U.S. Markets open in 6 hrs 13 mins

New Credit Score Model Looks to Transform Lending

Michael Schreiber

In a move that could have a wide-ranging impact on both the credit scoring business and the availability of credit to consumers, VantageScore Solutions today announced the latest version of its credit scoring model.

The company says that the new model, called VantageScore 3.0, has the ability to generate credit scores for tens of millions of consumers who previously didn’t qualify for a score, and will improve predictability by 25 percent over earlier models for consumers who have higher credit scores by looking at credit data in a more detailed way. It also excludes negative credit information that happened as a result of a natural disaster.

Among the most noteworthy features of the new model is the fact that collection accounts that have been settled or paid off will not be included in the score. Many consumers are unaware that currently, paying off a collection does not necessarily rehabilitate a credit score. Other scoring models, including the market-leading FICO score and previous VantageScore models, treat collection accounts — even those that have been paid off — as predictive of a consumer’s likelihood to pay off future debts. According to a press release announcing the new score, VantageScore is able to exclude the collection data and still make accurate predictions.

“This is potentially a big deal,” says Credit.com’s CEO Ian Cohen. “For tens of millions of consumers, one-time mistakes — maybe a single missed payment from a hospital visit — have led to years of subprime interest rates and denied credit. The issue has always been how to keep credit scores predictive so lenders can rely on them. These days, there are plenty of additional data sources available to do a ‘deeper dive,’ so to speak. This could help make credit more individualized. You actually see this approach in the insurance industry, and I’m glad we’re now seeing it in the credit scoring industry as well.”

[Related Article: FICO vs. VantageScore: 5 Differences You Should Understand]

VantageScore Solutions, which is owned by the three major credit bureaus (Experian, Equifax and Transunion), says it was able to improve predictability in the new model by including finer details in the data that is used to make up a score. For example, the scoring model doesn’t simply look at whether or not a student loan payment was made on time, but it can factor in how many days in advance of the payment due date that the payment was actually made. In the case of mortgages, payment patterns for different types of loans are treated differently. A 30-year fixed mortgage is not the same as an adjustable-rate mortgage or a home equity line of credit. VantageScore says this approach to the data is central to the model and it’s what allows for the exclusion of paid collection accounts. 

“We went into our efforts to build the VantageScore 3.0 model knowing that consumers with paid and/or unpaid collection accounts are higher risk that consumers without these accounts, thus this information can be usefully predictive.  This is especially true for unpaid collections,” said Sarah F. Davies, VantageScore’s SeniorVice President, Analytics and Product Management  “Using the more granular data in unique combinations that emphasized unpaid collections, we created variables that delivered greater predictive strength than paid collection variables. Given the mandate to deliver the most predictive model, these new combinations were incorporated into V3.0.”

VantageScore says it constructed the model using the anonymous credit data of 45 million Americans, (15 million files were supplied by each of the major credit bureaus), and examined the credit data in those files across two separate two-year time frames (2009 – 2011 and 2010 – 2012).

VantageScore Solutions’ scoring model competes for market share with FICO, the industry leader in credit scores. Among the changes that come with VantageScore 3.0 is the score range, which now uses the same spread as FICO (300 – 850). Previously, VantageScore’s range was 501 to 990.

In a 2006 court case, FICO argued that it had trademarked the 300-850 score range, however a 2011 appeals court decision ultimately rejected that claim and affirmed an earlier court ruling that FICO “obtained its trademark registration through fraud on the United States Patent and Trademark Office.”

According to the press release announcing the new scoring model, the purpose of the new score range is to “facilitate easier implementation for lenders, and more familiarity for consumers.” Indeed, consumers often don’t realize that they even have more than one credit score, and seeing different credit score ranges can be all the more confusing. VantageScore’s new scoring range will go some way in addressing that issue, though the thrust of the change may have more to do with boosting adoption of the scoring model by lenders. By some accounts – and these are unconfirmed — VantageScore holds less that 10 percent of the credit scoring market share, the vast majority of which is held by FICO. VantageScore’s CEO, however, argues that the new model will have a real impact on lending.

“The launch of VantageScore 3.0 advances the entire industry, and lenders will validate their portfolios and see for themselves how the model is transformative and extremely predictive,” says Barrett Burns, VantageScore’s CEO, though he concedes that lender adoption of the new scoring model won’t happen overnight. In fact, it will take at least a couple of months before even the credit reporting bureaus, who helped create the new model, fully incorporate VantageScore 3.0.

“Having been in a lender’s shoes, I know the sales cycle is long and lenders will need time to test the model on their portfolios,” Burns says. “One of the differentiators of the VantageScore 3.0 model is that it was developed with a lender’s implementation needs in mind, and the model opens up a tremendous opportunity to expand their customer base, so I think we’ll see the model gain traction relatively quickly.”

In addition to adopting a more familiar scoring range for lenders, VantageScore is touting 3.0 as a game changer in that it is able to score millions of previously unscorable Americans. The company says the new score does this in part by incorporating non-traditional credit data into its standard scoring model. Under this model, those with little or no recent credit data (available credit, revolving balances, late payments, account mix and age of credit accounts) are evaluated using non-traditional data such as collections, public records, rent, utility payments, telecom data and other inquiries. Further, when no recent traditional credit data is available, VantageScore 3.0 may look at credit data that is older than 24 months. According to VantageScore, including this data will make it possible to score between 27 and 30 million Americans who previously couldn’t get a score. This could potentially help to level the financial playing field for these consumers by giving them access to more affordable credit.

Likewise, consumers who have been affected by natural disasters — and as a result have missed payments on bills — may avoid being penalized via their credit score. Though the negative information will still appear on the credit reports, if a lender identifies a borrower with a “disaster reporting code,” then “information that would negatively impact a consumer’s credit score is ‘set to neutral’ so that consumers can continue to benefit from information that would have a positive impact.”

VantageScore has also launched a microsite, ReasonCode.org, in combination with the release of the new score. Reason codes are two characters long and come with a brief statements that explain to consumers why their scores are not better. Consumers often receive them when they’ve been notified that they have been denied credit, or have been given a higher than average interest rate on a loan. VantageScore has simplified the reason code language, and created ReasonCode.org as a searchable index for consumers to get more information about the codes and what they can do to improve their credit. 

More from Credit.com