A Small Group of Mortgage Lenders Survived Rate Hikes—Now They Face the Double-Edged Sword of A.I.

To paraphrase the opening line of Charles Dickens’ Tale of Two Cities, right now it’s both the best of times and the worst of times for independent mortgage banks (IMBs), or non-bank lenders. On the one hand, A.I. is beginning to reshape the contours of the entire mortgage ecosystem; new technologies are creating opportunities to streamline and drive efficiencies across the origination, underwriting and servicing sectors of the industry in ways that would have been almost unimaginable just a few years back. On the other hand, this new milieu of stubbornly high post-pandemic interest rates is putting the squeeze on nearly everyone. The Federal Reserve’s rate hikes are upending the playing field, pushing many IMBs out of business while driving a wave of consolidation among those who remain and are struggling to stay afloat. 

As with any period of major tumult, there are always those who manage to navigate turbulent waters with the finesse of a seasoned surfer. In the mortgage arena, a select few major non-bank lenders are emerging as formidable players, adeptly maneuvering through this pair of industry disruptions. These entities are not only weathering the storm but are also strengthening their positions and capturing significant market share amidst the chaos.

Mortgage origination and refinancing have plummeted since rate hikes

The Fed’s decision to begin raising interest rates in 2022 for the first time in more than five years to combat post-pandemic inflation cascaded directly into consumer financial products; from credit cards to auto finance, almost no corner of credit markets has been left unscathed by the Fed policy making. And among the areas most profoundly affected by these rate increases has been the housing sector. 

To put it in perspective, the average retail mortgage rate jumped from around 3 percent in early 2020 to over 7 percent in mid-2024. According to the Mortgage Bankers Association, mortgage applications as of June decreased by nearly 15 percent year-over-year, and over 30 percent since the Fed first started raising rates to combat inflation in 2022, highlighting the significant elasticity between higher rates and market activity. Data released from federally-backed lender Freddie Mac showed that total mortgage originations more than halved from $4.8 trillion in 2021—before the first rate increase took effect—to less than $2 trillion in 2023. And the refinancing segment of the mortgage market has all but vanished, dropping from $2.8 trillion in 2021 to just $310 billion in 2023, a nearly 90 percent fall-off. 

Another impact of higher rates is an increase in the number of buyers who are foregoing mortgage financing altogether, instead opting to pay cash. As of late 2023, about one in three homebuyers were paying cash for their home purchases, a marked increase over 2022 figures. At the top end of the market, the numbers are even higher as the percentage of homebuyers who pay cash is nearing 50 percent, according to data from Redfin – levels that haven’t been seen in over a decade. 

IMBs, which at its peak represented over 60 percent of all mortgage originations, have been the most brutally impacted by all these changes in consumer behavior. Unlike traditional banks such as USBank or Citibank, IMBs can’t accept deposits and must fund mortgages through wholesale credit markets or by selling loans to investors. Given their focus on mortgage and specialization in originations, nonbanks are inherently susceptible to market fluctuations – a vulnerability that has proven especially problematic since the Fed began raising rates, which in turn has forced IMBs to pass on their increased borrowing costs to consumers. 

As early as 2023, many of these IMBs began closing up shop in droves. Since then, every month, industry trades have begun to read like obituaries, listing the names of the latest players forced to shutter. “We estimate that 38 percent of lenders have left the market since the Fed started raising rates,”

Patty Arvielo, the CEO and co-founder of New American Funding, one of the nation’s largest independent lenders, told Observer. “But, from time to time, our industry needs to clean the house. These stats [showing consolidation among lenders and others going bust] show that many of these players were only in it for the re-fi game and when that dried up, they had nothing to fall back on. It’s periods like this one that enable the strong to get stronger. It’s good for the industry,” added Arvielo, who is viewed by many as the doyenne of the U.S. mortgage firmament. 

All of this contraction among IMBs has massive downstream implications as these lenders play a critical role in minority and low-income families getting access to homeownership. In 2021, IMBs originated 73 percent of the mortgages for minority households. Additionally, nonbanks have originated over 89 percent of the FHA loans and 83 percent of VA loans in 2021. As these entities have lost volume and have not had alternative lines of business to have revenue diversification, they have been susceptible to the rapidly increasing costs to originate mortgages due to their reliance on people. 

Is A.I. a silver bullet or a distracting shiny object?

Adding to the angst of IMB executives is the rise of A.I. and other adjacent technologies that are causing their own disruptive wake that is rippling through the mortgage industry. 

More than a handful of lenders have begun to experiment with A.I. in the hopes that the savings and efficiencies can make up for their losses in market share and help them recoup some of their mojo. The advent of foundational A.I. sources such as OpenAI and Anthropic have given birth to an entire ecosystem of applied A.I. that promises to revolutionize the mortgage lending space. 

Chris Bixby, managing director at Rice Park Capital Management’s venture strategy that focuses on investing in mortgage and real estate technology, breaks down the impact that A.I will have on the sector in three areas:

• Origination:  

The cost of originating a loan is now over $12,000—up from $8,000 just a few years ago, Bixby told Observer at an A.I. in Real Estate conference in Manhattan earlier this month. “Approximately 90 percent of origination costs are still people. A.I. automation, machine learning, and other related technologies are all key opportunities to be part of an efficiency play here.” 

Although it will be some time before the human element can be completely removed from the mortgage process, Bixby believes a sizable chunk of the workflow can be done better and faster by machine-learning algorithms which should ultimately translate into cost-savings for these non-bank originators. “Over the past two years, mortgage industry employment is down by approximately 30 percent. This constant ramp up and then cut jobs cycle that we have seen over and over can be smoothed out by better technology,” he said. 

• Underwriting:

When it comes to the underwriting side of the business, Bixby is less optimistic in the short-term. He believes that A.I. will only be able to drive marginal efficiencies until government sponsored entities (GSEs) like Fannie Mae and Freddie Mac rethink and update how their underwriting guidelines – most of which were developed decades ago and have not been able to keep up with the times. 

Specifically, Bixby believes that GSE lending guidelines are behind the 8-ball in understanding and capturing the way Gen Z and soon, Gen Alpha, make money—a radical paradigm shift from the way their parents and grandparents earned an income. 

“The ‘gig economy’ is a real thing and underwriters still struggle to account for that type of earning without unduly penalizing borrowers,” remarked Bixby. “The GSEs need to reevaluate the way they underwrite a new generation of first-time borrowers purchasing a home.”

Bixby thinks that FICO scores can’t continue to be held up as the end-all and be-all for determining credit worthiness. He suggests that bank statements or total family income may be an important way to underwrite those who earn money in less traditional ways even though it does not represent any more risk than a typical W-2 earner, and that A.I. can go a long way to help fold in these other reference points into the underwriting process as long as it’s sanctioned by government underwriters. “Unless the GSEs help remove the various impediments to adoption of new technology in the mortgage ecosystem, the industry will continue to be burdened by a boom-bust cycle of hiring and firing teams of people,” he added.

• Servicing: 

Finally, when it comes to loan servicing, Bixby believes that new A.I. technologies will have a major impact on customer relationship management. “AI-powered chatbots and virtual assistants can greatly improve the customer experience and handle routine customer inquiries, providing instant responses and freeing up human staff to focus on more complex issues. These tools also help maintain engagement with clients by sending personalized recommendations and reminders,” observed Bixby. He also feels that predictive analytics – helping IMBs better predict future borrowing patterns, identify potential opportunities, and detect fraud – will enable IMBs to tailor their marketing strategies and product offerings to better meet the needs of their clients and better mitigate risk.

A brave new mortgage world

All of these new technologies come at a cost, of course. And they require time to be trained and then successfully integrated into existing workflows. “If we are being honest, so far no mortgage company has really figured out A.I. In fact, I think that there is a risk that many IMBs will rely too heavily on A.I. and create a depersonalized experience,” said Arvielo of New American Funding, the largest lender to communities of color in the U.S. “And so far I haven’t seen any A.I. that can bridge the trust gap. From my view, it has to be another human marketing that mortgage.”

The new era of high interest rates coupled with what many have called the most transformative technology ever unleashed upon humanity is reshaping the mortgage market in profound ways, particularly for independent mortgage banks. While the challenges are significant, A.I. and other technologies offer a path forward but not a silver bullet. While industry consolidation is a reality and part of an evolving competitive landscape, the ability of IMBs to navigate this new terrain will depend on their ability to adapt to these changes and leverage their unique strengths amidst rapidly blowing headwinds. 

As IMBs navigate this brave new world, they must balance the efficiency and advantages brought by A.I. with a commitment to maintaining diversity and accessibility in the market. The stakes are high, not just for the brokers themselves, but for the many aspiring homeowners whose dreams hinge on their ability to secure fair and affordable financing.