Alternative Credit Scores: The Key to Financial Inclusion for Consumers

Transcription

1 WHITEPAPER Alternative Credit Scores: The Key to Financial Inclusion for Consumers May 2017

2 WHITEPAPER Alternative Credit Scores: The Key to Financial Inclusion for Consumers May 2017 Executive summary Consumer credit behavior has evolved over the past decade, along with the data and scoring systems that represented that behavior. As a result, many consumers have a difficult time getting access to affordable credit, or credit at all, due to no or poor traditional credit scores. ID Analytics examined credit applicant populations at ten key lenders to determine if the introduction of alternative credit data analysis tools would impact the eligibility of consumers categorized by traditional scoring models as thin-file, no-hit, marginal, or subprime for credit accounts and loans. The research analyzed data across 10 lending enterprises in the auto, telecommunication, credit card, and marketplace lending industries from , using the newest version of the company s credit score, Credit Optics Full Spectrum (FS). Part one of the study reveals that the inability to be evaluated by traditional scoring models restricts thin-file and no-hit consumers access to credit and how the addition of alternative data can dramatically increase the percentage of customers who would meet lenders eligibility requirements. In part two, the research demonstrates the impact of alternative data on marginal and subprime consumers whose creditworthiness is often underestimated from a traditional credit standpoint. This white paper shares the results of the research and includes multiple case studies and examples that demonstrate the ability of alternative data and modern modeling techniques to promote greater financial inclusion for many U.S. consumers.

3 2 Table of Contents Introduction 3 Study Methodology 4 Part One The Unscoreable Population: Thin-File and No-Hit Consumers 4 Part Two The Scoreable Population: Marginal and Subprime Consumers 7 Conclusion: Alternative Credit Scores Promote Financial Inclusion 10

4 Introduction 3 The Current State of Consumer Financial Inclusion The only constant in the consumer credit industry during the past decade was disruption. The financial crisis and new regulations forced many banks to tighten lending standards, while fintech innovation and marketplace lending provided new channels and sources of credit. Consumer credit behavior evolved, along with the data and scoring systems that represented that behavior. As a result, there s a gap in the U.S. credit ecosystem that leaves many individuals without access to traditional financial services. The majority of U.S. consumers rely on credit to achieve financial stability, but either have no credit score or a poor credit score that excludes them from obtaining traditional sources of financing. Nearly 20%, or 45 million, of U.S. adults fall outside the purview of traditional credit bureau scores¹ and another 32% of scoreable consumers have a poor credit score². Since you need credit to get credit many consumers are stuck in a Catch-22. Certain groups struggle with financial inclusion more than others, particularly young consumers, immigrants, and individuals with lower incomes. Research by ID Analytics reveals that one-third of millennials are unable to receive a score from a national consumer reporting agency and unscoreable applicants are four times more likely to be in the lowest income bracket than scoreable applicants. Consumers with poor scores can often be misclassified as marginal or subprime by traditional credit scores. Many of these individuals are formerly prime consumers who, often due to negative life events, struggled with credit responsibilities for a period of time but are now recovering. Traditional credit scores successfully predict whether known credit users are likely to default on their payments based on data typically taken from credit card, mortgage, student, and auto lending records. However, they lack information from key modern responsibilities that can provide additional predictive insights into a consumer s behavior. This incomplete perspective puts unscoreable consumers and those with poor scores at a disadvantage. In fact, an industry benchmark study of more than 300 lenders revealed that, of those surveyed, 58% agree that many consumers lacking traditional credit scores are low risk and 81% believe lowerincome borrowers need more help gaining access to financial services³. The White House Council of Economic Advisers (CEA) published a brief in June 2016 that supports the need for greater financial inclusion for consumers. Their findings showed that households that lack access to the mainstream financial system not only pay more for financial services, but also have little to no ability to build credit histories or cover emergency critical expenses such as rent, utilities, and mobile phone bills via low cost credit⁴. The CEA brief warns of an even broader impact for the macroeconomy: A lack of financial inclusion has the potential to hurt both equity and efficiency by reducing access to credit, which can be essential for entrepreneurship, homeownership, and economic development (Financial Inclusion in the U.S.). Alternative credit data and scores can help bridge the information gap left by traditional credit systems by providing the additional insight needed to more predictively evaluate the risk of nearly all U.S. consumers. 1 ² ³ state-of-alternative-data.pdf ⁴

5 Study Methodology 4 Is alternative credit data the missing link between underestimated consumers and access to credit? With this question in mind, ID Analytics conducted a study that quantifies the value alternative data can provide on both unscoreable (thin-file and no-hit) and scoreable (marginal and subprime) consumers eligibility for credit accounts and loans. The research analyzed data across 10 lending enterprises in the auto, telecommunication, credit card, and marketplace lending industries from , using the newest version of the company s credit score, Credit Optics Full Spectrum (FS). Credit Optics FS draws from the unique blend of traditional and alternative consumer credit data in the ID Network, one of the nation s largest networks of near real-time, cross-industry consumer behavioral data. The ID Network provides broader insights from data not typically analyzed in traditional bureau credit scores. This white paper reveals the results of the research and includes multiple case studies and examples that demonstrate the ability of alternative data and modern modeling techniques to promote greater financial inclusion for many U.S. consumers. Part One The Unscoreable Population: Thin-File and No-Hit Consumers According to the Consumer Financial Protection Bureau, there are approximately 45 million U.S. consumers who either have no credit history or lack sufficient information to generate a bureau score. Without a score, these individuals are identified as unknown, or high risk, to an enterprise and largely denied access to credit. High-risk consumers who are granted credit can expect to pay significantly higher interest rates or receive other less desirable terms and conditions on their loans. Of the 10 enterprise lenders covered in this study, 10-25% of their applicants were thin-file and no-hit consumers, meaning they could not be assigned a credit score by the national consumer reporting agencies. The research further revealed that not having a credit score had a significant impact on a consumer s ability to access credit. On average, these unscoreable applicants were only 25% as likely as scoreable consumers to gain access to the product or service they were applying for. The lack of a traditional credit history doesn t mean an individual is credit inactive. Consumers can build extensive history through modern credit responsibilities, such as marketplace lending and telecommunications, which fall outside of the national consumer reporting agencies visibility. This portion of the research illustrates how the incorporation of the additional insight, provided by alternative data and scoring, can improve financial inclusion for thin-file and no-hit consumers. Alternative Data Defined For purposes of this study, alternative data is consumer credit behavior not tracked by the national credit reporting agencies. Our research includes payment data from wireless, cable and utility accounts; online marketplace, payday and subprime lending; alternative billing methods; checking accounts; and other credit-relevant alternative data sources.

6 Marketplace Lending Insights Many applicants who are turned away from traditional credit look to online lending to fulfill their financial needs. ID Analytics examined the use of non-traditional loans at one of the top marketplace lenders to show the prevalence of consumer credit behavior outside the scope of traditional data. 5 Figure 1 depicts the percentage of unscoreable and scoreable applicants seeking online loans to purchase a car or home and to pay for medical and moving expenses. As demonstrated here, unscoreable consumers were more likely than scoreable consumers to apply for marketplace loans to fund life events across the board. Unscoreable vs Overall Population 12% 11% 10% 8% 8.2% 6% 4% 2% 4.6% 4.5% 2% 5.1% 2.1% 1.9% 0% Car House Medical Moving Unscoreable Scoreable Unscoreable applicants are: 2.4x s more likely to apply for a loan for a car 2.5xs more likely to apply for a loan for medical expenses 4.4xs more likely to apply for a loan to for moving costs 3.0xs more likely to apply for a loan to fund a vacation Figure 1. Unscoreable consumers apply for marketplace loans to fund life events at much higher percentages than scoreable consumers. These insights to marketplace lending suggest the benefit of alternative data to promote financial inclusion for consumers whose inability to be evaluated by traditional scoring models alone negatively impacts their access to credit. Alternative data is also relevant for the scoreable population. In this analysis, scoreable applicants were seven times more likely to apply for marketplace loans to pay off credit cards and three times more likely to apply for debt consolidation. From a risk perspective, this insight could signal the potential for a consumer to borrow beyond their credit capacity. Because not all marketplace lenders report ubiquitously to every traditional credit reporting agency, if at all, alternative data can be critical for gaining a complete picture of consumers modern credit responsibilities.

7 Telecommunications Insights Traditional credit scores are typically based on historical data taken from credit card, mortgage, student, and auto lending relationships. Thin-file and no-hit consumers who lack this type of history often exhibit modern consumer behaviors, such as telecommunications, that provide a more complete assessment of their creditworthiness. For example, only 33% of year olds have a credit card⁵, but nearly all (98%) have a cell phone⁶. Traditional credit scores frequently underestimate millennials creditworthiness due in part to their limited visibility into routine behaviors like cell phone bill payment. When unscoreable applicants were seen in the ID Network, for the purposes of this study, 70-90% had an account at a telecommunication provider. Telecommunications payment data can provide early insight into credit behavior for a large percentage of consumers, including millennials, who aren t being seen by traditional scoring models. 6 Additional Insights Promote Financial Inclusion for the Unscoreable Population Alternative credit scores with insight into non-traditional loans and modern credit responsibilities often identify a significant portion of unscoreable applicants as meeting lenders credit standards. In this research study, ID Analytics was able to predictively score 75% of no-hit and thin-file consumers with insights provided by alternative data. More than 80% of these individuals were seen outside of traditional credit bureau sources including wireless, cable and utility accounts; online marketplace loans, payday and subprime lending; and checking/savings accounts. Depending on the lender, 10-40% of their unscoreable applicants who did not activate would have been seen as credit eligible and added to the portfolio without an increase in the overall incidence rate. These results demonstrate a win-win situation for consumers and lenders alternative credit scores provide unscoreable consumers with improved access to the credit they need and deserve, while allowing lenders to grow their portfolios without incurring additional risk. Case Study: Scoring the Unscoreables This case study analyzed the portfolio of a top-10 U.S. credit card issuer to demonstrate how the application of an alternative scoring model could increase account activation rates for thin-file and no-hit consumers. The analysis found that 10% of the issuer s credit applicants were unable to be scored by traditional bureau standards. Using Credit Optics FS, ID Analytics identified 40% of the non-booked, unscoreable population as credit eligible based on the card issuer s acceptable incidence rate. 10 out every 100 applicants were not able to receive a score 2 out the 10 applicants were activated by the issuer 8 out of the 10 applicants were not activated 4 of the 8 were credit eligible 4 of the 8 were not credit eligible 4 of 100 applicants are considered credit eligible but not being accepted likely due to the offer associated with a bureau no score applicant This provider sees 5.6 million applicants annually meaning 224,000 unscoreable applicants could have been activated with no additional risk. ⁵ ⁶

8 This provider processes between 5 and 6 million applicants annually. Based on the results seen in this case study, 224,000 previously excluded applicants could have been activated, with no additional risk, by incorporating an alternative credit score. This means approximately 224,000 consumers annually at just one lender might have been granted access to credit, or significantly better priced credit, and are now able to build credit as part of the mainstream financial landscape. 7 In part one of the study, we revealed how the inability to be evaluated by traditional scoring models restricts consumers access to credit and how the addition of alternative data can dramatically increase the percentage of customers who would meet lenders eligibility requirements. In part two, we ll demonstrate the impact of alternative data on marginal and subprime consumers whose creditworthiness is often underestimated from a traditional credit standpoint. Part Two The Scoreable Population: Marginal and Subprime Consumers Out of the total number of U.S. consumers considered scoreable by the national consumer reporting agencies, 32% have a score that places them in a poor or bad credit class. For purposes of this study, subprime consumers were classified as those with a traditional credit score below 640. However, bureaus can miss the complete picture when it comes to understanding the creditworthiness of these applicants. Traditional scores offer only a partial view of consumer behavior and its associated risk, making it difficult for credit issuers to recognize recovering customers within a lender s subprime population. Marginal consumers who fall on the threshold of financial inclusion defined commonly as the nearprime population can be the most difficult to assess. The lenders in this study identified marginal applicants as those falling in a traditional score band of This narrow score band often makes it difficult to separate risk. Additional insight, from alternative data sources, can provide a more accurate assessment of consumers on the margin and help reveal the underestimated applicants in this population. Alternative credit information and scores capture many of the critical consumer behaviors that bureaus miss, allowing enterprises to see people who are on the path to credit recovery and increase consumer inclusion for underestimated applicants. Alternative Data Provides a Better Assessment of Marginal Consumers In this part of the study, ID Analytics looked at eight months of data and more than half a million records from a major retail card provider to demonstrate how an alternative credit score could increase consumer inclusion for the lender s marginal applicants. In this population, the marginal applicants fell within a traditional credit band of (Source: ID Analytics) Traditional bureau scores struggle to separate risk within such a narrow score band while alternative scores can help identify the percent of the population that is underestimated and safe to engage. This analysis revealed 40% of the population is below the retail issuer s overall incidence rate and considered credit eligible (figure 3). (Source: ID Analytics) The additional separation of risk using an alternative credit score allows lenders to identify marginal consumers whose ability and willingness to pay aren t wholly understood by traditional data and scores.

9 Marginal Risk Rank Ordering Bad Rate Bad Rate Decile Traditional Score Alternative Score Least Risky 7.02% 1.88% % 2.89% % 3.70% % 4.37% % 5.17% % 5.82% % 7.84% % 10.25% % 12.69% Riskiest Decile 8.80% 23.27% 8 Figure 2. The application of alternative scores to separate risk among the retailer s marginal population. Case Study: How Alternative Data and Scores Identify the Underestimated (Source: ID Analytics) This case study looked at a major marketplace lender with a score cut off and marginal population in the range. An analysis of one of the lender s applicants places this individual at the lower end of the marginal score band with a traditional bureau score of 646. When using an alternative credit score, the applicant s creditworthiness appeared to have been underestimated. Traditional bureau scores mainly look at five key elements oldest trade, total trades, total high credit, utilization, and delinquencies. The applicant analyzed in this case study showed the following attributes: Bureau Score: 646 Oldest Trade: 280 months Open Revolving Trades: 3 Total High Credit: $16,015 Utilization: 69% Delinquencies: 0 From a traditional credit standpoint, this consumer fell in the bottom third of all applicants seen at this lender, likely based on the high utilization rate and low number of open trades. In this case, the applicant did not receive a loan. When evaluated using an alternative credit score, the result was much different. Credit Optics FS scored this consumer in the top 6% of all applicants, indicating that the applicant s creditworthiness was significantly underestimated. How is this dramatic scoring difference possible? Additional alternative data insights revealed that this applicant opened an account at a telco enterprise more than two years ago, with zero missed payments over the life of the account. The additional perfect repayment history is important in assessing creditworthiness. Considering the majority of telco customers spend between $50-$150 per month⁷, this positive payment history would have benefited this marginal consumer when applying for the loan. This consumer had a broader and more consistent history of managing credit responsibilities than was communicated by bureau data alone the alternative data score revealed a credit eligible consumer who should have access to financing. ⁷

10 Recognizing Recovering Subprime Consumers Many applicants categorized as subprime have suffered negative life events or financial struggles, but are now recovering from past credit problems. As a result, they can also be underestimated by traditional scoring models, which are blind to many of the typical steps these consumers take to reestablish responsible credit management. Without access to traditional credit, these individuals often turn to marketplace or subprime lenders for financial support. Insight to these industries allows alternative credit scores to identify those consumers who are rebuilding their credit and have become credit eligible. ID Analytics looked at the portfolios of all 10 enterprises included in this study to determine the percentage of applicants who were considered subprime based on traditional credit bureau scoring methods. Subprime applicants were classified as those with a traditional bureau score below 640. Depending on the lender, the research identified as high as 50% of the enterprises applicants as subprime. Credit Optics FS was able to classify 14% of this subprime population as credit eligible. 9 Risk Rank ordering of Subprime Population 40% 35% 30% 25% 20% 15% 10% Credit Eligible 5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Case Study: How Alternative Credit Scores Identify Recovering Consumers (Source: ID Analytics) This case study illustrates how one applicant at a major marketplace lender was underestimated by a traditional credit score. This applicant received a traditional credit score of 612, placing them in the bottom 25% of all applicants clearly in the subprime portion of this lender s applicants. The consumer s bureau score reflected one delinquency in the past. As a result, this consumer was not offered a loan. When the applicant was assessed using Credit Optics FS, additional insight revealed a more complete picture of their credit history. More than two years ago, the same applicant was seen taking out an installment loan. The applicant had a low utilization rate on the loan of 36% and no new loans had been obtained. Based on this information, the alternative credit score put this applicant in the top 12% of all applicants at this lender and considers them to be a low credit risk. By adding an alternative credit score to the traditional bureau assessment, a lender can make a more informed decision based on the holistic view of the consumer s history.

11 Conclusion: Alternative Credit Scores Promote Financial Inclusion 10 This research study clearly shows there s an information gap in traditional credit scoring models that prevents a large percentage of U.S. consumers from achieving financial inclusion and that alternative data scores are an essential component in improving the current state of our credit ecosystem. Millennials aren t the only ones underestimated from a data perspective. When one-third of the nation s largest consumer generation can t be scored using traditional bureau data and a significant portion may be undervalued, there s an enormous opportunity to improve financial inclusion. The past 10 years have brought a lot of change to the consumer credit industry. As a consequence, traditional scores, while still important, no longer suffice on a stand-alone basis. The evidence of this study shows that better credit decisioning can be achieved using a combination of traditional and alternative credit data. This allows enterprises to be more inclusive with their lending strategies without increasing risk exposure. Financial inclusion for consumers will rely heavily on alternative data scores to identify individuals completely missing from the credit conversation and better serve those on the margin. About Credit Optics Full Spectrum In response to the increasingly dynamic marketplaces lenders and service providers compete in, ID Analytics introduced Credit Optics Full Spectrum (FS). Credit Optics FS is an FCRA-compliant credit score which provides predictive, incremental credit risk insights for the majority of through-the-door applicants. Credit Optics FS unique risk perspective is driven by the ID Network, a repository of consumer behavior data from a wider range of industries than other leading sources. By combining both traditional and alternative credit data, including telecommunications and online lending, Credit Optics FS delivers an assessment capable of dramatically improving credit decisioning on nearly every U.S. consumer. About ID Analytics LLC ID Analytics is a leader in consumer risk management with patented analytics, proven expertise and real-time insight into consumer behavior. By combining proprietary data from the ID Network one of the nation s largest networks of cross-industry consumer behavioral data with advanced science, ID Analytics provides in-depth visibility into identity risk and creditworthiness. Every day, many of the largest US companies and critical government agencies rely on ID Analytics to make risk-based decisions that enhance revenue, reduce fraud, drive cost savings and protect consumers. ID Analytics LLC is. Please visit us at ID Analytics is a registered trademark of ID Analytics, Inc. All other trademarks and registered trademarks are the property of their respective holders.

12 ID Analytics. All rights reserved.