Enter your e-mail address and password below.


Forgot your password? New User? Register Now.
   ARTICLE   |   From Scotsman Guide Residential Edition   |   October 2016

Scanning for TRID Risks

Change-of-circumstance audits should not involve a manual process

Scanning for TRID Risks

With the one-year anniversary of the implementation of the TILA-RESPA Integrated Disclosure (TRID) rules upon us, the mortgage industry is still grappling with the broad-sweeping changes this piece of legislation brought. Although the Consumer Financial Protection Bureau (CFPB) recently issued proposals to address several major issues lenders have faced with TRID, there are still kinks to be worked out in TRID-related procedures.

One such area is tracking change of circumstance, particularly when multiple TRID-mandated Loan Estimates and Closing Disclosures are issued to borrowers. Although TRID officially went into effect in October 2015, it wasn’t until 2016 that the industry shifted its eyes away from disclosure-timing requirements to take a closer look at change-of-circumstance issues.

Once originators had a handle on sending out Loan Estimates (LEs) and Closing Disclosures (CDs) in a timely fashion, they began to see the amount of work involved in properly reviewing these TRID forms to confirm that change-of-circumstance notices were provided to consumers when changes were made during origination. More importantly, they needed some way to determine that these changes were allowable and met TRID-tolerance requirements.

Today, many mortgage companies are attempting to manage the LE-CD review process manually, which could expose them to significant risk once the CFPB begins auditing lenders in earnest for TRID compliance. There are opportunities to automate this process, however, which will enable originators to reduce the time required to accurately log and track this critical information and decrease their level of risk exposure.

Ironically, it is technology that is at the root of this problem. Most companies have already moved to automated processes for producing and storing documents electronically. According to the “2015 Xerox Path to Paperless” survey, 78 percent of mortgage professionals report delivering disclosures or other documents to borrowers electronically, and 74 percent have implemented paperless origination and underwriting processes.

Manual data extraction

The ability to extract data from these electronically stored documents is not as easy as you might think, however. When changes to the loan amount, interest rate or fees associated with a loan take place, originators must issue a change-of-circumstance notice outlining the changes and also must issue an updated LE or CD to the borrower. Unfortunately, most loan origination systems (LOSs) are not storing the historical data needed to track all of the changes that occur during the origination process.

The issued documents are stored electronically, but the information contained in the documents is not easily accessible for reviewing, exporting, analyzing or reporting. This means there is no system-driven process to complete a side-by-side comparison of the multiple LEs and CDs present within an individual loan file. One workaround many mortgage companies use to address this issue involves rekeying data from multiple LEs and CDs on an individual loan. How do many of these companies store this data? You guessed it, via Microsoft Excel.

Specifically, processors are manually creating spreadsheets and entering all the critical data fields, including all points and fees data from each disclosure issued to each borrower. This methodology allows internal auditors to compare all issued disclosures to see where fee changes occurred, determine if the changes were allowable and if the fee changes remained within appropriate tolerance levels, and then verify whether proper change-of-circumstance notices were actually provided to the borrowers. As with any manual solution that involves rekeying data, however, this process is terribly inefficient and prone to error.

When CFPB audits of TRID compliance begin in earnest, mortgage
companies can point to a system-driven process to demonstrate compliance.

On average, auditors are comparing anywhere from four to eight disclosures for any given loan file. Using this methodology for LE-CD comparison, companies are spending an hour or more per full-time employee (FTE) per loan file to complete this process. Unfortunately, there’s no real way to track the error rate companies are experiencing by rekeying data, but common sense and past experience tell us it must be a fairly common occurrence. Whether these errors will result in significant violations is unknown, but this is an enormous — and easily fixable — question mark to have looming over your TRID-compliance status.

Electronic data extraction

Technology holds the key to solving this rekeying issue. Optical character recognition (OCR) technology has been leveraged to ingest and index documents based on their contents for years. OCR technology is normally associated with scanning a stack of paper documents and running the imaged documents through the OCR process to extract key data points. Early OCR solutions provided less-than-optimal accuracy, but the technology has significantly advanced and can now produce data-extraction confidence levels upward of 90 percent.

Although the presence of paper documents is less prevalent in today’s origination environment, OCR remains relevant because of its ability to extract data points from electronically stored documents, especially where the system of record cannot do so, as is the case today when mutiple LEs and CDs are issued for an individual mortgage loan.

Using OCR technology, processors can quickly and easily extract all key data points from the multiple disclosures and simplify the LE-CD comparison while retaining the integrity of the original data to avoid “false positives” and rekeying errors. Using this approach can reduce comparison turn times to as little as five minutes per FTE per loan file.

Once the data is extracted, loan processors then can match the information on the disclosures with the information contained within the LOS to further validate that the final documents match the LOS data. Down the line, when CFPB audits of TRID compliance begin in earnest, mortgage companies can point to a system-driven process to demonstrate compliance with TRID, particularly in regard to change of circumstance. They also will be able to provide documentation of that process via the final output, which can be saved to the loan file showing the detailed comparison matrix.

Drawbacks and solutions

Of course, no solution is perfect. Even though data-extraction confidence levels for the most advanced OCR-based systems are extremely high, no single solution can guarantee 100 percent accuracy. Thus, when companies evaluate OCR-based systems to conduct their LE-CD comparisons, they should note not only the system’s data-extraction confidence levels, but also the functionality in place to allow their loan processors to “triage” questionable extractions.

The ability to easily review only those extractions the system isn’t 100 percent confident in is critical to maintaining reduced turn times for comparisons. If the OCR system can achieve a 98 percent data-extraction confidence level, but processors must review 200-plus data fields in every single LE and CD scanned to find the 2 percent that is erroneous data, then that system failed to make the process more efficient.

Furthermore, the system should be flexible enough to accommodate individual naming conventions for fields within the LE and CD, because mortgage companies and their LOSs do have minor variations on that front. Field names between the LE and CD are fairly consistent across the board, but the comparison process should not be hamstrung because the OCR system continues to flag non-issues due to minor naming differences.

•  •  •

Even with the CFPB’s updated guidance on TRID, compliance with this broad-sweeping piece of legislation will continue to be a moving target. By leveraging advanced technology to automate and streamline key compliance and quality-control procedures related to TRID, however, mortgage companies can increase their level of confidence in their ability to meet the new requirements and reduce TRID-related defects, as well as general origination errors overall.


Fins A Lender Post a Loan
Residential Find a Lender Commercial Find a Lender
Scotsman Guide Digital Magazine

Related Articles



© 2019 Scotsman Guide Media. All Rights Reserved.  Terms of Use  |  Privacy Policy