The pace of technological change in mortgage lending in recent years has been staggering. Not only have new regulations radically reshaped mortgage lenders’ origination processes, but they have also made them increasingly reliant on software and systems in order to automate those processes and realize new efficiencies.
But as lenders are learning, technology is also an important differentiator in terms of driving customer satisfaction. Today, most borrowers are starting the home buying process online – thus, the mortgage process must be equally as simple and streamlined.
As we’ve been seeing, traditional lenders are now catching up with non-banks and “fintechs” and are deploying Web portals where borrowers can not only shop for the best loan to meet their needs, but also complete the application process and find out if they are pre-approved in a relatively short period of time. Making this possible is an ecosystem of applications on the back end of every lender’s website that handle important and complex functions, such as auto-verification of income and assets, as well as automated underwriting.
Putting such an ecosystem together, however, is no small task. Traditional lenders have been challenged in recent years to redesign their processes and find the right systems that can carry them forward in a new, highly regulated and exceedingly complex lending environment.
As explained by Abhinav Asthana, product manager for Wipro Gallagher Solutions, the mortgage industry has moved through three distinct stages of technological development during the past 10 years.
“In a way, it’s not too different from the technological stages that the Internet has gone through,” Asthana tells MortgageOrb. “In ‘Mortgage 1.0′ – say, around 2006 – we had lenders that had deployed advanced Web-based technologies, but the goal then was simply to process as many loans as possible without worrying about the quality and the regulations and whatever the risk aspects might be. So, even though we had products NINA – no income, no assets – there was, in fact, a design at the time.”
“But then it all blew up in lenders’ faces – and they did not know how to react to it,” he continues. “So, as a result, there was a major design change that occurred. This is when the industry entered its ‘Mortgage 2.0′ stage. At that time, lenders’ main focus was on designing processes and systems that ensured compliance, as well as loan quality.
“But those first two design strategies – one of which is based on greed and the other of which is based on [containing] the damage – did not do the borrowers any good,” he adds. “And, if there are no borrowers, there is no lending. It was around this time that many lenders first began to wonder, ‘Why is it that a customer who has a primary banking relationship with me does not want to borrow from me? What is wrong?’”
As a result, lenders have now entered what Asthana calls the “Mortgage 3.0” stage of technological development – where the processes are more focused on the borrower while at the same time maintaining a strong focus on compliance. This trend has been driven mainly by the non-banks and “fintechs,” which have little to no legacy in mortgage lending.
Designing an origination system that delivers an outstanding borrower experience while, at the same time, meeting an ever-increasing array of complex regulations, however, requires careful planning and striking a delicate balance, Asthana says. What’s more, lenders must be open to the possibility of more frequently redesigning their processes in response to new regulations. As such, lenders would be well advised to incorporate “design thinking” into their operations, he says.
“Design thinking” basically means visualizing the process from the borrower’s perspective. As more and more individuals prioritize the experience of getting a mortgage product over the product itself, mapping the borrower journey will become a critical component to optimizing loan origination systems and digital technologies in a way that is customer-driven, he says.
Borrowers don’t just want a convenient, tech-enabled and simplified mortgage experience – it must also be personalized, he says. Therefore, lenders must invest in technology that can help them deliver a more personalized – yet still automated – experience.
“Borrowing is a very important two-way trust relationship,” Asthana says. “The lender has to trust the borrower’s intent and ability to repay – and the borrower has to trust that the lender is not duping them with extra rates and fees. So, in terms of meeting the expectations – it’s more like the borrower has to meet the expectations of the lender; however, the industry is fast moving to the realization that the borrower is the center of the business. And lenders must now also meet borrowers’ expectations.”
As a result, lenders are now “re-imagining the lending process,” he says.
“In terms of redesigning processes, the customer must take the forefront,” Asthana says. “The industry knows that there’s going to be increasing regulations moving forward – so the question becomes how to redesign the processes so that the customer is at the center of everything. That’s what’s driving the next phase of design thinking in the industry.”
Interestingly, the best way to achieve this is basically to start with the borrower and work backward. Lenders must continually ask themselves, “What is it that today’s borrowers want?” In addition to a smooth, streamlined, online mortgage application process, borrowers also want to learn at the end of the process whether they are pre-approved. As such, it is very important to design a system that takes each borrower all the way through the process to the end. If a borrower gets even the slightest idea that he will be turned down for the loan during the process, he will likely abandon and the data will be lost – “and lenders don’t want to miss those opportunities,” Asthana says.
“Lenders also should stop considering new regulations as burdens and detriments to their business,” he adds. “It’s that mindset that creates overkill in the process. They should look at new regulations as enablers to do new things – and in the case of this current environment, that means putting the customer at the center of the process.”
“Lenders often design their processes with the regulations in mind – that is to say, they design their processes so as to avoid any fines or penalties,” he says. “But with the Consumer Financial Protection Bureau’s new TILA-RESPA Integrated Disclosures rule, for example, don’t you think lenders should be sending that type of information to the borrower within three days [of the closing] anyway? Isn’t the goal to help the borrower make the right decision? That’s the type of question that lenders should be asking themselves.”
Asthana says part of the goal for lenders now “is to break the process down into bits of information, which can then be fed to borrowers so as to guide them through the process in a way that they understand it.”
“If you look at loan quality and you look at customer satisfaction, you’ll see that they are closely linked together,” he says. “If you’re doing high-quality loans and you’re doing them right, then you’ll always have very satisfied borrowers.”
One concern for lenders in designing automated systems, he says, is that they could end up “filtering out” certain borrowers who might otherwise have been considered for a loan. This is partly because the only way to automate the process is to rely on data from third-party systems – and that data can sometimes be erroneous. If a certain database, for example, has wrong information for a particular borrower, that borrower could end up being “auto-declined” over and over, unless a live person manually reviews the file.
Worse yet, if a certain borrower is “auto-declined,” he or she may decide that he or she should not bother applying elsewhere for fear of the same result.
“It’s about using the right amount of automation, the right amount of artificial intelligence and the right amount of process re-orientation to determine – at a very early stage – whether that borrower will be able to work with that lender to get what they want,” Asthana says. “And it’s a very delicate balance – because lenders don’t want to lose an opportunity. They want to make sure they’ve tried everything possible before the borrower withdraws. But in order to do that, a lender must use multiple channels of engagement.”
Just like retailers, lenders also must be careful not to completely eliminate the human element from their services. Although today’s borrowers are desirous of a completely automated process, they also want the option to “zero-out” to a live agent if they have a more complex question or feel a need for higher-touch service. Asthana says that’s why many lenders today “switch the borrower over to a live agent at the end of the process – to have that live agent tell the borrower whether or not they are pre-approved.”
It is equally important that lenders provide multiple channels through which borrowers can reach live agents if needed. As such, many are now offering click-to-call and callback features prominently on their websites. Asthana points out, however, that there are no accurate stats on how many customers actually zero-out to a live agent.
An interesting question is how much artificial intelligence will take over the role of the human underwriter in the years to come. Asthana says Wipro has done a lot of research over the years to see how much data an automated underwriting system can draw upon in order to automatically render a decision. For example, might a system like Holmes (which was developed by Wipro and is sort of an equivalent to IBM’s Watson) be used to determine borrower intent or even sniff out fraud?
Asthana says although artificial intelligence is playing a more prominent role in the front end of the mortgage process, “There’s always going to be a degree of traditional underwriting that occurs regardless of how sophisticated a lender’s technology becomes.”
“A lender cannot write off its risk to automation – there must be individuals who are accountable,” he says. “The problem with technology is if it is done right, it will be right 100% of the time – but if something goes awry, it can just as easily be 100% wrong.”
“From a design perspective, the mortgage process must be an optimal combination of automation and manual underwriting,” Asthana says. “A lender can take the repetitive parts of the mortgage processes and automate them – but they must still leave that final element of judgment to a human being.”