Lenders: When was the last time you received a complete, error-free Closing Disclosure (CD) from your title and escrow processor in under an hour? Probably never. That’s because the current fee collaboration process between lenders and their title and escrow partners is akin to the horse-and-buggy transportation method of decades past: Slow, bumpy, and clunky.
In the current process, the lender sends the CD (or closing instructions) to the escrow associate, who manually types fee data received from the lender into the settlement statement, then “stares and compares” to ensure the data matches correctly. This tedious procedure is highly susceptible to human error, necessitating multiple document changes before a final statement can be transmitted back to the lender, allowing the mortgage closing to move forward at last.
This archaic process is further complicated by disclosure timing requirements imposed by the TILA-RESPA Integrated Disclosure rule, or TRID. Under rules implemented by the Consumer Financial Protection Bureau in 2015, the Loan Estimate (LE) must be provided to consumers no later than three business days after they submit a loan application. If a change in circumstance occurs, the lender can revise the LE within three business days, but the revised LE must be provided to the consumer no later than seven business days before consummation of the transaction. A Closing Disclosure (CD) must be provided to the consumer at least three business days prior to consummation. Prior to consummation, an additional three-business-day waiting period applies when there are changes to the CD that result in an increase to annual percentage rate, the addition of a prepayment penalty, or a change in loan product. Any discrepancies in the form can reset the clock on these timing requirements, causing even more delays.
This is especially problematic during busy periods such as the end of the month, when a flurry of last-minute activity could increase the probability of errors, place extra demands on overloaded staff, and increase lenders’ wait times for finalized documents – ultimately delaying loan cycle times.
Seeking to tackle these common pain points, title and settlement service providers are developing solutions that improve communication between lenders and their title and escrow processors, parse data from one file to another, and securely receive, track, and share documents. However, these solutions often require the use of and integration with specific loan origination systems and fail to completely remove the need to verify and rekey data if errors occur.
Rather than make fractional improvements to an obviously flawed process, Doma is addressing this common lender pain point with the same innovative approach we have taken to develop our suite of patented instant title underwriting solutions.
In June 2019, the U.S. Patent and Trademark Office issued to Doma a patent for “Machine Learning Using Multiple Data Types.” Our predictive analytics algorithms cull proprietary data sources on the property and assign a risk score to determine how safe a property is in terms of potential liens and other liabilities that may cloud a property’s title. The solution can deliver a final title commitment, free of defects, in less than a minute for 80 percent of refi transactions.
In December 2019, we received a second patent for “Predictive Machine Learning Models,” which enables us to optimize the value-maximizing point in the instant underwriting model by modifying the previously well-known Markov Chain Monte Carlo technique with unique title-specific mathematical components, turning a risk score into a profit/loss number. The model shows a low likelihood of claims, and if a claim is incurred, they are expeditiously processed.
“As far as I know, we’re the only company in our industry with academic and industrial expertise from astrophysics, data science, machine intelligence, supply chain management, operations research, and engineering disciplines collaborating with leading title and escrow professionals across the country,” said Christopher Morrison, Chief Operating Officer at Doma. “Their collective expertise puts us in the best position to leverage the extensive institutional knowledge we have within the title and escrow industry.”
Doma is currently applying these same methodologies and expertise to the fee collaboration conundrum, using natural language processing to parse CD data and map them to the correct fields. The loan origination software (LOS) agnostic solution will allow lenders to work within their preferred platform solution and significantly reduce time spent managing settlement preparation, result in higher accuracy of fees and payments, and eliminate time-consuming errors associated with manual data transcription.
Ultimately, loan processors will spend less time waiting for documents and see a reduction in friction and loan turnaround times. At the same time, automating our internal processors’ rote tasks will free them up to focus on solving complex title issues requiring special expertise. Finally, the technology will empower us to do more work with fewer resources, reducing operating costs for both us and our lender customers – achieving a truly instant close.
Writing for Scientific American in 2019, Doma CEO Max Simkoff observed, “With an ‘it ain’t broke, why fix it’ mentality, companies in this market were keeping the same old processes in place, with personnel repeatedly entering the same information into multiple computer programs – a system that’s expensive, rife with errors and easy to improve through automation. We bring a forward-thinking, innovative, creative, technology-driven operating model to the table.”
