ALTERNATIVE CREDIT SCORES: opportunities and challenges
As new institutions step into banks’ traditional lending domain, alternative credit score models are causing disruption in the way loans are underwritten and applicants are assessed. Thanks to technology innovations focused on data science, revenue opportunities abound as new entrants look to serve a large, overlooked yet creditworthy market. In this article, Adi Ghosh and Harsh Sharma examine the changing landscape of lending; the opportunities for new entrants in the alternative credit score market; and the risks and challenges that emerging and traditional players face.
As the financial services industry looks to the future, there is widespread recognition to leverage technology innovation not only to improve compliance and operational efficiency, but also to create new sources of revenue—particularly in the lending market. One technology discipline that is receiving an increasing amount of attention is the application of big data and data science improvements to support alternative credit score models. These models will allow firms to extend traditional banking services to the underserved or the “thin credit file” market.
Changing Lending Trends
Collectively, post-crisis regulations have led to increased capital constraints, higher compliance costs and a reduction in spending for revenue-generating technology projects in traditional lending institutions. The result has created an opportunity for newer and nimbler entrants to the market, such as Zest Finance and Sofi—online lenders who are less constrained by regulations since they are non-depository institutions.
These institutions have stepped into typical traditional banking strongholds by lending to consumers who may not qualify under traditional underwriting models and standards. And they are leading the steep increase in peer-to-peer lending by utilizing alternative credit scores and models to provide access to funds to underserved small businesses and individuals.
New Market Entrants
At least 64 million consumers in the United States do not have a FICO credit score, according to Experian. But, 10 million of these so-called “unscoreable” consumers are prime or near-prime consumers, while a large portion of the remaining consumers have either professional jobs or low liability levels. Their income distribution is also in line with consumers who do have a FICO credit score. Clearly, there is a need to determine creditworthiness outside of traditional models.
To better serve this customer segment, several new alternative score providers have entered the market, each using different methods to calculate credit scores. These include:
- ScoreLogix: Uses employment and income data in specific zip codes
- Lenddo: Uses Facebook profiles to analyze risk
- Kabbage: Uses e-commerce histories from sites like Amazon for underwriting purposes
- Kreditech: Uses consumers’ online data and machine learning to provide access to higher credit and digital banking services
What’s more, the so-called “deposit behavior score” tracks how people manage their money by reviewing their checking and savings accounts for routine overdrafts, lack of payroll direct deposits and frequent or large intermonth balance fluctuations.
Determined not to be left behind, traditional credit bureaus have also started to develop their own alternative credit scores. TransUnion launched Credit Optics Plus to predict risk by tracking stability factors such as changes of address, cell phone service and other personal data. This score may benefit young people or recent immigrants who may have stable incomes and employment but thin credit histories. It could also penalize those who frequently relocate due to their profession.
Traditional lenders are responding to these new entrants by using alternative data in their own underwriting processes. Nearly two in three lenders reported tangible benefits within the first year of using alternative data.1 Additionally, existing credit bureaus such as Experian and TransUnion have already developed or are close to creating their own alternative credit score models to help lenders better serve consumers.
It may be some time before banks can emulate organizations like Sofi to transition to “FICO-free” credit scoring models, but there is definite merit in developing or leveraging existing alternative credit score models. In fact, new market players using alternative credit scores have realized several growth opportunities.
- New consumer segments: Consumers without a FICO score represent underserved and creditworthy homeowners, professionals, retirees and new immigrants.
- Differentiated customer experience: In addition to expanding the opportunity to offer tailored products and a friendlier customer experience, alternative data points and algorithms used for credit analysis allows the lender to streamline and automate loan applications—and deliver faster loan decisions to applicants.2
- Enhanced underwriting processes: The traditional underwriting process can be enhanced by many non-conventional variables such as credit card and banking transactions, social media presence, utility bills and employment history. This can lead to reduced credit risk through improved risk modeling and monitoring.
Challenges and Risks
The use of alternative credit scores presents tremendous opportunities. However, it is not without risks and challenges.
Regulations: A major challenge for any organization is federal regulation compliance. For example, organizations looking to use an alternative credit score must comply with the Equal Credit Opportunity Act, Fair Credit Reporting Act, Fair Housing Act, Gramm-Leach-Bliley Act and UDAAP, to name a few.
Data accuracy and objectivity: Keeping these regulations in mind, ensuring data accuracy and objective application of the methodology should be of paramount importance. Additionally, alternative credit scores are still new and their successful adoption have yet to proven on a universal scale. Meanwhile, FICO scores have been around since the 1950s.
Manipulation: While the ways in which an applicant can acquire a higher FICO score are well known, the wide range of alternative credit score models make it more complex for the average applicant. At the same time, alternative credit scores are more open to manipulation since an applicant could create the illusion of wealth or prominence through clever social network maneuvers, for example. To ensure accuracy and mitigate risk, a balance should be created between new models and traditional factors, such as income and savings.
Lenders can take three main steps before and after utilizing alternative credit scores to control their risks.
- Develop a strong compliance program before the implementation of alternative credit scores in the underwriting process to ensure compliance with all relevant regulations. Staff should be well trained and knowledgeable in order to quickly identify any risk areas and adapt to evolving regulations.
- Establish an enterprise-wide data governance structure that is cross-functional and includes representatives from all stakeholder groups with decision-making authority for accurate data and model calculation. In addition, procedures for ongoing analysis of data groups and their correlation to borrowers’ credit histories should be created.
- Conduct regular internal and external audits to ensure that procedures are followed without an adverse impact. Furthermore, champion-challenger testing should be implemented to validate changes to the organization’s decision logic and risk policy, and to try out any strategies on a representative sample before rolling it out across the entire business.
The use of alternative credit scores in the traditional underwriting process is still in its infancy. Organizations that have already implemented alternative credit models should periodically undertake research initiatives to gain additional insight and update models accordingly. Firms should also invest in research to compare performance across different models, including multiple alternative credit and FICO scores, to check for patterns and seek opportunities to synergize and improve scoring models.
In addition, increased collaboration across lenders, issuers, investors, technology partners and regulators is needed to aid in this initiative. An industry-wide group through appropriate associations and forums should be established where organizations with diverse knowledge and expertise can participate in knowledge sharing opportunities, resulting in more intensified research and development efforts.
Adi Ghosh is a Washington, DC-based Director focused on the primary and secondary housing finance market. He works closely with key industry participants to increase operational efficiency and bring strategic alignment in the housing finance space through technology solutions and value-driven partnerships. Adi has over 15 years of product development and business advisory experience across mortgage-backed securitization, loan origination, servicing and delinquency management.
Harsh Sharma is a Business Consultant with Sapient Global Markets based out of Washington D.C. Harsh has spent his career working in a variety of capacities within the financial services industry. He has worked on derivatives operating models and cash management operations for major financial institutions. Additionally, he has worked on multiple initiatives within the primary and secondary mortgage markets covering mortgage investment analysis, operations