Who’s Bankrolling Non-QM?

Jeff Miller, CEO and founder of Truss Financial Group, opened his broker shop in 2015, helping real estate investors refinance out of high-rate hard money loans. Today, the company prioritizes DSCR loans, bank statement loans, and home equity products.

Although he enjoyed helping clients escape higher rates and impending balloon payments, he described those early days as “very, very frustrating.”

“In 2019 and 2020, at the trade shows, there was maybe one Non-QM company,” Miller said. “Fast forward to current industry conferences, and almost no banks are advertising conventional products — instead, virtually everyone is advertising seconds and Non-QM files.”

That’s the room filling up. And the crowd has given originators new places to find business in an otherwise muted mortgage market. Several of the nation’s top lenders and wholesale platforms, including Rocket, UWM, and Pennymac, have ventured into investor lending, equity extraction, and other alternative lending products.

A recent survey by A&D Mortgage found that 74.5% of brokers reported growing Non-QM volume, and 7.1% now operate exclusively in the channel.

Miller pointed to several possible drivers of Non-QM growth, including the channel’s proven credit performance and the overall growth of the gig economy. The composition of the market is changing, too. According to S&P, DSCR loans now account for more than half of securitized Non-QM loans, replacing bank statement mortgages as the dominant product.

So who is paying for the party?

Nonbank lenders generally rely on outside financing to fund loans before they are sold or securitized. Banks remain part of that machinery, but recent research published by the Federal Reserve indicates that a growing share of the money is coming from private credit.

What’s Private Credit?

Private credit is, as the name implies, private. It has also gone by private debt, direct lending, nonbank lending, alternative credit, and, more cryptically, shadow banking. Regulators now tend to favor the considerably less ominous “nonbank financial intermediation.” Regardless of the label, the basic idea is the same: Credit is being extended outside traditional bank lending and public debt markets, often through privately negotiated transactions that can be considerably harder for outsiders to see. 

Federal Reserve researchers have flagged that opacity as a potential financial stability concern, writing that “the lack of transparency and understanding of the interconnectedness between private credit and the rest of the financial system makes it difficult to assess the implications for systemic vulnerabilities.” 

That opacity is not new. What changed after the 2008 financial crisis was the scale of the market and the role it began to play in replacing credit that had traditionally flowed through banks. Private credit existed well before the crisis, but the upheaval that followed helped transform it from a relatively specialized corner of finance into a major alternative to traditional bank lending. 

The numbers show how dramatic that transformation became. Federal Reserve researchers found that invested capital in private-debt funds grew from roughly $129 billion in 2008 to more than $750 billion by mid-2024, excluding another roughly $278 billion of undeployed capital. Including business development companies (BDCs), the U.S. private-credit market had reached approximately $1.34 trillion by 2024.

Federal Reserve Vice Chair for Supervision Michelle Bowman said in May 2026 that the post-crisis reforms strengthened the banking system but also had an unintended consequence: some lending migrated outside regulated banks as certain loans became more expensive for banks to hold. Since 2015 alone, she said, banks’ share of U.S. corporate lending has fallen from 48% to 29%.

At the same time, much of the expertise needed to make those loans was already sitting inside Wall Street.

Lehman Brothers provides one of the clearest examples of Wall Street credit expertise migrating into private lending after the financial crisis. Joris Fletcher was a managing director and head of European private credit at Lehman Brothers Europe. After Lehman’s collapse, Fletcher became managing partner and senior portfolio manager at Neovara, the independent credit manager created in 2010 to take over Lehman’s European mezzanine business. Neovara continued managing the fully invested Lehman fund and initially planned to raise another vehicle.

A similar career migration occurred at Bear Stearns, although not through a direct fund spinout. Jason Mandinach was a vice president on Bear Stearns’ agency CMO desk within its structured-products business, giving him experience in mortgage-backed securities before Bear’s 2008 collapse. In 2010, he joined PIMCO and eventually rose to managing director and head of Alternative Credit and Private Strategies. He now leads product strategy across PIMCO’s corporate, asset-based, and other public and private credit businesses, including private lending.

But nearly two decades after the subprime crisis pushed more lending outside banks, banks are becoming deeply connected to that market again.

In October 2025, after the failures of subprime auto lender Tricolor and auto-parts supplier First Brands raised concerns about credit quality, JPMorgan Chase CEO Jamie Dimon warned, “When you see one cockroach, there are probably more.”  

Yet JPMorgan, alongside five of the largest U.S. banks, had spent years building substantial exposure to the same nonbank financial system Dimon now warns about. On the bank’s first quarter 2026 earnings call, CFO Jeremy Barnum said roughly $50 billion of JPMorgan’s $160 billion in core nonbank financial institution exposure is what the bank considers private credit, including back-leverage facilities and lending to BDCs.  

JPMorgan’s position illustrates a broader contradiction in the market: Banks have reduced some forms of direct lending while increasingly extending credit to the private funds and other nonbank lenders that have taken their place.

Large-bank commitments to private-credit vehicles climbed from roughly $8 billion in 2013 to $95 billion by the end of 2024, according to Federal Reserve researchers. Banks now provide credit lines and liquidity to private-credit funds while also entering origination partnerships, warehouse arrangements, and risk-transfer transactions with them.

Separate Fed research has described an increasingly explicit intermediation chain: banks lend to private-credit providers, which lend to mortgage companies, which then lend to borrowers. But the question of interconnectedness has returned in a new form.

It’s essentially a party where everyone is running tabs for everyone else.

The bank fronts money to the mortgage lender. The lender passes some of the loans to a private fund. The private fund looks like it now owns the risk — except it may have borrowed from a bank to buy those loans in the first place. Other investors may be holding pieces of the same exposure.

By midnight, the room is full of IOUs, but nobody has a complete ledger showing who ultimately owes whom.

That is the Fed’s problem. Regulators can see some of the transactions, but not the full chain of funding relationships between banks and nonbanks. So if the economy tanks and losses start piling up, they may not know who is truly stuck with the bill — or whether the partygoer holding the bad debt also owes money to someone else across the room.

The risk may look like it changed hands. In reality, everyone may still be on the same tab.

Follow The Yield

To understand why investors are willing to assume that risk, it helps to begin with what separates modern Non-QM mortgages from the subprime loans that collapsed before 2008.

Credit market expert Dhruv Mohindra wrote an article on the JPMorgan Asset Management website to highlight the emergence of an improved alternative lending channel, dispel skepticism, and build investor confidence in modern Non-QM securitization.

“Bottom line: pre-crisis underwriting was a great deal for borrowers. Today’s Non-QM lending: uh, not so much,” he wrote.

But Mohindra isn’t just any bond investor. He was a firsthand witness to the 2007 mortgage credit meltdown while serving as Associate Director of ABS Portfolio Management & Surveillance at Bear Stearns Asset Management.

According to SEC filings, Mohindra was part of an email exchange that captured a major turning point in the lead-up to the subprime crisis. After receiving a notice from Citi marking two mortgage securities previously valued near 99 down to 30 and 22, the fund’s founder and senior portfolio manager, Ralph Cioffi, asked him, “Dhruv what do you think?” Mohindra replied, “I think we have to live with the mark.”

As banks saw the value of the mortgages securing those lines decline, they issued margin calls or pulled warehouse financing altogether. Fed Vice Chair Donald Kohn later said those margin calls “put a number of originators out of business.”

In his 2020 article, Mohindra uses market data to show that modern Non-QM is fundamentally different from 2006 Alt-A and subprime loans. Modern Non-QM borrowers were mainly self-employed, averaged 700–750 FICO scores, carried lower LTVs, and underwent rigorous asset documentation. He also pointed out that modern securitizations feature over twice the AAA subordination levels of 2006 deals with excess spread and issuer risk retention.

If the borrower is paying far more than they would on a conventional mortgage, investors are being compensated well for taking the extra credit risk. As Mohindra so bluntly put it: “A bad deal for them (the borrower) is a good deal for us (the lender).”

The fast prepayment speeds (25%–40% CPR) on those loans made it obvious that borrowers were given “a bad deal.” Instead, it was clear borrowers would use them as temporary, short-term financing out of necessity, which also means a large share are refinancing or paying off their mortgages quickly.

The trade-off for bond investors is that they get a safer, shorter-term deal. Issuers can call the deal after two years once the remaining collateral balance falls to 30% of the original pool balance.

Capital market analysts like Whalen refer to those investors as “yield-chasers.”

Non-QM mortgages fit the bill: They produced income, carried higher yields than agency loans and were secured by residential property. As Whalen explained: “It’s not like you’re financing a credit card portfolio. There’s actually an asset somewhere that backs the loan.”

Investors — including private credit funds, private equity funds, institutional asset managers, insurance companies, and family offices — could buy the loans outright or supply lenders with capital to originate them, giving smaller mortgage companies an alternative to bank financing.

Miller’s experience shows how secondary-market appetite can eventually translate into more products and more lenders competing for borrowers.

However, Whalen believes that yield-chasing capital follows a predictable pattern. When central banks cut interest rates, those investors “reach for yield” by moving down the credit spectrum into riskier assets. “You always see crowds of people going into these asset classes in good times, and then they run out in bad times,” he said.

Whalen’s concern is that mortgage originators who rely on recent yield-seeking funds as their primary funding source face operational risk if those funds experience redemptions or liquidity issues. He described recent yield-chasing entities, like Blue Owl Capital, as players that “just kind of appeared out of nowhere because there was so much money looking for a yield.”

Mohindra also brought up that point in his piece, saying they meet with all issuers to discern between those who aspire to be in the mortgage business longer term, versus those who are looking to exploit higher rates while they last.

For lenders, Whalen advised seeking out legacy institutional buyers, such as PIMCO, which has been buying whole mortgage loans for over 30 years across multiple economic and credit cycles.

The First Stress Test

The staying power of that capital mattered because Mohindra ended his 2020 article with a direct challenge to comparisons between modern Non-QM and pre-crisis subprime lending: “Zero. That’s the total amount of losses ever incurred by Non-QM bondholders to date.”

But the onset of COVID-19 immediately stress-tested his thesis. The Non-QM funding and securitization markets seized up; DBRS reported that issuance halted, delinquencies jumped, and lenders, including Angel Oak and Deephaven, curtailed funding and cut staff.

When the smoke cleared, DBRS reported that actual bond losses remained very low during the first half of 2020. Later, S&P’s review of 2020–21 Non-QM deals found many classes benefiting from low or zero accumulated losses, deleveraging and rising credit enhancement, leading to numerous upgrades.

Even though Vista Point outsourced day-to-day loan servicing to a third-party company, CEO and Founder of Vista Point Financial Holdings Bill Ashmore said his company supervised the loss mitigation strategy for several hundred million dollars of Non-QM loans originated around the onset of COVID-19.

“We were the best performing Non-QM assets on our securizations in that 2020 era than anybody out there — other than one or two other firms,” Ashmore said. “Reason being is that we only had discipline in our pricing and our credit, but also, afterwards, in terms of managing those assets.”

Overall, Ashmore said private credit is too broad a category to make a single judgment about whether it is safe or risky. A fund holding residential mortgages presents a different risk profile than one making unsecured corporate loans or equity investments in technology businesses.

“You better do your homework,” Ashmore said. “What are they investing in? Do you believe in their thesis, in terms of what their assets are investing in? Do you believe that they’re going to be a good manager of those assets after they originate them?”

He also cautioned that some sources of capital can become less dependable when perceptions of risk change. “When they see that there’s risk, they sometimes leave the building, and there may not be any fire,” Ashmore said.

He points to life insurance companies as the gold standard of permanent capital. As the U.S. population ages, retirees are purchasing record numbers of life insurance annuities to secure their retirement income. When an insurer sells an annuity, it incurs a long-term financial obligation that requires predictable, steady interest payouts over decades. To back these annuity liabilities, insurance companies must buy long-duration, cash-generating assets.

“They’re investing with companies like me, so that I will go out and manage those assets,” Ashmore said. “I will acquire them, price them, and manage them, and they will get the benefit of the coupon in that particular pool of loans. So, for me, this is a very big thing that will continue to fuel the market for the foreseeable future.”

But more liquidity can’t replace credit discipline. Ashmore warns other lenders that rely on private credit to maintain strict pricing and credit discipline rather than simply chasing short-term volume or high-yield coupons. His company, for example, maintains conservative portfolio parameters, holding an average borrower FICO score of 735 to 740 and a weighted average LTV of 70% or less.

“We’ve been through crises before, whether it be, you know, a great financial crisis or whether it be COVID,” he said. “As I said, I’m not happy about my default ratio back in 2008–2009, but I didn’t go out of business.”

Reading The Risk

That credit discipline is being tested most visibly in DSCR, the product now driving much of Non-QM’s growth. Bank of America Securities reported in June that DSCR and investor loans make up about 50% of all Non-QM collateral and 30% of all Non-QM securitization issuance. BofA analysts project a 46% increase in Non-QM issuance to $100 billion by the end of the year.

As the A&D survey shows, more brokers are exclusively committing to Non-QM or DSCR lending, such as Fernando Corona, CEO and founder of Remote Lender. Corona reported that as of August he’s closed 71 loans totaling over $25 million to date this year, mainly as a solo producer supported by a four to five person remote team.

“I truly believe the reason why loan officers are stuck with like two to four loans a month is because they’re presenting themselves as ‘I can do any loan,’” Corona said. Ironically, he said that expanding his product range would limit his business. “It’s going to create more operational drag [and] slow me down. I’ll provide a worse service to the client because my back-end team doesn’t know how to support it as well.”

By hyper-focusing exclusively on DSCR, Corona was able to standardize his entire operation so every file followed the exact same process that remaining dedicated to his investor lending niche will allow him to build a more streamlined, scalable business.

“In order for growth to happen, the problems that I’m solving need to be fairly similar. And if they’re too different, then my fulfillment and operations ends up being too different, and so I’m not allowed to scale,” he said.

But as Non-QM and DSCR lending continue to surge, fraud risks tend to increase as well. Because DSCR relies on property value and rental income instead of W-2 forms and other personal documentation, bad actors have targeted the space.

For example, a major fraud ring in Baltimore colluded with an appraisal management company (AMC) to artificially inflate appraisals in gentrifying neighborhoods and extract cash-out equity before defaulting.

Brokers also reported significant loan fallout when real estate agents’ optimistic rent projections failed to match the official appraiser’s Form 1007 rent schedule, forcing borrowers to bring more cash to the table or take higher interest rates.

Vista Point Chief Revenue Officer Allison Ashmore said she believes the Baltimore fraud ring was the biggest problem to disrupt DSCR. She confirmed that Vista Point and Brokers Advantage company did not face any exposure to the fraudulent loans. “I think that, because of that bad apple that spoiled the bunch, it made everyone want to be like ‘hey, let me dig into these DSCR assets.’”

Investors began looking beyond the information presented on a data tape, she said, asking whether short-term rentals were legally permitted, whether projected rents were supportable, whether leases backed the income being represented and whether the collateral itself made sense.

“Our leadership team was at Impact Mortgage, which was an alternative lender that survived the crisis. So from our perspective, we always have our loss mitigation hats on,” Ashmore said. “So we have always been very, very particular about the collateral that we either lend on or the collateral that we acquire.”

In addition to reviewing the appraisal, she recommends that an originator working outside a familiar market pull up satellite imagery or Google Maps and examine the surrounding neighborhood. That way, she said, the originator can spot any obvious collateral issue before a borrower moves further into a transaction.

Ashmore applies the same caution to first-time real estate investors. A borrower qualifying for a DSCR loan does not necessarily mean the property will be a successful investment. She said investors should consider potential repairs, insurance costs, cash reserves, local short-term-rental restrictions, and whether the property could still work if the original rental strategy changes.

“It depends on your market, and it depends on how much cushion you have saved,” she said.

There’s also a difference in who has authority over the credit decision, Ashmore said. For a broker, it can create delays when a lender has to stop and ask an outside investor whether a particular collateral or underwriting exception is acceptable. Lenders with greater control over their own credit box, she argues, can give originators more certainty about whether a loan will actually close.

That becomes particularly important as conventional lenders move further into Non-QM, including Rocket, UWM, Pennymac, and others.

Large lenders may have highly efficient systems built around government or agency mortgage production, but Ashmore said more complicated Non-QM files can require a different kind of flexibility. Multi-entity bank statements, unusual investment properties, and other non-standard scenarios may not move cleanly through a process designed around uniform milestones.

She argues that product availability alone does not erase the experience required to handle exceptions. As more lenders enter the market, Ashmore believes execution — knowing where a file can break and having the authority to make the credit decision — becomes part of the competitive distinction.

Further advancements in technology could help narrow that gap, but Ashmore said lenders cannot simply place an AI tool on top of an existing mortgage workflow and consider the process automated.

Ashmore argues that established Non-QM lenders have an advantage over newer entrants because they have spent years learning where complicated loans tend to break. That experience, she said, gives longtime lenders more history with unusual income structures, collateral, and borrower scenarios, helping them distinguish between an acceptable exception and a risk they are unwilling to take.

Where DSCR Could Break

What experienced lenders see as risks within individual files, Whalen sees accumulating into broader credit, valuation, and liquidity risks as the DSCR market expands.

ATTOM reported in the spring that rental yields were declining year-over-year in 54.8% of the counties studied. The dip in profitability came despite rent increases outpacing home price increases in 55% of counties studied.

As property prices remain elevated, rental yields are compressing, making it harder for properties to naturally meet standard 1.0x or 1.25x coverage thresholds. To keep deals moving, Whalen pointed out that market participants are resorting to “no-ratio” loan options or requiring larger down payments to force the property cash-flow math to work.

U.S. HomeLife Mortgage rolled out its new “no-ratio” financing for real estate investors who can’t meet the standard DSCR ratio. Scenarios may include purchases before rents are in place, cash-out refinances to fund improvements, and longer-term financing to replace hard money loans. The program offers up to 75% LTV, potential structures reaching 90% CLTV, no rental income verification, and faster closings.

Earlier in 2026, Newfi Executive Vice President of Sales John Wise flagged a rise in delinquencies among DSCR loans with ratios below 1.0, where rental income fails to fully cover the mortgage payment.

A more recent report from Fitch Ratings shows that mortgage delinquencies are rising for vintages between 2023 and 2025. While the broader housing market is seeing an uptick in late payments, newer Non-QM loan vintages are deteriorating at a much faster rate than older debt, signaling a shift in credit quality.

Research from Bank of America Securities shows that larger loan balances are increasingly entering Non-QM pools, and higher balances consistently correlate with higher delinquency rates and faster prepayment speeds.

Despite these increases, Fitch states actual losses remain minimal relative to its default expectations for the asset class.

A drop in interest rates triggers quick refinancing waves, with Fitch reporting Non-QM Constant Prepayment Rates topping 17–19%, allowing troubled borrowers to exit or modify loans before liquidation occurs.

Because current market momentum relies heavily on tight credit spreads and heavy secondary market appetite, Whalen claims that DSCR will be the first subset within the Non-QM sector to suffer a loss of market liquidity when credit conditions tighten or economic stress returns

Both Ashmore and Whalen are cautious of how today’s private-credit market will behave in a substantially worse environment.

Ashmore said the rapid post-COVID expansion has not yet experienced every kind of credit cycle. “It’s too early to tell,” he said.

So who owns the risk? It depends on where the loan sits when trouble arrives. The originator may carry it before the loan is sold, a warehouse lender may be exposed through its credit line, and private funds, insurers, or bond investors may ultimately hold the mortgage. Because banks also finance many of those private-credit firms, some exposure can travel back toward the regulated banking system. Non-QM has emerged from past disruptions with limited bond losses, but today’s larger, DSCR-heavy market has not yet endured a full credit cycle at this scale. The risk has not disappeared. It has been divided, financed, and passed around, and the next downturn will reveal who is still holding the tab.