Mortgage robotics makes workers more efficient, not obsolete

Register now

The mortgage industry is no stranger to technological change — the shift toward a digital mortgage has been a long and steady process. As many mortgage companies have already discovered, robotics technology can be deployed at scale and integrated with myriad processes, creating a unique opportunity to rethink the entire loan manufacturing process.

"I've been in the industry for 21 years — that was right at the conversion from fax to digital," said Dominick Marchetti, chief technology officer at LoanDepot. "Whether we get the application through fax or digitally, our file flow in the industry has largely been the same."

"This is the first time we've had a non-linear, non-waterfall approach to lending, and as soon as that happens the world changes," Marchetti said.

To understand exactly how robotics can reshape how lenders and services approach lending, look no further than MUFG Union Bank.

At Union Bank, the introduction of robotics to the mortgage division stemmed from a recommendation by a single employee in the company's technology group. The lender needed to consolidate its loan files into a single system to boost efficiency, a process involving roughly 800,000 documents. That employee suggested using robotic technology from Kapow Software, a firm that has since been acquired by document and data capture technology developer Kofax.

By using RPA to facilitate the movement of these documents, a process that could have taken weeks if done using manual labor was finished within a few days.

Today, Union Bank has a group of employees who devote some of their time toward developing additional use-cases where robots could facilitate certain processes. Robots are used in conjunction with an optical character recognition process to move data from documents into systems. Bots update the indices for adjustable-rate mortgages. The software also produces checklists for systems to ensure compliance when onboarding new loans in the servicing division.

"With anything that we believe we're doing today that can be done by a robot, we come up with a proof of concept," Cheryl DeRoche Johnson, managing director of innovative solutions at MUFG Union Bank, said.

The time savings Union Bank found using RPA adds up quickly. Time is not just saved because the processes are expedited, but also because robots can accomplish multiple tasks at once.

"In the past you had to do things very linearly," Amory Booher, senior vice president of risk at BBVA Compass, said. "In some cases, robotics are enabling those things to happen simultaneously or nearly simultaneously."

As everything from employment and asset verification to pulling an appraisal happens at the same time, the loan manufacturing process goes from occurring in a very linear manner to a dynamic process. And reviewing processes for their applicability to RPA also affords companies a golden opportunity to decide whether those tasks are even necessary in the first place.

"You start taking something that takes six to eight weeks in the linear fashion, and now maybe you're just limited to the 10 or 12 days required by TRID," Marchetti said.

Time-savings is only one part of the equation. Robotics also offer an opportunity to dramatically reduce expenses.

"RPA is the ultimate and final low-cost location for labor," said Michael Henry, a principal at auditing and advisory company KPMG. "Instead of looking for an $11-an-hour file clerk in whatever the fashionable country is today, I can now have a robot for between 50 cents or $1 an hour to do the same function. And that cost is only going down."

Robotic process automation technology can cost half as much as a full-time offshore employee, according to the Institute for Robotic Process Automation. Compared with outsourcing, robots have other benefits, too — as Henry puts it, "there's no political unrest in robot-land."

Beyond being cheaper, robots can also perform processes faster and more consistently than human employees, improving turnaround times and reducing error rates. And bots can be programmed to record their actions, creating an audit trail that can help keep companies in compliance.

Because of such labor-related advantages, there is a natural inclination to view RPA as technology that outright replaces humans. And to an extent, that fear is founded.

A study published last year by researchers at Oxford University in the United Kingdom found a 98% likelihood that the occupation of a loan officer was one of the most likely to be computerized. It also theorized that 47% of all U.S. workers are at high risk of their occupations becoming potentially automated at some undetermined point in the future.

"This argument has been going on since the beginning of the Industrial Revolution really," said Andre Fuochi, global head of marketing communications at Blue Prism, a robotic process automation software provider.

"With some of these jobs, they're going to be automated," he said. "The technology's here, it's just a question of time."

But Fuochi and others stress that the focus of implementing this technology is not in reducing the number of staff. And many lenders and servicers find that implementing robotic process automation will involve hiring additional workers to get the software up and running and then to maintain it.

More than being about job elimination, the shift toward robotics is about allowing employees to do more in their jobs.

"We'd much rather have a SWAT team than a platoon," Marchetti said, referring to the way in which automation technology can perform certain tasks better than human employees. Real people are always needed by lenders though because, as he put it, "it's still a relationship play."

As a result of the added time workers will have to do more fulfilling tasks and the requisite skills presented by new technology, mortgage firms will need to adjust what they look for in who they hire. Ultimately, experience with robotics itself could become a prerequisite for many lending-related occupations.

"At some point in the future, it's going to be a base-level skill set," Booher said. "It will be something everyone gets their hands dirty with."

With employees doing less of the manual work that once took them so much time; they are free to perform other activities. That presents lenders and servicers alike with an opportunity to build capacity and to grow their companies.

Along those lines, RPA lays the foundation for more advanced robotics technology, like artificial intelligence and machine learning. With these tools, lenders can begin to move away from the yes-or-no decision-making that guides many choices lenders make in terms of prospective borrowers. This is what drove loanDepot to integrate robotics into mello, the company's new lending platform that features a web-based consumer portal and a mobile point-of-sale system supporting a fully digital mortgage application process, Marchetti said.

"We can suddenly say we have a really good view of what the loan looks like at the point of origination," Marchetti said.

Furthermore, companies can begin to "create a corpus of all the information out there and use that corpus to start making active decisions using real-time data that's more similar to a consumer's Uber or Amazon experience," Marchetti said. So as programs become smarter, more consumers in the gray area of underwriting will be brought into the fold.

"We're not going to just look at the happy-path customers who happen to be in the database," he said.

For reprint and licensing requests for this article, click here.