HousingWireHousingWire
AI-based tools are now substantially reducing both the time and labor of routine mortgage processing. Traditional procedural steps—such as employment proofing, income verification, fraud detection and title checks—have long resulted in an unnecessarily lengthy time for a mortgage application’s approval. But the potential benefits of the AI revolution in mortgage lending aren’t just speculation; they are already producing meaningful results.
Underwriting software applies machine learning algorithms to rapidly process applicant data, discern patterns, compare to current market conditions, identify red flags and quantify risks. Because these algorithms avoid the human errors that can be made by loan processors, title agents and appraisers, the result is more accurate and higher-quality approval decisions.
Additionally, by working 24/7, AI-based underwriting also dramatically reduces the time it takes to process an application: from the industry’s current average timeline of 30-to-45 days, AI-based methods have reduced this to a mere eight minutes. Good-bye to mounds of paperwork, sluggish turnaround times, and idiosyncratic approval decisions, and hello to a speedier, more routinized and efficient process that is to the benefit of both lenders and borrowers.
But the benefits of reduced costs and risks go beyond the lender to include both the economy and society. By enabling the expansion of scale efficiencies, the overhead costs of mortgage origination are reduced, with both the savings and benefits being ultimately passed on to homeowners. Reduced risk at the individual application level multiplicatively results in reduced risk at both the firm and industry levels.
And reduced risk in the mortgage lending industry naturally reduces the exposure of the entire economy to a downturn in the housing sector—like what happened in the Housing Bubble and subsequent Great Recession of 2008. And, achieving the twin goals of minimized risk with maximized lending promotes a greater rate of homeownership, for which the benefits go beyond just the economic to the social.
Critics may express concern about the substitution of human decision-making with technology-driven methods. But there are benefits to be considered for this transition as well. Unlike individual underwriters and loan officers, algorithms don’t become tired, make arbitrary mistakes, get distracted or have the same biases (whether explicit or implicit) that humans are prone to.
Human judgment is susceptible to both error and idiosyncratic variation. But since the decision-making process is standardized across applicants, AI is considerably less vulnerable to the accusations of discrimination that have plagued the lending industry for decades. Say hello to expanded homeownership and goodbye to redlining and bias.
To those that are still skeptical, these desirable outcomes are not just pure speculation: they are already happening. As an example, just look to the Rocket Companies (parent of lending giant Rocket Mortgage).
In recent years, Rocket has transformed the way in which consumers buy and refinance homes, beginning with the very visible way in which homebuyers apply for a mortgage. Just last year, Rocket rolled out a “groundbreaking AI-powered platform” called Rocket Logic that reduces closing times by 25%, as reported by Forbes. This improvement in productivity saw Rocket experience a larger increase in loan originations than its nearest industry competitor UWM Holdings (the biggest mortgage lender in the U.S.)
But this gain isn’t just limited to Rocket’s bottom line, but extends to Rocket’s employees as well. Armed with algorithm-based tools, Rocket’s employees have increased their compensation and benefits in addition to their productivity. The median annual compensation for Rocket employees is currently over $90,000, compared to just $77,000 at UWM.
Additionally, Rocket provides a wide range of supports to help their team members advance – including college tuition, training through LinkedIn Learning, and an internal mobility program. This allows Rocket to focus on retaining higher-skilled (and higher-paid) employees, rather than relying on low-wage, low-skilled workers that can be easily rotated in and out as the economy and housing cycle fluctuate.
This achievement prompted Rocket’s CFO to assert that they are “delivering tangible, transformative results across our organization” that have “unlocked over one million team member hours through AI automation with benefits that continue to compound”. It’s no wonder that Rocket continually makes Fortune’s list of 100 Best Companies to Work For.
Ultimately, embracing AI in mortgage lending isn’t about adapting, it’s about advancing.
Mortgage lending will no longer be confined to the narrow scope of clerks, clipboards and calculators. It will instead evolve into a faster, fairer, and more efficient system—and one that better serves not just lenders and borrowers, but the public as well. The positive change that AI will bring won’t just be large, it will also be equitably and democratically shared. For anyone navigating the complex journey to homeownership—regardless of their political leanings—that’s a change worth celebrating.
Kevin C. Gillen, Ph.D., is an economist who holds positions as both a Senior Research Fellow with the Lindy Institute for Urban Innovation and an Adjunct Professor of Finance at Drexel University. 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.
