This article is from the archive of our partner
Scientists and mathematicians comb through roughly 950 different data points when calculating an American consumer's credit score, the all-important number that can determine an individual's purchasing power on everything from an auto loan to a home mortgage. But only about 15 percent of those data points can actually predict the probability that a person will pay back her debt. The major credit reporting bureaus are constantly tweaking the mix they examine to give themselves the most current snapshot of consumer spending and debt.
Shifting through and modeling this data is the chief aim of VantageScore, a 10-person, Connecticut-based startup founded in 2006. One of the goals of VantageScore is to evaluate a greater portion of the population and, by extension, give more people access to credit. The company estimates that its latest 3.0 model allows it to score at least 27 million new consumers, a group larger than the population of Texas. "I look at the credit score as a gateway," says Barrett Burns, president and CEO of VantageScore Solutions. "Our main mission is to create the opportunity for people to have access to responsible credit. Even if a person gets a lousy score — if you do not know what your score is, then you do not know how to change your behavior."
Access to credit now consumes much of the discussion surrounding the housing market, which has not rebounded as fast as economists had hoped following the global financial recession. Scores of first-time home buyers, who typically make up about 40 percent of home sales, remain locked out of the market thanks to more stringent lending requirements. If mortgage lenders dropped the required credit score by just 50 points (on par with prerecession lending), then 12 million more people could gain credit and potentially purchase homes, according to a recent analysis by Mark Zandi, chief economist for Moody's Analytics. "It is difficult to see how the economy can achieve full employment over the next few years unless housing leads the way," Zandi wrote in a May research note.
VantageScore is not the only company to look for different ways to evaluate consumers' creditworthiness. Experian, one of the three big credit reporting bureaus, purchased a small company in 2010 to allow it to track rental payments as a way to measure a person's likelihood to pay off debt. A handful of financial startups have also entered the mix; one uses information from a consumer's social-media feeds, combined with private data from the credit bureaus, to determine a person's creditworthiness. Within the industry, there's talk of leaning on a payment as simple as a cell-phone bill or a person's remittances to his home country to factor into the almighty calculation of the credit score.
The problem with these different measurements? They all look to the past without predicting a consumer's ability to pay off debt in the future, says Annamaria Lusardi, a professor at the George Washington University School of Business and personal-finance expert. "We need a way to think about people's potential," she says. "The problem with having a static view is that just because a person, as an example, might be unemployed now does not mean that he will be unemployed in 30 years."
Far better, Lusardi says, would be to instead examine a person's current level of income, or the ratio of personal income to debt as a way to evaluate a person's creditworthiness. That's a strategy that Chi Chi Wu, a staff attorney at the National Consumer Law Center, also supports. Credit scores should also take into account the tough economic times individuals may face during a recession, like unemployment or a temporary inability to pay one's bills, Wu says. "Credit scores should distinguish between people who are a hot mess versus those who have just fallen on hard times. Giving them bad credit scores just keeps them down," she adds.
VantageScore does not incorporate social media into its credit-scoring models. Instead, the company leans heavily on the following ways to predict someone's credit score: The person's ability to pay bills on time, the history and type of credit he uses, and his total credit-card balances and total debts. A red flag pops up if someone accumulates huge amounts of credit very fast.
In the latest version of its model, VantageScore no longer takes into consideration paid medical debt or other types of paid debt that get kicked to a collection agency. The company no longer finds it "mathematically significant," Burns says. Consumer advocates, like Wu, applaud them for excluding this debt that can plague millions of Americans and easily drag down scores. The company has also tried to take into account the effect of natural disasters, like Hurricane Sandy, by pulling out the negative consumer data caused by these predicaments. Burns likes to call these tweaks "nifty differentiators."
Right now, VantageScore sells its credit-scoring models back to the three major credit-reporting bureaus. The long-term goal, however, is for Vantage Score to become its own household name, even if the federal government's major lenders, like Fannie Mae and Freddie Mac, still do not accept VantageScore as a primary credit scorer.
Advocates for first-time homebuyers and more diverse populations hope this will change and that a greater share of lenders will be open to these different algorithms to predict credit scores. "We're not looking for ways to get people into the system that are not qualified. We're looking for other metrics to identify people who will be successful," says Gary Acosta, cofounder and CEO of the National Association of Hispanic Real Estate Professionals. "A large percentage of first-time homebuyers or people from the Hispanic population have thin credit or very little traditional credit. But they have buying power, job stability, and strong assets. The marketplace eventually will be responsive."
Clarification: This story has been updated to make it clear that VantageScore only excludes paid medical bills and other types of paid collection debt from the latest version of its scoring algorithm.
This article is from the archive of our partner National Journal and part of our Next Economy coverage.