A consumer group and credit industry trade group released a study on credit scoring inaccuracies and its affects on borrowers and lenders -- which were found to be both positive and negative.
The "Credit Score Accuracy and Implications for Consumers" study analyzed more than 500,000 individual credit files, and found that scores reported by the three repositories varied substantially.
The joint effort between the National Credit Reporting Association (NCRA) and the Consumer Federation of America is an attempt to point out inequities in the mortgage credit approval system and to weigh the impact of those inequities.
In the merged credit files that were analyzed, a number of inconsistencies were found.
When a lender plans to hold a loan in portfolio, he or she will order a credit report with scores from a credit reporting agency. That agency will compile the borrower's credit scores from all three credit repositories -- Experian, Trans Union, and Equifax -- using "merging logic," and will then supply the merged credit information to the lender.
The study found that among 1,704 of merged credit files studied manually, almost one in 10 was missing a credit score from at least one repository.
The same ratio out of 1,545 files had blatant mistakes such as credit activity of a totally different person in the file, or the file reporting the consumer as deceased. About one out of 10 files contained additional reports and information not relevant to a consumer's credit history.
Only 21% or one out of five of the total files had a range of fewer than 20 points between the lowest and highest scores; 31% had a range of 50 points or greater between scores; and 5% had a range of 100 points or greater between scores.
The average range between highest and lowest scores was 43 points, and the median range was 36 points.
Files with both good and bad credit were susceptible to large point ranges, although consumers with poor credit may be slightly more so, according to the study.
Merged credit scoring in mortgage lending protects borrowers from mistakes when omitted information, but it heightens the risk of borrowers being penalized for mistakes where erroneous information was added.
Accuracy in credit scoring is important to the mortgage lending process, the report said, and lenders heavily rely on Automated Underwriting (AU) systems for the approval process. But while AU systems are efficient, they do contain some shortcomings, NCRA said.
AU systems also can tell a broker right away if Fannie Mae or Freddie Mac will buy a particular loan. Since AU systems have become so integral to the mortgage lending process, it's important that credit scoring, a major factor for approval and at what price, be accurate.
Incomplete reporting of information, or an error in omission, can make a consumer appear either more credit worthy or less credit worthy, the report said. So not only is a false low score harmful to the consumer, a false high score is harmful to a lender.
The report conservatively estimated that 40 million consumers, or 20% of the 200 million with credit reports, are at risk of being misclassified into the subprime category, and at least 8 million of those would be misclassified upon application.
The study concluded that many consumers, especially those with very good credit, are unharmed by the inconsistencies found, and some may even benefit from them. However, tens of millions consumers are at risk of being penalized for incorrect information in their credit reports, in the form of increased costs or decreased access to credit and vital services.
The study suggested several ways to improve the system, including establishing meaningful oversight of the development of credit scoring systems. It also includes recommendations to help consumers stay on top of their credit report.
Credit scoring was introduced by Fair Isaac and it developed three different credit scoring models specifically for each repository: the FICO score for Experian, the Empirica score for Trans Union, and the Beacon score for Equifax. The term "credit score" also is commonly referred to as "FICO score," according to Fair Isaac.