Both the mortgage industry and the real estate industry that it supports are complex businesses subject to a high degree of regulatory oversight. While these businesses are technically separate, any chief economist will tell you that the housing markets are tied tightly to the mortgage markets. Without access to funding from mortgage lenders at an affordable interest rate, the housing industry would fail.
While that may seem far fetched, we witnessed a near collapse of the housing industry during the financial crisis of 2007-2008, when mortgage backed securities lost nearly all of their value on the open markets. Had the federal government not stepped in with additional funding and new legislation, the recession could have been much worse.
Today, artificial intelligence (AI) is helping mortgage lenders and servicers do more business more profitably while maintaining compliance with investors and regulators. Quicken loans may be the example that springs to mind for many, but there are many other success stories, and many more to come.
In this guide, we’ll show you where AI has changed the industry and the promise it holds for the future, but first a look at the technology the industry relies on today. Over the years, many events have served to shed light on industry problems that innovative technology providers have stepped forward to solve. Here are some of them.
The industry’s driving force: technology.
When local banks first began making home mortgage loans, loan committees would make individual loan decisions based on their knowledge of the credit worthiness, the real estate collateral and the character of their prospective borrower. To make these decisions, they relied upon the facts they had at their disposal, generally supported by paper documents. Mortgage companies operated from paper forms and managed their businesses with file folders filled with documents. It was a slow, manual process but one that worked.
As the industry grew and more Americans decided to become homeowners, mortgage bankers found their work much more difficult and time consuming. One of the industry’s first technological innovations was computer software to make the completion of mortgage loan applications simpler. This wasn’t mortgage automation, per se, but it opened the door to using computerized systems for processing mortgage loans.
The paperwork required to make an application for most mortgage loans had been standardized by the federal government by the middle of the last century. In 1934, the Federal Housing Administration began insuring mortgage loans and in 1938 Fannie Mae (then the Federal National Mortgage Association) was created as a government agency to ensure a reliable and affordable supply of mortgage funds throughout the country. In 1968, Ginnie Mae, the Government National Mortgage Association, was formed as the second government-sponsored enterprise (GSE) and tasked with providing liquidity for government-insured mortgage loans. In 1970 Freddie Mac, then known as the Federal Home Loan Mortgage Corporation, entered the business as the third GSE to add more liquidity to the market.
Fannie’s Form 1003 mortgage application became the de facto standard for the industry, making it clear exactly what information lenders needed in order to consider making a loan. By 1995, Fannie Mae had released software that would allow lenders to automatically receive an underwriting decision from Fannie Mae electronically by submitting the information collected on the 1003 and borrowers’ authorization for the lender to pull their credit scores.
While it would take another decade to catch on industry wide, this was the first giant step in automating the mortgage process because virtually all lenders were collecting the same data on the same forms and sending it electronically for automated underwriting. For the first time, it became economically feasible for mortgage tech providers to create new systems because they could be sold to many lenders.
What really opened the door for innovation was the passage of the Uniform Electronic Transaction Act (UETA) and the Federal Electronic Signatures in Global and National Commerce Act (1) (“E-Sign”) in June 2000. Now, borrowers could give their consent electronically to have their data shared between parties involved in mortgage loan origination.
Almost overnight, the industry saw new workflow and automation tools either offered as functionality built into their existing loan origination system (LOS) software or integrated with it that allowed them to collect more information from borrowers electronically, verify that information electronically, get the loan underwritten, and then clear any stipulations and clear the loan to close.
By 2007, mortgage loan origination was at an all time high, exceeding $3 trillion annually. Some investors found it much more profitable to loan funds to riskier borrowers at a higher mortgage rate. Unfortunately, this type of innovation in mortgage products introduced loans that investors ultimately considered too risky. When the bottom fell out of the market, the federal government stepped in and Congress passed the Dodd Frank Wall Street Reform and Consumer Protection Act, which ushered in the age of RegTech (regulatory technology).
This was followed by the foreclosure crisis that saw many changes made on the servicing side of the mortgage business. This was basically focused on automation that would ensure that mortgage borrowers could get the attention of a single point of contact as they worked to get their mortgages back on track.
As the industry worked to build back consumer trust, the borrower experience became more important. As a result, a flood of digital Point of Sale tools entered the market with the promise of making it easier for consumers to find, evaluate, and apply for a new mortgage loan. Today, lenders are getting very close to fully digital mortgage transactions, but the technologies in use are still, for the most part, workflow automation tools. As long as the transaction doesn’t generate any exceptions, the technology works perfectly. But when something doesn’t fit quite right, the loan falls off the virtual assembly line and a human must step in to deal with the problem. Very few loans make it all the way through our process today.
But that is all poised to change. Artificial intelligence (AI) will empower industry automation to do more of the tasks that today fall to human workers. AI in mortgage will change the way the industry operates. Mortgage lenders as well as those engaged in mortgage servicing will benefit. Let us show you how.