You may have seen or heard TV and radio ads for “Mortgage approvals in minutes!” The “approved in minutes” concept has the attention of many consumers, but behind the marketing hype is the same old, slow mortgage approval process (in the data center). While you may upload some information on your phone for the front end of the mortgage process, it’s equally important to have faster processing in the data center.
The mortgage experience will be completely transformed when an actual approval, including underwriter support, takes one to two days, rather than several weeks. We need technology to help get us to a real world of rapid approvals. Artificial intelligence (AI), machine learning and automation will play a key role in making single-day mortgage approvals a reality.
AI will make the gathering, reviewing and verification of mortgage documents much faster and more accurate. It can turn the application process into an “assembly line” system that not only moves faster but has fewer errors and is more efficient. The end result is faster transactions, happier home buyers and improved productivity for banks and other lenders.
This eWeek Data Point article features industry-level predictions from Steve Butler, founder and General Manager of AI Foundry, who sees the following changes coming in the mortgage industry as a result of advances in AI technology.
Data Point No. 1: Lenders Will Turn to AI to Reduce the Cost and Time to Close a Mortgage
Today, it takes about three weeks to close a mortgage, starting from the time that all the required documentation for the mortgage application is submitted, through providing verification of information, processing, reviews and underwriting, and to finally closing. AI will drive a new generation of software robots that automatically process mortgages, replacing slow and costly manual processes. We will begin to see approval times drop to one to two days, and this will become the norm.
Data Point No. 2: AI Will Make OCR Technology Obsolete
We will see steady declines in the use of OCR (optical character recognition) technology in document processing. OCR was invented nearly a century ago and is losing its utility today, because it can’t do anything intelligent with the images it scans. OCR will give way to AI, which enables machines to “read and react” to content. This will enable a boom in “white collar automation,” where manual document processes (such as mortgage processing) are replaced by software-based robotics processing.
Data Point No. 3: Mortgage Processing Will Come Back On-Shore; AI Will Lead the Way
Most Americans do not know that when they apply for a mortgage, often their personal financial information goes overseas to processing centers in India, the Philippines and other low-wage countries. We have seen more of this onshoring recently, and this trend will continue and grow in 2019. This is driven by a number of factors, including regulations, increased customer complaints and companies gaining a better understanding of the true costs of offshoring. This will cause a spike in back-office AI and automation investment as mortgage lenders try to keep processing costs down while complying with emerging privacy regulations.
Data Point No. 4: ‘Instant Mortgages’ Will be Exposed
Consumers will get more savvy and educated about claims of “instant mortgage approvals.” It still takes weeks to process a mortgage and “approved in minutes” applicants can still get turned down by underwriters when going through the full mortgage application process. Mortgage lenders will turn to back-office automation to accelerate the traditionally slow mortgage-approval process, so consumers can have a true “Uber experience,” where mortgages are approved in one to two ays, not weeks, while having visibility into the status of their mortgage during the entire process.
Data Point No. 5: Real Estate Firms/Websites Will Become ‘Single Source’
We will see more real estate firms and websites becoming one-stop shops for the home buyers, with a horizontal model similar to how automobile dealers offer financing and maintenance. The new model will allow buyers to access everything they need (financing, legal counsel, accredited inspectors and contractors, etc.) from their realtors, rather than relying on referrals and research. With AI driving back-end mortgage processing, the silos will be torn down between lenders, inspectors, realtors and others in the home-buying scenario.