New data shows troubling trends in mortgage file accuracy
A new report has revealed that approximately 11.5% of all US mortgage loan file content contained errors or missing information over the past decade, leading to an estimated $7.8 billion in increased costs for consumers.
The findings, reported by LoanLogics, highlight systemic issues within the mortgage sector, despite significant investments in technology.
Decade of discrepancies costs consumers billions
The analysis by LoanLogics, which serves a substantial portion of the US lending market, examined data from 2014, 2019, and 2024. The study focused on two primary types of discrepancies: “Doc to Data,” where information claimed in a file is incorrect or absent, and “Doc to Doc,” where supporting documentation is incorrect or missing.
The combined error rate fluctuated, starting at 9.7% in 2014, peaking at 13.3% in 2019, and then falling to 11.4% in 2024. The decade’s average for combined errors settled at 11.46%.
“Our results show zero material improvement in loan file quality after a decade of industry investment and supposed innovation,” said Craig Riddell, EVP of market development at LoanLogics. Riddell attributed these persistent issues to “inappropriate application” of technology, incomplete training, rushed implementations, and aging integrations, which create “unexpected and costly data conflicts that necessitate manual intervention.”
Staffing fluctuations impact error rates
The spike in error rates observed in 2019, according to Roby Robertson, EVP of origination technology strategy at LoanLogics, “correlates to higher mortgage volumes across the industry.” He suggested this was “likely due to fluctuations in inexperienced staff brought on to deal with the increased workload.”
Conversely, the decrease in errors in 2024 coincided with reduced mortgage volumes. Robertson noted that lenders responded by “reductions in workforce, leading to more experienced and knowledgeable staff,” which in turn brought error rates down, though they remained near the 10-year average.
Call for enhanced automation
Robertson emphasized the need for better technological solutions. “As we continue to see new creative lending approaches emerge to serve more and more of the borrower population, it is imperative that companies work to solve their data problems with better automation and technology,” he stated.
LoanLogics’ technologies are used to process more than half of all loans in the US each year, having extracted and organized nearly 16 billion data elements and processed over 1.34 billion documents since 2005.
What are your thoughts on the latest findings? Share your insights in the comments below.
