A woman and man stand together smiling in front of their new home in North Carolina. They received a loan not based on credit scores, but on character.

#183 Summer 2016 — Financial Well-Being

Challenging the Almighty Credit Score

A majority of mainstream lenders base loan approvals on a hotly debated three-digit score. Are there better, fairer ways to assess risk?

Imelda Carranza and Raul Gomez are all smiles after closing on a home in Raleigh, North Carolina. The two were able to purchase the residence after they were approved for a loan from the Latino Community Credit Union. Photo courtesy of Ricardo Brandiarian

 

A woman and man stand together smiling in front of their new home in North Carolina.

Imelda Carranza and Raul Gomez are all smiles after closing on a home in Raleigh, North Carolina. The two were able to purchase the residence after they were approved for a loan from the Latino Community Credit Union. Photo courtesy of Ricardo Brandiarian

Credit scores go hand in hand with buying a home, renting an apartment, or leasing a car. But as more and more companies begin to rely on that three-digit number for a bevy of purchases, including non-loan items such as insurance and utility rates, some are recognizing that those scores don’t tell the whole story. For years, there’s been debate about how credit scores should be calculated, but will that discussion eventually result in a change? As with all things credit, only time will tell.

What Makes a Credit Score?

Credit scores typically range from 300 to 850 and are calculated using information from a person’s credit report, which includes data about their credit history, the amount of debt they owe, and delinquent payment information. It’s well known that paying certain bills on time and having a low balance-to-limit ratio will positively affect a credit score. But the exact algorithms used to calculate the figures aren’t known.

Ninety percent of lenders in the United States use the Fair Isaac Corporation’s scoring model, or “FICO scores,” as part of their decision-making, according to the company.

But there isn’t just one FICO score. Different FICO versions are used for auto lending, mortgages, and credit cards. The three major credit bureaus use different versions, too. In mortgage lending, for instance, Experian uses FICO Score 2, TransUnion uses 4, and Equifax uses 5, according to myFICO.com.

Last year, a new version of this credit-scoring model was released. FICO Score 9 was said to be the company’s most predictive model yet because it includes rent payments in the credit score calculation.

Community advocates, representatives of CDFIs and credit unions, and financial counselors have long touted the importance of looking to non-traditional credit like rent payments to evaluate a person’s creditworthiness.

“In the advocacy community, we’ve known for more than a decade that rent payments are much more predictive of how you’ll pay a mortgage than how you pay a credit card [is],” says Jim Carr, a Coleman A. Young Endowed Chair and professor in urban affairs at Wayne State University. “Yet the system has been very slow to . . . have a systematic report of rent payments.”

A couple of years ago, Experian become the first credit reporting agency to include positive rental payment data into a person’s credit history, according to a report the company released. Experian evaluated 20,000 leases using VantageScore 3.0, another credit scoring model, and 95 percent of renters experienced a score increase or a neutral impact. “Not only can the reporting . . . make a difference to those looking to build credit history, but it also may help thin-file or underbanked consumers become scoreable . . .,” the report indicated.

Disputed and past due payment information would not be included in a credit report because there isn’t a standard definition of “past due” in the rental sector, according to Carr. However, if past due rent is referred to a collection agency and then reported to the credit bureaus, that information would appear on a person’s credit report, as it does today.

“It’s unfair to have only the negative accounts reported by way of collections but not the positive payments, which is one of the reasons why rent reporting should be encouraged. Far more people would benefit,” Carr says.

Newer scoring models, like FICO 9 and VantageScore 3.0, could expand the lending market and directly affect the credit of African-American and Latino populations, says Carr, who is also a consultant for VantageScore Solutions. VantageScore estimates that mortgage lending to African-Americans and Latinos could be increased by 16 to 32 percent over 2013 levels.

However, there are a few issues. Rent payments must be reported to a credit bureau for that history to be part of a credit score analysis, but reporting is both voluntary and costs money.

What’s more, even if the monthly payments were reported to the credit bureaus, most lenders use outdated versions of FICO, which don’t take rental payments into account. It could take years for a company to update its system to utilize the newly released version of FICO, if they decide to do so at all, according to myFICO.com.

Carr disagrees. “It’s foot dragging,” he says. “I’ve heard estimates [to make the transition] in the hundreds of millions, and I’m saying if it takes you hundreds of millions of dollars to simply accept a score, then you need to fire your IT person because they don’t know what they’re doing. That makes no sense at all.”

Because the credit-scoring field has been FICO-driven for decades, a bill was introduced in the U.S. House of Representatives to allow Fannie Mae and Freddie Mac to consider using alternative scoring models when deciding what loans to purchase. Rep. Ed Royce, R-California, and Rep. Terri Sewell, D-Alabama, introduced the “Credit Score Competition Act of 2015” late last year, saying that the two government-sponsored enterprises (GSEs) rely on credit scoring models that don’t take into account simple items, such as rent payments. “Lower-to-middle income Americans [who] are qualified to buy a home but are unable to gain entry to the market due to their FICO score, or lack thereof, will specifically benefit from the GSEs using other credit scoring models,” a release from both representatives stated.

The bill was referred to the House Committee on Financial Services.

The Use and Abuse of Credit Scores

Throughout its history, the country’s credit scoring system has routinely been criticized. In 2012, it was deemed ineffective in a scathing analysis by the National Fair Housing Alliance (NFHA). Credit scores didn’t accurately predict the default rates of borrowers, the report concluded, and they “have a disparate impact on people and communities of color.”

While there is no mention of race on a credit report, the NFHA study, and many in the community development world, say the scores have racially disparate impacts for several reasons.

One is historical. Discriminatory rating systems used from the 1930s through the 1970s (even after the passage of the Fair Housing Act of 1968) labeled communities of color as “high risk.” “Segregated neighborhoods formed and people of color had limited access to affordable, sustainable credit. . . .[They] were relegated to using fringe lenders and paying much more than they would otherwise have had to,” the report stated.

Even if their credit score was above subprime levels, African-American and Latino communities were also targeted for subprime loans, which carry much higher interest rates and faster repayment periods. Because of this, those groups were more likely to have a poor payment history, the report indicates. (Payment history is a component of the credit score calculation.)

NFHA also observes that some fringe lenders only reported negative data to credit bureaus. Part of the credit score equation consists of the length of a borrower’s credit history, but if a lender doesn’t report positive information, longtime payments won’t appear on a borrower’s credit report. This has led to a cycle of thin credit histories, and more subprime loans.

Finally, some problems are baked into the scores themselves, says Carr. “If you have discriminatory modeling assumptions like, for instance, all borrowers must have credit cards open that they should be using regularly . . . you have been automating discrimination.”

While the NFHA report argued that credit scores do have a disparate impact on communities of color, a 2010 study by the Federal Reserve Board found no evidence of such, and FICO has also denied the claims.

Having lenders make decisions based on a calculated score is efficient, says Brenda Cude, a professor in the Department of Financial Planning, Housing and Consumer Economics at the University of Georgia.

The main alternative—individualized underwriting—takes time. For instance, it’s a four- to six-week process to review an application for a business loan at the Fresno Community Development Financial Institution, according to Credit Manager Jeremy Hofer.

Also, consistent use of credit scores could be a fairer way to evaluate borrowers, says Cude.

“I’m not sure we would want the opposite where someone, with all their biases and subjective notions, individually reviews people and says ‘you’re in’ or ‘you’re out.’ That sounds pretty scary to me,” says Cude, recalling a banker from her hometown who discriminated against certain groups of customers seeking loans.

Wayne State’s Carr is not quite sold on the need for credit scores at all, citing that home purchases ramped up to the greatest extent post World War II up until the late 1970s. “There was no such thing as a credit score. So the argument now that we have to have credit scores because you can’t make a quality loan [without them] is nonsense.”

Carr, also a visiting fellow for the Roosevelt Institute, a nonprofit think tank, said one of the main reasons lenders embraced credit scores so wholeheartedly was because they thought it would remove the chance of anyone citing discrimination as a reason for not getting a loan. But ironically, by using scores that have discriminatory modeling assumptions, lenders are discriminating by the thousands, he says.

Cude is ambivalent about credit scores. Though she believes they can be more objective than an individual review, she doesn’t believe credit scores are a particularly good indication of whether a person knows how to handle their finances. “When people talk about credit scores as indications of your financial responsibility, I think that’s a crock,” Cude says.

The professor teaches a money skills class, which requires that students obtain a copy of their credit report and estimate their FICO score. Cude has seen high scores for students who’ve piggybacked on their parents’ credit history, and low scores for students who never use credit, but pay their rent on time and have a savings account. Their scores are low because of their limited credit history. “They’re showing indications of being responsible that will never show up in a credit report or a credit score,” she says.

Cude doesn’t believe there’s a one-size-fits-all system. She believes in different points of entry for those seeking loans, much like what happens at credit unions and CDFIs, where representatives look at other factors besides credit scores to make a decision.

Building Credit

Some institutions are trying to counter the historical disparities in credit scores with targeted credit-building efforts. The Hartford Community Loan Fund, which serves low-wealth residents in Hartford, Connecticut, recognizes the correlation between credit scores and communities of color. CEO Rex Fowler says the nonprofit has secured the credit scores for the neighborhoods it serves and discovered that in the North End of Hartford, home to primarily low-income residents of African-American and West Indian descent, 55 percent have credit scores that don’t meet the minimum score (640) needed for the company’s construction rehab loan program. (The company does offer an asset-based program that will allow a score of 600 if there’s a “good explanation,” Fowler says.)

Fowler explains that they started their credit-builder program in order to help clients become “bankable”—able to access various types of affordable credit at a range of institutions. The program entails extending a small loan—anywhere from $100 to $1,000—expressly for the purpose of generating an official record of paying back credit.

There’s no minimum score to qualify for the program, but a resident must show they can pay the loan back within a year. In the past, the Hartford Community Loan Fund has provided a credit-builder loan of around $200 to allow a person to obtain a secured credit card at the “friendly” credit union across the street.

The loan and the secured credit card give a person two accounts, also known as trade lines of credit, because the two entities report payment information to the major credit bureaus, Fowler says.

Major lenders typically require three trade lines of credit —and a good credit score—before approving a loan, according to counselors at the National Affordable Housing Network (NAHN), a HUD-approved counseling agency that primarily helps those seeking homeownership. To help a person build traditional credit, counselors usually recommend opening a prepaid credit card with a local credit union or bank, as the Hartford Community Loan Fund has helped its credit-builder customers do.

“Usually once people have one or two lines of credit, more people are willing to trust them and extend them credit. But getting the first one or two can be very difficult,” says Dorothy Czehura, a credit counselor at NAHN.

The credit-builder program at Hartford Community Loan Fund has been in existence for about three years, but participation dwindled after two HUD-certified counseling providers closed in the last couple of years. The providers referred clients to the program.

Setting People Up to Succeed

Credit scores are used to decide not only whether to approve a loan, but also what interest rate to set, in what’s known as risk-based pricing. Luis Pastor, the CEO and president of the Latino Community Credit Union in rural North Carolina, thinks this makes no sense.

“You get the loan, or you don’t get the loan,” he says. “But we don’t charge you more because you have a poor credit history because no one can convince me that charging someone more is going to make it easier to pay back that loan.”

The credit union serves 62,000 members in a mainly low-income community of mostly Latino immigrants with non-traditional credit histories.

The loan application process begins with a personal interview, or what Pastor calls an “intensive” conversation. Credit union representatives then see if the applicant understands what it means to take out a loan, followed by a look at factors such as how long the applicant has worked at their current job and lived in their home. Credit union staff review water and utility bill payment information for those who do not have traditional credit histories.

The credit union began in 2000 after unbanked Latinos were targeted in a rash of robberies in the area. On a Friday night, most had all their payday earnings with them, says Pastor, as they hadn’t yet wired money to their families back in their native countries. These residents were viewed as “walking ATMs,” he says. Some residents were unbanked due to trust issues, language barriers, a lack of documentation, or the little-to-no experience they had with banking.

In Mexico, two-thirds of the population doesn’t have a checking account, so why would they be expected to open an account immediately after coming to the states, Pastor asks. Given all of this, credit scores aren’t very accurate for his customers. “Understanding the immigrant mentality is very relevant to understanding how we could qualify our members [for loans],” he says. “The system is set up in a way that’s not working for . . . immigrants coming to this country. Why would we use a system that’s not working?”

Part of the Mix

Credit scores do play a part in the decision-making at Fresno CDFI, which specifically works with small businesses in Central California. However, the company isn’t FICO driven and they don’t have a fixed cut-off credit score, according Hofer. Credit, collateral, and cash flow are all taken into account when deciding whether a person is approved for a business loan, Hofer says.

If an applicant has two out of the three aforementioned items, Fresno CDFI representatives are generally comfortable approving a loan request.

Hofer cited an example of a 26-year-old who sought a loan after getting an opportunity to take over a lucrative company. In the business acquisition world, he would have needed to have assets, a steady income, and good credit to make this happen. While this low-income applicant didn’t have the credit portion of the equation, he had somebody who was willing to back his dream and offer a down payment.

They gave him the loan.

Focus More on the Report, Not the Score

The two NAHN counselors Shelterforce spoke with know that when it comes to credit scores, there is systemic bias, but they generally don’t talk about it with clients.

“We focus more on what they can control and what they can do versus what they have no say over,” says Czehura.

When a person is getting ready to apply for a loan to purchase a home, NAHN counselors begin by pulling the person’s credit report. Then, depending on where the person needs assistance, counselors can help applicants with credit education and setting up a credit improvement plan, if needed. It’s really about the detail of the applicant’s credit report rather than the credit score itself, according to Czehura.

Typically, a person applying for a loan doesn’t know that several credit scores are pegged to their name because they usually find a free service that monitors one number. There have been instances where NAHN counselors have pulled up a person’s credit report and score that wasn’t even close to what the applicant had.

“Our purpose is really to educate individuals so they can not only improve their score to get into a house, but to continuously improve their score to benefit their life,” Czehura said.

It’s most difficult for a person between the ages of 30 and 40 to begin establishing credit, Czehura said. To begin the process, counselors suggest opening up a credit card for something the person needs, even if they could pay for it immediately. The worst type of credit a person can get is a department store credit card, said Sharon Chebul, a counselor who has been working at NAHN for more than a decade. Credit unions and banks are the best options, she says.

If a person doesn’t have three trade lines of credit, non-traditional credit like utility and rent payments can be used to prove that a borrower can pay back their debt. NAHN counselors have used non-traditional credit to help clients secure loans, but the payment history should be for more than two years.

There can be times, however, when even rent or utility payments are not good predictors. Someone not paying their water bill should be a warning sign for potential lenders, but if the person didn’t pay their bill because they live in Flint, Michigan, for example, where the water has poisoned residents, the non-payment makes more sense, Carr says.

Improving one’s credit takes time, and while Chebul believes that credit scores do reflect what’s on a person’s credit report, she, like Cude, doesn’t believe a credit score is indicative of a person’s trustworthiness.

“Sometimes it’s medical. Sometimes there’s no way to avoid it.”

(According to Carr, Vantage Score 3.0 gives less weight to charged-off medical debt. FICO 9 will also reduce the impact of overdue medical debt, which affects 43 million Americans, according to a Forbes article.)

Looking Ahead

In June 2016, Fannie Mae was expected to require lenders to use “trended credit data” when underwriting single-family borrowers through the GSE’s automated underwriting program. But in mid-June, the GSE postponed those plans due to technical issues with a new version of its underwriting program. (The company used to be FICO-driven, but it replaced credit scores with a proprietary credit risk assessment in 2000, according to a commentary from Eric Rosenblatt, the vice president of credit risk analytics and modeling at Fannie Mae.) Trended credit is a longer view of a borrower’s credit history, up to two and a half years. When used, it can show that one late payment, for example, was not reflective of a person’s ability to pay back debt, according to Fannie Mae’s website.

Meanwhile, CDFIs and credit unions that are willing to take a more nuanced approach to the role of credit scores in underwriting continue to sprout across the United States, and Hofer of Fresno CDFI believes those institutions are going to continue to grow.

“Anytime you have . . . a subprime market, you’re going to have . . . efforts that are community driven [to better serve that market],” he says.

While there are options available to obtain more predictive credit scores, there isn’t complete agreement on what lending institutions should use to determine credit-worthiness. However, many agree that a lot more scrutiny should be placed on the lending world.

Building or improving one’s credit, or waiting for a change to happen in the credit scoring world, is going to come down to one thing: time.

Editor’s note: This article has been updated to include VantageScore’s estimate about mortgage lending to African-Americans and Latinos, which could be increased by 16 to 32 percent over 2013 levels.

OTHER ARTICLES IN THIS ISSUE