TheMReport — News and strategies for the evolving mortgage marketplace.
Issue link: http://digital.themreport.com/i/1491913
32 | M R EP O RT FEATURE The same system should be able to place loans into other buckets based on the particular loan program or level of risk and complexity. For instance, suppose a borrower is applying for a loan program and the borrower meets all the requirements, but there are questions about the borrow- er's income. The lender's system should be able to note the poten- tial discrepancies and place the loan in a "gray" bucket, indicating the loan requires some human intervention before the loan can be approved. The same system could also make similar determinations based on whether the borrower has a credit issue or a complex tax return file. When underwrit- ing and other departments receive the file, it will be fully marked by the system with items that need attention, thus improving efficiencies even on loans that require human assistance. Once the items are investigated and it is determined the borrower does in fact qualify, the file can be sent back to the green bucket and the lender can get it done. Why Smarter Systems Are Needed U ltimately, a lender's system should leverage artificial intelligence that allows it to learn which bucket to place a loan based on decisions made on previous loans. As a result, a lender's "green" bucket of re- fi-ready files may start off small but will expand over time as the AI "learns" which loans will not require human intervention. In the meantime, a lender's staff will be able to focus on more sophisti- cated problems, so the lender can better optimize its staff resources. Over time, both the system and the lender's staff learn to operate more efficiently. For example, suppose someone is buying a home through a trust. Perhaps the system has not been programmed to deal with trusts, but it could place the loan in a bucket designated for a specific underwriter who has experience working with trusts. The same could be applied for borrowers of FHA loans or a borrower who owns multiple properties and has extensive documentation. The system will not only triage these loans and put them in the right spots but can eventually learn to make its own decisions automati- cally without involving anybody. Ultimately, a lender's pricing engine should also be able to identify the moment when a current homeowner qualifies for a refinance and preapprove the borrower before they have even thought about refinancing. The lender can then offer the bor- rower a "mortgage health check" and demonstrate how much they could reduce their payment, then provide a link that allows the borrower to move forward. Can't Sleep on Innovation W hen the housing market contracts as it has today, most lenders respond by cutting costs, typically by letting people go. Rarely do lenders respond by investing in new technology unless there's an immediate bene- fit to be realized. But this is exact- ly what they should be doing—in fact, they should be doing it all the time, whether push-button refis are their ultimate goal or not. That doesn't necessarily mean lenders need to build their own technology. But they should be constantly assessing and reas- sessing their needs as well as the availability of new technologies to ensure they are optimizing their operations for the current and future market. One rule of thumb for choosing technology: don't pick the prod- uct—pick the company. Most tech- nology vendors will claim their products can be easily integrated with a lender's other systems and be customized for how they do business. And yet, not all vendors will deliver and very few are in the business of building new, cloud-based technologies. In fact, this is precisely why so many lenders are still dependent on legacy software and technol- ogy vendors that not only fail to help them control costs but typ- ically create additional expenses. For this reason, it may be wise to look beyond the handful of off-the-shelf products that have dominated the mortgage technol- ogy landscape for the past decade and find a vendor with in-depth experience building new technol- ogy and the ability to customize anything they have already built to fit your needs. Today's market may not be conducive to refis, but it's a per- fect time to innovate better ways of originating them. While the ability to fund loans without any human involvement may sound like a stretch, it's probably going to happen one day. In fact, in the back of every innovator's mind is the idea: "If we don't do it, some- one else is going to." Speaking for my own team, at least, every line of code is written with this idea in mind. The bottom line is lenders can choose to feel nostalgic for the good old days or they can put in the work now to make the next refi market even better. As the late Hubert Humphrey once re- marked, "The good old days were never that good, believe me. The good new days are today, and better days are coming tomor- row. Our greatest songs are still unsung." CARLOS SA is the CEO of MILOS, an IT consulting firm serving the mortgage industry. He has more than 25 years of IT experience and an extensive track record of designing, build- ing, and implementing mortgage technologies. Prior to founding MILOS, Sa served as Head of Information Technology for Mortgage Network, where he and his team built the company's entire IT infrastructure, including its accounting systems, loan origination system, pricing engine, online rate lock, and online borrower portal. He can be reached at CSa@ MilosSolutions.com. One rule of thumb for choosing technology: don't pick the product—pick the company.