One axiom of lending has held true for decades — a consumer's past behavior is the best predictor of his or her ability to repay a loan.
While that is as true today as it ever was, the data lenders use to assess creditworthiness is undergoing a sea change.
Consumers with tarnished or no credit history can now build their profiles by demonstrating stable cash flow, along with on-time rent and utility payments, to credit reporting firms. On the flip side, the big three credit reporting agencies, Equifax, Experian and TransUnion, have just begun excluding paid-off medical debt from consumer credit reports. New medical bills won't appear on a person's record for at least a year, up from six months, to allow for more time for repayment or negotiating payment plans.
And the recent boom in buy now/ pay later loans is also generating data that could help a lender evaluate a consumer's creditworthiness.
These changes hold the promise of democratizing credit, benefiting millions of consumers — including newly arrived immigrants — who were previously consigned to the subprime category because of past financial difficulties or thin credit files.
"With credit scores becoming more reflective of consumers' actual financial picture, to say we're entering a golden age of credit scoring innovations wouldn't be an understatement," said Silvio Tavares, president and CEO of VantageScore, one of two companies that considers data from multiple sources, including the three major credit reporting bureaus, to assess individual borrowers' creditworthiness with a three-digit score.
A golden age?
FICO, founded more than 60 years ago as Fair, Isaac & Co., and based in San Jose, California, pioneered the familiar three-digit credit scoring system, ranging from 300 to 850 in 1989. Use of FICO's credit score expanded significantly in the mid-1990s, after Fannie Mae and Freddie Mac began requiring them on mortgage applications. VantageScore, of Stamford, Connecticut, developed a competing version of the three-digit score in 2006.
Of course, the underlying reason FICO and VantageScore have grown so ubiquitous is that they work. They do a very good job of predicting consumer behavior and have helped tens of millions of consumers qualify for credit cards, auto loans and mortgages.
Problem is, they may have worked too well.
Millions of would-be borrowers who may have possessed adequate financial wherewithal have been locked out of the conventional credit box by the credit scoring system. Research published earlier this year by Experian and Oliver Wyman estimated nearly 106 million Americans either have files so thin they're rendered credit invisibles, or can only obtain credit at elevated subprime rates due to problematic histories. The National Foundation for Credit Counseling puts the number at 132 million Americans with no or poor credit.
This credit-impaired population includes disproportionate numbers of low-income, younger and minority consumers, according to the Consumer Financial Protection Bureau. Adjusting the data sets on which credit scores are based with an eye toward boosting inclusion could have dramatic benefits, making credit available to more consumers, even narrowing the racial homeownership gap, "something everyone wants to see happen," said Karan Kaul, a principal research associate in the Housing Finance Policy Center at the Urban Institute.
"People have been talking about using alternative credit data for a very long time," Kaul added. "The issue they ran into previously was that there was very little availability of such data."
To Tavares's point about a "golden age," things are changing rapidly. Technology is making it easier to collect alternative credit data, and institutions as disparate as community banks and fintechs as well as FICO and VantageScore are scrambling to put it to work. Like Tavares, Kaul contends a transformation involving the data used for credit scoring is underway, though he says it will be years before its effects become mainstream.
"A lot of things have changed in recent years to allow us to move from merely talking about this type of data to actually beginning to operationalize it, but it's going to take a very long time … for volumes to kick in," Kaul said.
The shift in the handling of medical bills — which took effect in July — can be seen as a start. It will affect about $88 billion in outstanding debt. Although large bills remain on the books, they will no longer cast such a large shadow over a borrower's immediate credit prospects.
"The way medical debt traditionally was handled — often caused by an unexpected event and handled by intermediaries like insurance companies — wasn't very predictive of consumers' actual creditworthiness," Tavares said.
Proving payments
Historically, credit scoring organizations have relied heavily on basic metrics such as income verification, public records and repayment data from lenders, including mortgages and home equity loans, credit cards, auto loans and student debt, in evaluating creditworthiness.
Defaulting on a loan or missing a rent payment could hurt a borrower's credit score with no immediate way to reverse the damage. And for the unbanked, building a credit score could prove to be extremely difficult.
For Wemimo Abbey, who co-founded the firm Esusu to help people build credit by documenting on-time rent payments to the credit reporting agencies, the sting of financial exclusion is hard to forget.
Immigrating to the U.S. at age 17 from Nigeria to attend college in Minnesota, Abbey and his mother were turned away from several banks when they tried to get a loan in 2009.
"My mother worked for 20 years at the post office in Nigeria, but she had no financial track record here," Abbey said, adding that ultimately the pair was forced to borrow money in costly installments to make ends meet.
Years after completing college and graduate school and working for a stint at Goldman Sachs and at a couple of other startups, Abbey founded Esusu in 2016 with a classmate, Samir Goel, to help consumers with thin or blemished credit records get a foothold.
Through Esusu, consumers opt in to have their on-time rental payments reported to the three major credit bureaus through their landlords, who pay a fee to Esusu.
"For consumers, this is a tangible way for them to expand their access to credit and cheaper debt, and for landlords it creates an incentive for renters to pay on time," Goel said. Landlords typically only report missed rent payments to credit bureaus, creating negative history consumers will likely have trouble correcting, he added.
While Esusu competes with about a dozen other companies in the emerging rent-reporting fintech niche, Abbey said the growth opportunity is large.
"Rent represents 35% to 55% of a consumer's monthly expenses and only about 10% of that data is being captured and fed to the credit bureaus in a meaningful way," Abbey said.
Indeed, Kaul of the Urban Institute described rent as one of the biggest issues in alternative credit. Getting on-time payments reported has the potential to aid millions of consumers who don't own their own homes but make their payments in a timely manner. Renters who live in properties owned and managed by large, multifamily operators often have their payments reported, "but a disproportionately large share of renters live in homes owned by mom-and-pop landlords" who don't report to credit bureaus, Kaul said. "That's where the framework breaks down."
Rent represents 35% to 55% of a consumer's monthly expenses and only about 10% of that data is being captured and fed to the credit bureaus in a meaningful way.
Checking the box
According to Kevin King, vice president of credit risk and marketing strategy at LexisNexis Risk Solutions, lenders will use alternative credit data on as many as 800 million consumer credit applications in 2022. Much of that data will come from publicly available sources, King added,
noting LexisNexis Risk Solutions collects information on asset ownership, education history, court records and consumer credit inquiries.
"If you think about it in that context, this is not something that is just finding its footing," King said.
There are new wrinkles, however. One of the biggest trends involves consumers giving permission to third parties to share payment flow data from their checking and savings accounts. For consumers who make their banking data available, there is no fee. Companies make money marketing the alternative scores they generate to lenders.
The practice holds enormous promise for allowing consumers with damaged and thin credit files to demonstrate positive payment history and savings behavior, though there are problems that have to be worked out and adoption to date has been limited.
Checking account data "is where we start getting into areas that are newer and have less adoption," King said. "That's an area, I think, that is more on the frontier."
FICO unveiled a credit score, FICO XD, aimed at the so-called credit invisible population in 2014. It's based on consumer permission data, primarily cable, cell phone and utility payments, as well as data collected from LexisNexis Risk Solutions detailing how often consumers changed addresses.
According to a spokesman, FICO has generated 4.5 million FICO XD scores over the past eight years for consumers who were previously unscorable.
FICO has also been piloting an even broader alternative score, UltraFICO, built around cash-flow data from consumers' checking accounts along with savings behavior. To be considered for an UltraFICO score, consumers must provide the lender with access to their checking and savings account history. To date, more than 35,000 consumers have shared their financial data and received UltraFICO scores.
"FICO is taking the learnings from this pilot to evaluate new ways to enhance the offering to help broaden financial inclusion through the use of banking data," the company's spokesman said.
Experian, too, has invested heavily in consumer-permission checking data, launching Experian Boost in 2019. The service aids consumers with problematic pasts by extracting transaction data from deposit accounts or credit cards in order to identify cell phone, utility and other
recurring payments. According to Experian, nearly 8 million consumers have seen their FICO scores rise after using Boost, with an average increase of 13 points. The feature has resulted in a total of 88 million points being added to credit scores nationwide, Experian said.
Jill Castilla, president and CEO of the $354 million- asset Citizens Bank of Edmond in Oklahoma, said a family member who used Boost saw their credit score jump significantly. "It was very effective and will provide them with the ability to build a strong credit score without having to apply for credit they may not be ready for," she said.
As a banker, Castilla said she was open to alternative credit data, "since the current system is far from perfect, both from the credit provider and consumer perspectives," she said.
There is broad agreement that alternative data gleaned from checking accounts is valuable, King said. "There's really no debate on that piece," he added. "Everybody wants this."
Kate Hao, CEO of the New York-based fintech Happy Mango, called alternative credit data, particularly cash-flow information gleaned from primary checking accounts, "a necessary complement" to traditional credit scores. Alternative data can help identify individuals with what Hao termed "artificially high" credit scores people with unblemished payment histories who may have just lost a job or source of income.
"Additional cash-flow analysis can help filter out people who can't afford the credit they're applying for," Hao said. "That should be a very powerful tool."
King, however, said that for alternative data to be really effective, it has to be collected broadly across the full marketplace, a standard that may take years to achieve.
King added that it could be years before services like Experian Boost, FICO XD and UltraFICO make serious headway in reducing the number of credit-damaged and credit-invisible consumers.
"The kind of person who sees an ad on television or hears it on the radio and takes the initiative to go online and is willing to share their information tends to be the kind of consumer that has diligently taken care of their credit," King said. "You're not necessarily getting the kind of consumer … you would want for financial inclusion. You're getting more information on consumers you already know pretty well. That's the challenge."
Barring a comprehensive campaign aimed at persuading more people to grant access to their checking and savings data, "this journey is going to take quite some time," King said.
Machine power
Along with tapping new streams of data about consumers making on-time rent and other routine bill payments, fintechs and other lenders are developing their own data models to evaluate consumers who fall outside the scope of traditional credit reporting agencies.
One example is Nova Credit, a San Francisco-based firm founded in 2016 that has gained traction with a service it provides to other lenders that translates immigrants' credit histories from their home countries to an equivalent score in the U.S.
Another is Mission Lane, a fast-growing startup spun off from LendUp in 2018. Over the past few years the Richmond, Virginia, firm has extended credit lines to 2 million consumers with tarnished or limited credit history using a proprietary underwriting formula. The system analyzes
diverse types of payment and loan data with machine learning plus consumer input to determine the creditworthiness of mostly underbanked consumers.
Similarly, the $1.2 billion-asset Armed Forces Bank in Fort Leavenworth, Kansas, developed an in-house underwriting model that relies on alternative data such as rent, utility and telecom payments to help it make personal loans of $750 to $15,000 to younger enlisted soldiers, a group that payday lenders and other more predatory providers have historically targeted.
Armed Forces' personal loan portfolio, which totaled $5.4 million at March 31, 2021, rose to $22.1 million a year later. That proved successful enough to prompt Paul Holewinski, CEO of Armed Forces' parent company, Dickinson Financial, to plan on expanding the program to Dickinson's other bank subsidiary, the $2.3 billion-asset Academy Bank in Kansas City, Missouri.
According to King, it's "much less common" for a community bank to develop a proprietary underwriting model based on alternative credit data. Despite Armed Forces' results, King said he doesn't expect to see many institutions follow in its footsteps. Most will be content to rely on data from third parties, and "rigorously test them, challenge them and leverage the very best ones."
Hao, a former bond trader with Morgan Stanley, founded Happy Mango in 2014 to bring greater transparency to consumer credit reporting. Her New York-based fintech collects and analyzes clients' income and spending data to provide lenders with a fuller picture of their creditworthiness.
"We wanted to create a better tool for lenders to assess credit and avoid the disasters we had" during the financial crisis, Hao said.
Assisting so-called credit invisibles, including newly arrived immigrants, has become a big part of Happy Mango's business. Some of its newest clients include recently arrived immigrants from Afghanistan.
Most immigrants "are financially literate," Hao said.
"With a little help from coaches about how the U.S. financial system works, they can improve their profiles. They can use the tools available to them very effectively," Hao added.
Moving slowly and cautiously, VantageScore and FICO are now working with the three major credit reporting agencies to access additional streams of alternative data, often drawing on new consumer permissioned-data approaches coming from the global open-banking movement.
Teaming up with Pentadata, a Burlingame, California- based financial data platform, and Akoya, a data aggregator based in Boston, VantageScore will soon be able to tap into consumer data from several of the largest U.S. banks. That should expand its funnel for alternative financial data consumers may directly supply to enrich their credit records.
"There has been no fast track for immigrants coming to the U.S. to prove their creditworthiness, even if they have a long and impeccable credit record in their home country," VantageScore's Tavares said.
FICO is working with the consumer financial data aggregator Finicity to augment the company's traditional credit scores with new streams of alternative data consumers may directly add to their credit records through the three credit reporting agencies.
"A lot of immigrants and young people have data that fall outside of our traditional credit scores and we're creating pathways for them to add their own data to build up a score or repair blemished scores," said Sally Taylor, vice president and general manager of scores at FICO.
A two-way street
Through its U.S. base in Costa Mesa, California, Experian earlier this year launched Experian Go in the U.S., which targets an estimated 28 million credit invisibles. It creates a simplified pathway for them to begin building their own Experian credit report without going into debt.
A consumer with no credit history can set up an Experian membership and begin creating a credit report within minutes.
Eligible users are invited to connect to Experian Boost, which enables them to add cellphone, utility or video streaming services directly to their Experian credit report, said Greg Wright, chief product officer for Experian Consumer Information Services.
"We believe that every individual deserves the opportunity to reach their fullest financial potential and have access to fair and affordable credit and we're continuing to look for more ways to enable lenders with the right technology, data and analytics to score consumers and empower them to play an active role in their financial health," Wright said.
In addition to consumers gaining the ability to help build their own credit histories, digital technology enhancements have made it easier for consumers to easily monitor and track their credit scores.
For years, credit reporting companies only allowed consumers one free credit check a year, but in the wake of widespread data breaches, many financial services provide free credit monitoring to individuals whose data had been exposed.
Those practices expanded free access to credit scores and reports offered free as a marketing service from some banks, credit card issuers and financial services providers like Credit Karma.
"There's been a massive change over the last five years as banks, credit card issuers and others have made credit scores more accessible to consumers, and people have become much more savvy about credit and are learning how to manage it and maintain healthy financial habits," said Liz Pagel, senior vice president, consumer lending business leader at TransUnion.
BNPL on the radar
VantageScore and FICO each update the algorithm used in generating their three-digit scores approximately every three to five years, so it will take time for the full array of emerging alternative payment-data streams to make their way into the credit scoring systems banks and other lenders use to grant credit, according to FICO's Taylor.
"A lot will depend on how the credit reporting agencies feed consumer data to us," she said.
Momentum is accelerating for BNPL firms in the booming installment-loan arena to begin sharing borrower data with credit reporting companies to expand transparency around the risk these loans pose.
The Consumer Financial Protection Bureau said in a June blog post that it's become important for BNPL lenders to furnish both negative and positive data about borrowers' loans and repayment history for the overall health of the financial services ecosystem.
The CFPB has urged buy now/pay later lenders to adopt standardized loan codes and formats and suggested that credit reporting companies incorporate this data into their core credit files as soon as possible.
TransUnion, Experian and Equifax have each announced different road maps to begin accepting certain BNPL payment data from point-of-sale lenders. Although many BNPL firms are not yet reporting borrower data, the stage has been set for companies to take voluntary steps to share relevant elements of this data to benefit the broad lending ecosystem.
TransUnion conducted a two-year study of 9 million BNPL borrowers, which ended in late 2021, that found most point-of-sale loan applicants are more likely to be younger with thin or no credit history. That means these borrowers are likely considered to be below prime.
In upcoming months, TransUnion will begin enabling BNPL loans to be received through the traditional credit reporting process. That should give financial institutions the option to consider these trade lines as part of their existing credit data information.
Experian is working with the largest BNPL providers to collect data on all types of point-of-sale loans to create a comprehensive view of consumer payments, including the status of BNPL loans that have been paid and are outstanding.
Equifax sees the opportunity to incorporate data from BNPL fintechs like Affirm, Klarna, Afterpay, Sezzle and Zip into its consumer reports as a boon for the overall consumer credit industry, according to Sid Singh, president of Equifax U.S. Information Solutions.
"Consumers who are paying BNPL loans on time should get credit for paying them, but lenders should also have visibility into how many consumers might be overextended with loans that aren't reported," Singh said.
Equifax created a new category to display a consumer's payment history with various BNPL providers. That will provide lenders and credit-scoring partners with more accurate information about prospective borrowers' wherewithal to repay new loans.
Although BNPL lenders and credit reporting firms are still in the early stages of determining how exactly to apply point-of-sale loan data to consumers' credit files, consumer advocates are concerned about potential negative effects.
The use of rent data in credit models is allowing people with low credit scores or thin credit files to get loans. But consumer advocates warn the use of that data could harm the very people it’s supposed to help.
The CFPB's proposed standardized approach to reporting BNPL loan data as a revolving credit trade line on individuals' core reports could be a positive step, according to Ed Mierzwinski, senior director of federal consumer program at U.S. PIRG, a consumer advocacy nonprofit based in Washington.
"The worst case would be a BNPL reporting system only 'good' enough for lenders to figure out how to keep consumers on a BNPL hamster wheel — able to buy more impulse BNPL stuff that they can't afford without improving their scores enough to eventually qualify for decent revolving credit cards or affordable installment credit to take out, for example, car loans," Mierzwinski said.
BNPL data is only one area where Equifax plans to expand its consumer-credit analysis. Over the past 18 months the Atlanta-based company has purchased 11 different companies to expand its access to alternative consumer credit data to make its credit assessments on its 136 million consumers more precise and up-tothe-moment, Singh said.
"The boom in online commerce has created more signals to capture consumer traditional and alternative financial data, which combined with advanced analytics is moving us closer to getting a 360-degree view of more and more consumers," Singh said.
