In the mortgage industry, no word carries more stigma than subprime. For anyone who lived through 2008, it brings up memories of falling home values, rising foreclosures and a financial system that nearly came apart at the seams. Pinning that same reputation on non-QM lending is a mistake the industry can no longer afford.
In the years since the crisis, the industry has unfairly lumped non-QM borrowers in with subprime borrowers. That misconception is costing the industry more than it realizes and doing real damage to borrowers who deserve better from the system designed to serve them.
To understand why, it helps to remember what subprime actually was, because it was not one thing. It started with a legitimate purpose. Subprime and its close cousin Alt-A were built for borrowers with real income who struggled to document it through conventional means. Those early products carried reasonable discipline and served people who were genuinely creditworthy but poorly served by a system built around W-2s and tax returns.
Then the guardrails came off. Lenders started offering stated-income loans, where borrowers could write down whatever income they wanted, and no one checked. Then came NINJA loans, which stood for No Income, No Job, No Assets, where the borrower did not need to prove anything at all. By the mid-2000s, the original idea had been buried under products handed to people with no realistic shot at repaying them.
Non-QM was built to meet the legitimate need to serve borrowers who did not fit the agency credit box, but it operates under an entirely different set of rules. The borrowers it serves are not people who cannot afford their homes. They are people whose financial lives do not cleanly translate into government forms.
According to Nomura’s 2026 Securitized Products Outlook, non-QM loans originated in the second quarter of 2025 carry the highest credit scores the sector has ever recorded. Since the fourth quarter of 2022, lenders have reduced the share of low-FICO and high-LTV loans by approximately 50%.
The credit box has tightened over time, not widened, despite the sector’s nearly nine-year existence. The trajectory runs in the exact opposite direction of what the market took heading into 2008.
The workforce changed, but the underwriting didn’t
Agency lending was designed for salaried employees with a W-2, buying a modest home, planning to pay it off over thirty years and staying put. That model made sense when it was drawn. That now describes a shrinking share of Americans’ actual earnings.
Estimates of the self-employed population range from 16.5 million to more than 27 million, depending on how you measure it, according to the Center for American Progress. All measures agree that these are not marginal earners.
Self-employed workers generate $156,000 in annual income on average, compared to $123,000 for traditional wage earners, according to St. Louis Federal Reserve data. The people the conventional mortgage system most often turns away are, in many cases, among the most financially capable people in the market.
The problem with conventional underwriting is that it reads a tax return and makes a decision. For many self-employed borrowers, that tax return won’t tell the whole story. For example, business owners who write off legitimate expenses, pay themselves through a pass-through entity or have income that fluctuates seasonally can appear to be a credit risk on paper, even when they are anything but.
Non-QM underwriting gives lenders room to ask different questions. Instead of “What does your tax return say?” it asks, “What does your actual financial situation look like?” A borrower with 12 months of consistent bank deposits and 30% down is not a risk. They are a paperwork problem, and there is a big difference between the two.
Regulatory environment prevents a subprime repeat
The subprime collapse was about more than bad loans. It was about what happened to those loans the moment they were made. Lenders were packaging mortgages and selling them off within days, meaning they had no financial incentive to care whether the borrower ever made a payment. That was somebody else’s problem now. When you remove that accountability entirely, you get a machine that actively rewards lenders for making these bad loans.
Dodd-Frank addressed this directly. Under the Credit Risk Retention Rule, finalized by six federal agencies in 2014, sponsors of mortgage-backed securities must retain at least 5% of the credit risk of the assets they securitize. They cannot offload or hedge that position during a specified holding period.
The rule exists to make sure the people packaging and selling mortgage securities have real money on the line if those securities fail. The originate-to-distribute model that triggered the subprime collapse now carries a mandatory cost.
The Ability-to-Repay framework goes even further. Every lender, including non-QM lenders, must verify that the borrower can afford the loan before closing. We are talking about a full review of income, assets, employment, monthly debt obligations and credit history. Non-QM lenders go through the same exercise.
The difference is that instead of a W-2 and a tax return, they might use 12 months of bank statements or an asset schedule. The documentation looks different. The standard does not.
And here is the part of the argument the industry rarely makes loudly enough. A QM lender gets a legal safe harbor if a loan goes sideways. A non-QM lender gets no such protection. When a non-QM loan defaults and enters litigation, the lender must prove that the underwriting was sound. The system punishes bad non-QM underwriting in a way the subprime market of 2005 never did.
Non-QM volumes rising
The numbers tell the story pretty clearly. According to Nomura’s 2026 Securitized Products Outlook, non-QM origination volumes are on track to reach $150 billion in 2025, up from $50 billion in 2022, with room to reach $180 billion if rates decline. Non-QM hit a record 8% share of total mortgage origination volume in July 2025, up from around 5% the year before.
More originators are offering it, more issuers are bringing deals to market and the institutional investors coming into the secondary market, insurance companies, pension funds and private credit funds are not the kind of buyers who take on risk without doing their homework.
There is one risk worth being straight about. Nomura points out that as volumes grow and more lenders pile in, there will be pressure to loosen credit standards. That has happened before, and it could happen again. The rules put in place after 2008 make it harder than it was, but rules only work if the people following them actually care about why they exist. The subprime era did not fail because there were no guidelines. It failed because the people running the machine stopped asking whether any of it made sense.
Non-QM lending is not a concession to risk. It is the mortgage industry updating its picture of who a creditworthy borrower actually is in an economy where income is increasingly non-linear, non-traditional and no longer tied to a single employer sending out W-2s every January.
The subprime era failed because lenders stopped asking whether borrowers could repay. Non-QM is defined, legally and practically, by the requirement that they must. Those are not the same thing. They were never the same thing.
Victor Kuznetsov, Managing Director, Imperial Fund Asset Management
This column does not necessarily reflect the opinion of HousingWire’s editorial department and its owners. To contact the editor responsible for this piece: zeb@hwmedia.com.
