The housing market crisis that began in 2007 brought about a monumental change in the mortgage industry. Not only did it help bring forth one of the worst economic depressions in U.S. history, it also was a catalyst for the government to bring forward new regulations to help ensure something like this did not happen again in the future. The government instituted new regulations, including the Dodd-Frank Act, which among other things created a new agency, the Consumer Financial Protection Bureau (CFPB).
Once the Dodd-Frank law was enforced in 2010 – which defined new rules while integrating existing rules (example: TILA and RESPA) – both loan originators and due diligence organizations frantically tried to update their software and processes to ensure adherence with the rigid rules implemented. These changes were the largest regulatory changes in the mortgage industry in over 40 years. Alignment by all loan origination and due diligence systems to new regulations related to TILA-RESPA Integrated Disclosures (TRID) rule was required by Oct. 3, 2015.
These new regulations required significant investment to update existing company software. It also brought additional processes into existing workflows and required changes across the entire life-cycle of systems. Implementing these new regulations and processes demanded significant effort owing to an already complex due diligence process. Review times doubled from prior to the implementation of the regulation.
Recognizing that adherence to the new regulations was mandatory, and understanding that clients did not want to see an increase in their costs, due diligence firms needed to figure out ways to be more efficient in their processing. The good news is, while these new regulations were being introduced, there was a massive boom in the market around artificial intelligence, business process automation and machine learning. These technologies allow due diligence organizations to reduce costs while aligning to the regulation that clients need.
Any due diligence provider that wants to ensure relevance going into 2020 will be looking to adopt the following technologies:
Artificial intelligence: Artificial Intelligence engines, like Wipro Holmes, is needed to build next-generation Mortgage Document Recognition Platforms. Unlike the common OCR technologies that were limited to being able to recognize only common documents like Tax Form, Loan Estimate, etc., these new AI engines can read the document in a similar way a human can and is able to determine unstructured documents and derive information. This technology will drive an increase in accuracy and provide a major reduction in turn-around times.
Business process automation: Implementing Process Automation will automate data entry into proprietary systems without the need for a system rewrite. This type of Process Automation allows organizations to leverage existing platforms without needing to engage the software development team in making any system changes. It removes the risks and time associated with data entry from a loan document to internal systems, which will provide efficiencies and increased accuracy.
Business rules engine: Recognizing the aggressive evaluation of the rules, Rating Agencies require due diligence firms to have the ability to evolve quickly. With a User Defined Business Rules Engine, the business will be able to alter the compliance rules directly. Extending out the rules engine to automate the raising of tests, exceptions, and grading will allow for faster turn-around times which will reduce costs.
Machine learning: In addition to the near-term technology solutions mentioned above, I think we need to start looking very closely at our ability to automate the entire process via Machine Learning. Machine Learning will look at all past due diligence work and based on the data from the loan, quickly determine all the exceptions and automatically send the results to the client/seller for review. Not only will this be valuable in the secondary market, it will also become a tool for the loan origination market to ensure alignment before loan closure.
In the next five years, we could arrive at a future where a client can send a loan document, and within one hour the due diligence organization can send it back with due diligence results at a greatly reduced cost. Full automation will revolutionize the industry and will allow a near 100% of all loans to be audited and graded instead of just the limited pools that are done today due to costs.
The first due diligence organization that embraces this automated future will become the leader in the due diligence space. If you do not believe this future is possible, just spend a couple of days in southern California and observe the numerous driverless cars operating in the streets. If we can train computers to operate a car, there is no reason why we shouldn’t think that we can train a computer to audit a loan file.
David Christiansen is chief information officer for Opus Capital Markets Consultants.