AI’s Threat To Mortgage Jobs: Automation’s Impact On Your Future

What’s Ahead?

All that begs the question, what lies ahead for the mortgage business? Will, for example, the processor position eventually merge with loan officer duties? Who else’s job in the mortgage production food chain can be incorporated into that of another’s? Or extinguished completely and forever?

There’s already enough software floating around out there so that some jobs can be eliminated altogether. The originator, if he is so inclined, can pre-approve the borrower at the point of sale, obtain the verifications of employment and incomes at the push of a button, and request an appraisal or an automated valuation simply by lifting that finger again. Ditto for ordering proof of the borrower’s homeowners and flood insurance and contacting the title agency. All by using automation, the entire package can be deposited on the underwriter’s desk or, more accurately, on his computer screen.

More specifically — but not to single out any one program as better or worse than any others that do the same — lenders can pre-approve a borrower using Blue Sage’s digital platform, snag the VOE and VOI with TRUV, order and receive the appraisal through Opeteon, a consumer-direct portal, and nail an electronic closing, perhaps even remotely, with Snapdocs. Then, the last human standing, the underwriter, could turn to LoanLogics for quality assurance before sending it to the closer and the settlement table.

You could perform all these services on an a la carte basis, or you could use a platform like Blue Sage’s to send an order through in its entirety, from start to closing. Either way, you name the task, and there’s an AI application for it. If there’s not, there likely soon will be.

There are so many automated solutions out there that they’re a dime a dozen. Well, not exactly a dime, but you get the point. Plunk has an AI-powered analytics platform for real estate investors. Bonzo offers a relationship management program that reduces the time loan agents spend communicating with prospects. Valigent has a data collection suite that supports both Fannie Mae and Freddie Mac’s requirements.

ReadyPrice has a shop, lock, and deliver loan exchange platform. There’s even technology from Talk’uments that provides compliant transactional clarity on mortgage disclosure documents in four different languages — Spanish, Chinese, Korean, Vietnamese, and Tagalog — for borrowers with limited proficiency in English.

The list goes on and on and on. Which ones will survive is anyone’s guess. But they and many others are out there, and they have to be at least considered. That what’s members of the National Association of Realtors did at their annual confab in California in November, where realty agents and brokers discussed what lies ahead for their portion of the production line. Mortgage needs to do the same.

No One Job In Danger

Let’s start by thinking about who’s job is destined for the cutting room floor. I put this question and others to a handful of professionals at the Mortgage Bankers Association’s annual convention in Philadelphia this fall, and almost to a person, they agreed: No one’s job is in any sort of danger. Eventually, perhaps, some positions could become extinct. But for now and the immediate future, AI won’t eliminate anyone. Rather, it will allow the same number of people to be more efficient and to do more with less.

Mortgage coach Ron Vaimberg sees AI as a ‘moving target.” So many new programs are coming out every month that it’s difficult to comprehend.” But, the Jefferson Valley, N.Y., consultant added, it’s “not going to end the business as we know it. Rather, it creates an opportunity to get things done faster and better.”

“AI (won’t) replace humans as much as improving their ability to do their jobs,” Michael Sachdev, CEO of Snapdocs, told me. “Everyone’s job will still exist, but they’ll be able to do more much faster.” At the same time, he added, it will “go a long way” toward eliminating the hiring and firing boom-and-bust cycle that plagues the mortgage business.

Sachdev noted that so many people have left the field so far in the current down cycle that many jobs have already changed. And when the market picks up again, as it always does, companies will have to rehire former workers if they are still available and train new ones to replace those who have gone on to other lines of work.

Predictive vs. Generative

The San Francisco-based executive pointed out that the difference between predictive and generative AI holds the promise. The former is a type of machine learning that uses advanced algorithms to analyze large datasets to identify hidden patterns and relationships between variables. With the latter, algorithms automatically produce content in the form of text, images, audio and video. These systems have been trained on massive amounts of data, and work by predicting the next word or pixel to produce a creation.

It’s generative AI that holds the promise of flattening out the hiring-firing-hiring cycle because it allows humans to do more with less, Sachdev said. In other words, with AI, you won’t need as many people when the market is running on all cylinders, so you won’t need to let as many people — maybe not anyone — go when the market sags.

“All things being equal,” Sachdev said, “we’re going to need fewer people; that’s just a reality. But right now, it’s incredibly difficult to run a business through hiring-and-firing cycles, one after another. AI will make the business much more sustainable, provide a great borrower experience, and lower costs.”

Sameer Ahluwalia, the previous leader of Sourcepoint mortgage business, agreed with the thinking that lenders should not be thinking about hiring more people to handle what’s in their pipelines but “how do I build my pipeline with fewer people?”

“It costs a lot of money to hire and train people every time the market picks up steam after a downturn,” Kirill Klokov of Truv also pointed out. “And lenders do that after every down cycle.”

Not everyone is on board with this kind of thinking, though. Casey Cunningham, founder and CEO of Xinnix, a sales, operations, and leadership performance company, sees a “complete shift” on the horizon. Loan officers who embrace AI will likely need fewer assistants, she said. And those same loan officers will become less like salespeople and more like advisors who helped build the company brand.

Obsolete Underwriters

The processor’s role will change, too, and “radically so,” the Atlanta-based Cunningham predicted. “The position will be much more customer service oriented than it has been. Processors won’t just be pushing papers and hunting down documents. There will be much more direct focus on clients services.”

And let’s not forget underwriters, who hold perhaps the most critical role in the lending assembly line. “AI is going eliminate them,” the consultant ventures. In the underwriting role, she said, “AI will have so much functionality that it possibly could approve a loan instantaneously without even having a screening process.”

Klokov of Truv also thinks loan processors could one day become unnecessary. Technology, he said, “can eliminate all manual processing so your staff can be more sales focused.”

Of course, one of the problems with AI is not AI itself but the adoption of the same. And on that score, Firstsource’s Ahluwalia said the mortgage sector is woefully lagging. “Extremely slow to adapt,” agreed Vaimberg, the New ork-based sales trainer. In many cases, older originators are not even willing to give technology a try.

“The mortgage business is 10-20 years behind the banking industry,” said Ahluwalia. “Not just in technology but practically everything. The power of AI has increased immeasurably, but adoption is still slow.”

Eventually, though, the mortgage workforce will “age out,” said Vaimberg. The old guard will die off, replaced by younger folk who have grown up with technology and are “more inclined to accept AI and what it can do to make their jobs easier and be more productive,” he added. “The question is, how long will it take?”

Moving forward, Ahluwalia offered this advice: Lenders “need not be first with everything. But you have to be a fast follower. Have some patience and measure your outcomes. Nobody has a silver bullet, but you have to move ahead. Work with people you trust. But keep them honest. If what they commit to doesn’t work, don’t pay them.”

And here, Carmine Cacciavillani of Blue Sage in Englewood Cliffs, N.J., gets the last words. “Don’t try to build your own programs. Rather, make use of what’s already available. Use someone else’s (system) and incorporate it into your own systems,” he said. “You’ll be able to double your productivity.”

But don’t wait until the market gets back on its feet, Cacciavillani also advised. “Now is the time to invest, not when so many leads are coming through the door that you don’t have time to even think about technology.”

Finally, let’s roll back to my local grocery store, where I suggested that the company’s motive was profit, pure and simple. But to the contrary, the clerk at the register told me, electronic checkout stations allowed her to do more of the other chores required of her. “There’s so much work to be done that we need more employees,” she assured me, “not less.”