still higher legal exposure.
Don’t treat flexibility as permission to get sloppy. “How will people cheat this?” is a necessary question.
Use data to sanity-check the narrative. Vibes are not underwriting.
Be honest about what non-QM solves. It’s powerful for certain borrower profiles—not a universal affordability fix.
Watch where the capital is flowing. Secondary market appetite drives product availability.
Watch / listen + resources
Featured guests:
Sign up to receive more Risk & Roll episodes and content!
Comments