Credit reports contain information that employers, landlords, and financial institutions use to make decisions such as who to hire and promote, who to rent to, and who to approve for credit and for how much. However, more than 45 million U.S. consumers lack sufficient credit history to either generate a credit report or a credit score (CFPB 2016; Lacagnina 2023). Consequently, these individuals may experience limited employment options, insufficient income, poor living conditions, and reliance on high-cost alternative financial products.
To address these negative consequences of traditional credit scoring, fintechs, credit bureaus, and financial institutions have begun collecting alternative data and developing new, supplemental scoring models to better assess the creditworthiness of consumers who were previously unscorable. This Payments System Research Briefing examines the effect of including alternative data in consumers’ credit reports; the role of fintechs, credit bureaus, and financial institutions in collecting alternative data and developing new scoring models; and potential concerns about the use of alternative data, such as adverse consumer outcomes, encroachment on privacy, and bias in scoring models.
Traditional credit records
To create a consumer’s credit report, credit bureaus collect information from credit accounts such as credit cards, student loans, installment loans, and mortgages. This information includes the types of credit accounts a consumer holds, the age of those accounts, the amounts owed, and the consumer’s payment history. With this information, they then create a credit score to evaluate consumers’ creditworthiness using models such as those offered by FICO (originally Fair, Isaac and Company) and VantageScore._
These models have strict requirements for generating a consumer credit score. To have a FICO credit score, for example, a consumer must have at least one credit account that has been opened for six months or longer and at least one credit account that has been reported to the credit bureau within the past six months (myFICO 2023). According to the Consumer Financial Protection Bureau (CFPB), nearly 26 million U.S. consumers were “credit invisible” in 2015 because they did not have credit cards, loans, or other lines of credit. An additional 19 million were “credit unscorable” because their credit history was so limited or dated that the credit bureaus were unable to determine their credit risk._
Although some products such as credit-builder loans and secured credit cards can help consumers with no or low credit scores build their credit history, these options also have barriers to entry. For example, both a credit-builder loan and a secured credit card require prepayments. For a credit-builder loan, consumers first must make fixed payments with interest to the lender. For a secured credit card, consumers must first deposit cash—usually equal to the credit limit—to open the account. These products are often costly for consumers. Moreover, if the consumer fails to make payments on time or defaults on the debt, they will build negative credit history, compounding their difficulties accessing credit.
Other options for accessing credit without a credit history require consumers to have a relationship with someone who has both the financial resources and willingness to absorb the risk of helping them build their credit. For example, some consumers can get a cosigner with a credit history to apply for a credit product with them. Other consumers may be able to piggyback on someone else’s credit account as an authorized user. However, not all lenders report authorized users to credit bureaus. Even if they do, some credit-scoring models do not count accounts on which an individual is only an authorized user.
Alternative credit records
To improve credit access and better predict future repayment behavior, some researchers have suggested looking beyond traditional credit records and using alternative data in credit-scoring models to supplement or support lending decisions (Campisi 2021). In practice, some fintechs, credit bureaus, and financial institutions have started collecting alternative data and developing new scoring models.
Alternative data fall into essentially one of two categories: other financial data and nonfinancial data. Other financial data include information such as bank account balances; assets; receipt or payment of child support; and rent, utility, and subscription service payments. Other financial data may also include nontraditional information that can be gleaned from consumers’ use of fintech services. For example, a consumer may work at will as an Uber or Lyft driver, cash out their fares to a debit card provided by a fintech company such as Chime, and use that debit card to pay their rent and utilities. Alternatively, a consumer may use Buy Now Pay Later (BNPL) services to purchase goods and services with credit without having to open a credit card account. None of those transactions would necessarily be visible in a traditional bank account or to a credit bureau but could nonetheless help establish a consumer’s credit history.
In contrast, other nonfinancial data include public records; educational history; employment history; personality traits; relationships with acquaintances, friends, and family; and digital footprint. Public records, for example, could reveal credit blemishes such as bankruptcies or liens that typically stem from a debt or delinquency. Education and employment history might provide information about a consumer’s income and ability to make payments. Personality traits and relationships might provide insights about a consumer’s lifestyle and spending habits. And a consumer’s digital footprint might confirm whether they are who they say they are and live where they say they live.
Fintechs’ role in record collection
Fintechs are using alternative data to help improve credit records in a few ways. First, they are offering consumers services that leverage alternative financial information to improve their credit scores. Second, they are offering consumers alternative credit products that can help them establish credit histories. Third, they are leveraging alternative data to facilitate the credit approval processes.
Several fintechs leverage alternative data such as rent payments to help consumers improve their credit scores. For example, RentPlus and RentReporters allow consumers to enroll in a credit-building service by providing information about themselves and their property manager or landlord. RentPlus and RentReporters then work with the property owner to automatically report rental payments to the credit bureaus. However, the services are not free: RentPlus charges the consumer a fee that varies based on the property management company, while RentReporters charges the consumer a one-time signup fee and a recurring fee for either a monthly or annual plan. MoCaFi, a free alternative financial services network for unbanked and underbanked consumers, also facilitates credit building through rent payments and allows users of its bill pay app to opt in to having their rent payments reported monthly to the credit bureaus. And Self, a provider of traditional credit-building services such as loans and secured credit cards, also enables consumers to use its app to link payments for rent, cell phone, or utility bills, which Self then reports to just one credit bureau, TransUnion, for a monthly fee.
Other fintechs offer consumers alternative credit products such as BNPL loans. BNPL loans are available to all consumers, including those with no or low credit scores, and provide credit in a predetermined number of installments of equal amounts. Although many BNPL loans, especially short-term loans, have not typically been reported to credit bureaus, credit bureaus are seeking to include consumers’ BNPL loans on credit reports (Paul 2022). In addition, BNPL provider Affirm is now partnering with FICO to build a credit-scoring model that will enable some BNPL loans to be factored into credit and lending decisions and to be reported to the credit reporting agencies (Mullen 2023).
Fintechs are also substituting algorithms and alternative data for in-person interaction between lenders and borrowers (Di Maggio, Ratnadiwakara, and Carmichael 2022). As a result of the data collected, these alternative underwriting models may also be able to identify individuals currently overlooked by standard measures of creditworthiness, such as credit scores. Mortgage originations are one area where fintechs’ use of alternative data has gained popularity. Online lender Quicken Loans is now the largest mortgage originator in the United States (S&P Global Market Intelligence 2020).
Credit bureau incorporation of alternative scoring models
Besides fintechs, credit bureaus are also leveraging alternative data to supplement the traditional credit reporting products they offer to financial institutions. The bureaus have developed some products in partnership with scoring model developers and fintechs and developed other products independently. All of these products require financial institutions to use specific credit-scoring models and thus may also require upgraded credit reporting software. In addition, some products require consumers to opt in to having their data collected through a third party.
In 2016, Equifax partnered with FICO and LexisNexis Risk Solutions, a data and technology provider, to create FICO Score XD. This product uses alternative financial data—such as public records and phone, cable, and utility payments histories—to generate scores for consumers who are credit invisible on the same scale used for standard FICO scores. The results of its pilot with many of the largest U.S. lenders found that most credit applicants who were previously unscorable could be scored and that as many as half of credit applicants received scores of at least 620—a common threshold for loan underwriting (HelpMeBuildCredit 2018).
In 2019, Experian partnered with FICO and Finicity, an open banking platform provider, to create an UltraFICO Score that factors in alternative data such as how long accounts have been open, the frequency and recency of bank account transactions, evidence of cash on hand, and history of positive account balances (Brown and Tarver 2021). Currently, consumers of participating financial institutions can grant permission for Finicity to access data from their checking, savings, and money market accounts and for prospective lenders to review their data in the decision-making process.
In addition to these products, Experian and Equifax also independently offer alternative credit-scoring models. In 2019, Experian introduced Experian Lift, which uses both traditional credit data and alternative data to help lenders mitigate risk and improve predictive capabilities for credit-invisible and thin-file consumers. That same year, Experian also introduced Experian Boost, a free service that allows consumers to contribute information directly to their Experian credit file by connecting their bank, credit card, or service provider to Boost. In particular, Experian will consider up to two years of positive payment history for consumers’ rent, internet, utility, phone, or streaming services bills in consumers’ Experian credit files._ In 2023, Equifax announced an expanded mortgage credit report that includes insights from consumers’ phone, television, and utility bills to determine creditworthiness.
Finally, VantageScore—a credit-score provider founded by Experian, Equifax, and Transunion—recently introduced an updated version of its scoring model that uses machine learning techniques to find patterns in the credit data of those who are credit unscorable to provide more accurate scores for this segment of the population. This updated scoring model eliminates requirements to have a minimum number of accounts and a minimum length of time for accounts to be open (DeNicola 2023).
Benefits and considerations
Thus far, studies suggest alternative data have advantages in determining consumers’ creditworthiness. A recent study finds that using alternative financial data to assess consumers’ creditworthiness resulted in both a higher probability of a consumer being approved for credit and a more favorable interest rate for that credit (Di Maggio, Ratnadiwakara, and Carmichael 2022). Another study from Andrew Davidson and Company, a provider of risk analytics and consulting for residential loans, finds that a consumer’s positive bill payment history is strongly correlated with a future positive mortgage payment performance; the correlation is even more pronounced when multiple types of bill payments, including mobile, cable, utility, and streaming service payments, were considered. This finding implies that positive bill payment history could streamline the underwriting process, help more consumers secure loans, and potentially lower mortgage rates for borrowers who might otherwise be overcharged (Equifax 2023). Finally, a study using German data finds that alternative nonfinancial data—namely, information that is easily collectable from a consumer’s digital footprint—can predict consumer default equally as well or better than credit records collected by credit bureaus (Berg and others 2018). However, the study also finds that using credit bureau scores and alternative nonfinancial data in concert predicts consumer default much better than using either alone, suggesting that information on a consumer’s digital footprint complements, rather than substitutes for, credit bureau data.
Providers and pilot programs also report that alternative data can improve credit reporting and thereby access to credit. Fintechs that leverage alternative data such as rent payments to help consumers improve their credit scores report that their customers can experience up to a 40-point increase in their score over a 12-month lease. Fintechs that offer alternative credit services have expanded the availability of credit to consumers with no or low credit scores. And both fintechs and credit bureaus that use alternative data in their underwriting models have found that alternative data not only helps identify individuals currently overlooked by standard measures of creditworthiness, but also results in credit scores that often align with traditional guidelines for extending credit.
However, using alternative data in credit decisions can also have negative effects for consumers. Once consumers agree to have their alternative data reported to a credit bureau, any delinquent payment or questionable activity may negatively affect a consumer’s credit report. Moreover, the CFPB has noted that some alternative data (such as social circles) may not be related to a person’s own financial conduct, and the use of these data could make it more difficult for people to improve their credit standing (Kreiswirth, Schoenrock, and Singh 2017). In addition, some types of nonfinancial alternative data, such as education history and social media usage, may not necessarily have a broadly agreed upon or empirically established relationship with creditworthiness and may be correlated with characteristics protected by fair lending laws (Prove 2021).
Indeed, many financial institutions and other lenders have not adopted alternative scoring models, suggesting they may be uncertain about the benefits of alternative data relative to the costs. Similarly, many consumers have not opted in to alternative data collection and credit-building practices, potentially due to privacy and other concerns.
Fintechs, credit bureaus, and financial institutions are collecting alternative data to develop scoring models that supplement traditional credit reports. While alternative data can improve both credit-scoring accuracy and credit access, adoption of these data is not yet standard due to several factors. Most alternative credit-scoring products require consumers to opt in to data collection and lenders to use particular scoring models and software; not all alternative data are clearly correlated with creditworthiness; and some uses of alternative data have the potential to negatively affect a consumer’s credit report. It may take some time for the benefits of alternative data to outweigh the costs to financial institutions. However, refining the types of alternative data used in credit scoring and expanding consumer knowledge and comfort with alternative credit-building products may help improve adoption of alternative data and consumers’ access to credit and financial well-being going forward.
Both FICO and VantageScore provide models for calculating credit scores based on information collected by credit-reporting agencies. However, VantageScore’s models can score more consumers than FICO’s models. Consumers who are new to credit or have not used their credit account for a while may not have a FICO credit score but may have a VantageScore credit score.
Toh (2023) examines the barrier traditional credit scores pose to obtaining affordable credit in the United States and discusses efforts to address that barrier.
Experian will look for bills or rent that have at least three payments in the last six months (including one payment within the last three months). Only online residential rent payments made to select property management companies or rent payment platforms are eligible for Boost. Rent payments made with cash, money order, personal check, or on a mobile payment transfer app (such as PayPal, Venmo, or Zelle) are not eligible.
Berg, Tobias, Valentin Burg, Ana Gombović, and Manju Puri. 2018. “External LinkOn the Rise of FinTechs—Credit Scoring Using Digital Footprints.” Federal Deposit Insurance Corporation, July.
Brown, Jerry, and Jordan Tarver. 2021. “External LinkWhat is UltraFICO?” Forbes Advisor, May 20.
Campisi, Natalie. 2021. “External LinkFrom Inherent Racial Bias to Incorrect Data—The Problems With Current Credit Scoring Models.” Forbes Advisor, February 26.
CFPB (Consumer Financial Protection Bureau). 2016.”External LinkWho Are the Credit Invisibles?” December.
DeNicola, Louis. 2023. “External LinkEverything You Need to Know about the VantageScore 4.0 Credit Scoring Model.” CreditKarma, June 6.
Di Maggio, Marco, Dimuthu Ratnadiwakara, and Don Carmichael. 2022. “External LinkInvisible Primes: Fintech Lending with Alternative Data.” National Bureau of Economic Research, working paper no. 29840, March.
HelpMeBuildCredit. 2018. “External LinkFICO XD—Introducing Alternate Data.” September 2.
Kreiswirth, Brian, Peter Schoenrock, and Pavneet Singh. 2017. “External LinkUsing Alternative Data to Evaluate Creditworthiness.” Consumer Financial Protection Bureau, February 16.
Lacagnina, Chris. 2023. “External LinkHow Many People Are Credit Invisible in the US?” Moneygeek, April 2.
Mullen, Caitlin. 2023. “External LinkAffirm, FICO to Create BNPL Credit-Scoring Model.” Payments Dive, May 10.
MyFICO. 2023. “External LinkWhat are the Minimum Requirements for a FICO Score?” MyFICO.com.
Paul, Trina. 2022. “External LinkBNPL Loans will Soon be on Your Credit Report: Here’s What You Need to Know.” CNBC Select, March 20.
Prove. 2021. “Six Biggest Problems with Alternative Data.” July 12.
S&P Global Market Intelligence. 2020. “External LinkQuicken Overtakes Wells Fargo as Nation’s No. 1 Mortgage Originator.” July 16.
Toh, Ying Lei. 2023. “Addressing Traditional Credit Scores as a Barrier to Accessing Affordable Credit.” Federal Reserve Bank of Kansas City, Economic Review, vol 108, no. 3.
Terri Bradford is a payments specialist at the Federal Reserve Bank of Kansas City. Ying Lei Toh, an economist at the bank, provided helpful comments. The views expressed are those of the author and do not necessarily reflect the positions of the Federal Reserve Bank of Kansas City or the Federal Reserve System.