Alternative Credit Scoring in the US: Innovators Applying Data Science to Unlock Financial Potential of ‘Thin-File’ Individuals

March 1, 2017     By : Elena Mesropyan

Over the past 10 years, over a third (34.2%–36.9%) of the US population has been steadily categorized as having a subprime credit score (<620, which is considered to be a bad credit score). And even that part of the population should be considered ‘lucky’ in comparison to those not having a credit score at all. Meanwhile, credit score scales are used for all lending decisions in the formal financial sector, creating certain hurdles for subprime categorized population and those fully excluded from the formal sector.

In 2015, there were 26 million credit invisible consumers in the US, according to the report by Consumer Financial Protection Bureau (CFPB). Moreover, the estimates suggest that ~8% of the adult population has credit records that are considered unscorable based on a widely-used credit scoring model.

Richard Cordray, the CFPB Director, emphasized that “A limited credit history can create real barriers for consumers looking to access the credit that is often so essential to meaningful opportunity – to get an education, start a business, or buy a house. Further, some of the most economically vulnerable consumers are more likely to be credit invisible.”

Although those people are credit-invisible for the traditional sector, their everyday activity and alternative records represent a meaningful and vast source of precise hallmarks of their level of sustainability, resilience and credibility. Smart Data, hence, as opposed to Big Data, is bridging the gap for the un- and underbanked, plugging them into the formal sector based on sophisticated use of advanced technology and available records from alternative sources (social, web, etc.).

The following companies have been taking a different approach to trustworthiness evaluation in order to ‘humanize’ FICO and have developed their own algorithms and tools to make consumers’ lives better.

Alternative Credit Scoring in the US: Innovators Applying Data Science to Unlock Financial Potential of ‘Thin File’ Individuals

Cignifi has developed an analytic platform to deliver credit and marketing scores for consumers using mobile phone behavior data. Cignifi’s big data engine enables mobile network operators and their banking and insurance partners to identify qualified leads for savings, credit cards, insurance, personal loans and more from among tens of millions of pre-pay customers.

FactorTrust provides alternative credit data, analytics and risk scoring information that lenders need to make informed decisions about the consumers they want. The Atlanta-based company utilizes an internal team of predictive analytics experts and statisticians combined with a deep, proprietary database of underbanked consumer loan performance and best-in-class third-party data sources.

First Access provides a credit scoring solution for microfinance institutions and other emerging market lenders using data from consumers’ prepaid mobile phone history to assess applicant creditworthiness.

eCredable gives consumers a letter grade (A through F) based on their payment information that the company collects from the landlord, utility provider, insurance company or other vendors to verify timely payments. The report is called an AMP credit rating (where AMP stands for “all my payments”), and eCredable will share it with a potential landlord or lender with the consumer’s permission.

Happy Mango applies an aggregation technology to pull in financial information with the user’s permission and help lenders make credit assessments based on information not found in a credit report.

James by CrowdProcess (a scientific computing company behind the credit modeling tool) is a one-stop shop for credit risk management that allows to easily create, validate, deploy and monitor regulation-ready, high-performing predictive models. The solution claims to increase acceptance rate by 10% while decreasing the default rate by 30%. James’ cloud infrastructure is scalable by design, accommodating peak usage automatically. James can be easily integrated with other existing software or existing workflows.

Juvo is a mobile identity scoring platform to provide financial inclusion services to unbanked customers. Juvo uses algorithms to analyze both subscriber and usage data in real-time, to generate financial ID scores to unlock access to credit and basic financial services. The company also provides intelligence and reporting tools to offer mobile operators with actionable analytics and insights to drive their user engagement and growth. Juvo has a market footprint of 23 countries across four continents with a reach of over 350 million subscribers.

Neener Analytics has built a platform that correlates Personality & Behavior using their unique method by which to correlate social media data to score financial risk. To prove the model, Neener is the first to connect a consumer’s Facebook, LinkedIn or Twitter profile, with Personality & Behavior and an accurate FICO prediction. Neener can accurately predict/project a consumer’s fico score range almost 80% of the time. Neener is a web-based B2B (SaaS) that uses Personality & Behavior analysis to look at a consumer’s social media footprint to score financial risk for the underserved consumers who are thin-file, no-file or challenges (35-40% of US consumers) to repair the disconnect between this group and the businesses that want to serve them.

SharedLending is a platform where people create their unique CORE Profiles. The company’s algorithm has no connection or correlation with FICO. Their model analyzes responses to 5 human characteristics: Productivity, Resilience, Finance, Health and Education.

Trooly delivers Instant Trust™ services that verify, screen and predict trustworthy relationships and interactions. Working with financial institutions, peer-to-peer marketplaces, marketers and employers, the Instant Trust rating service is designed to fill a “trust gap” caused by the speed of modern commerce and community, which requires an instant evaluation of potential reward and risk – without the trust-building interaction history and feedback loops that people use to evaluate relationships offline.

TrustingSocial provides a consumer credit score based on social, web and mobile data. It provides three major scores on an applicant: 1. Authenticity: Measures how likely an online identity is real and trustworthy; 2. Network quality: Measures the depth of the network, based on interactions with other people and their background; 3. Financial Credibility: Estimates a person’s income percentile based on her work history and education.

ZestFinance develops big data underwriting technologies to give lenders a better understanding of risk. They have reinvented underwriting, enabling more accurate credit decisions, increased credit availability for borrowers and higher repayment rates for lenders. ZestFinance’s technology identifies good borrowers in the near-prime credit segment where traditional credit scores indicate lower credit quality, giving the company’s Basix loans the ability to serve borrowers overlooked by banks.

There are certainly more companies working toward a better understanding of what one’s creditworthiness really means and is comprised of. Examples include 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), ScoreLogix (uses employment and income data in specific zip codes), etc.

If we have missed some other interesting examples that you believe should be included, please reach out at follow@letstalkpayments.com or elena@letstalkpayments.com.

Elena Mesropyan
Follow me

Elena Mesropyan

Elena is a Market Research Analyst at LTP. She is a research professional with a background in social sciences and extensive experience in consumer behavior studies and marketing analytics. She is passionate about technologies enabling financial inclusion for underprivileged and vulnerable groups of the population around the world.
Elena Mesropyan
Follow me