Identifying opportunities and overcoming challenges for the thin file consumer

In today’s digital age, data is often referred to as the new gold. Financial institutions, lenders, and businesses rely heavily on consumer data to make informed decisions. However, a significant portion of the population falls into a category known as “thin file” consumers. These individuals have limited or no credit history, making it difficult for them to access essential financial services. But there is hope on the horizon, thanks to the power of artificial intelligence (AI) and machine learning (ML). 

‍

The thin file challenge

Thin file consumers, also known as credit invisible or credit unscorable, are individuals who don’t have a sufficient credit history to generate a traditional credit score. This lack of credit history can result from various factors, such as being young, recent immigrants, or simply not having a history of using credit products. Unfortunately, without a credit score, these individuals often face significant barriers to accessing financial services, including credit cards, loans, and even housing. There are 62 million thin file credit consumers, (22 percent of the adult population) according to Experian. 

Credit unions face the risk of losing the most crucial demographic “Generation Z” due to their inability to extend credit to thin file consumers. Generation Z is an important group for credit unions because of their size, potential for long-term memberships, alignment with digital services, shared values, and diversity. 

‍

How AI can help thin file consumers

Artificial intelligence, specifically machine learning, has the potential to transform the landscape for thin file consumers by addressing the challenges they face. Here are some key ways in which machine learning can make a difference: 

Predictive modeling: Machine learning algorithms can predict risk for thin file consumers by being able to dig deeper into more data and analyze it using more sophisticated math. These models can more accurately assess an individual’s likelihood of defaulting on a loan even when data is missing from a credit file, helping lenders make more informed decisions about extending credit. 

‍Alternative data analysis: According to an article from the Federal Reserve Bank of Kansas City, alternative data falls into essentially one of two categories: other financial data and non-financial data. Other financial data include information such as bank account balances, assets, receipt or payment of child support, and rent, utility, and subscription services payments. Nonfinancial alternative data could include public records, employment history, and digital footprint. Responsible use of AI can incorporate additional data sources to enhance accuracy, identify what types of data will help predict risk, and address compliance-related challenges with some alternative data. 

‍Financial inclusion: AI-powered tools can enable financial institutions to offer tailored financial products to thin file consumers. By assessing risk more accurately, and by removing bias, these institutions can create products that cater to the unique needs of this underserved population. 

‍

Challenges and ethical considerations

While alternative credit data hold great promise, they also raise concerns about privacy, data security, and potential bias in algorithms. Striking a balance between increased financial inclusion and protecting consumers’ rights and interests is a crucial challenge that the industry must navigate.

Many fintech companies are relying on alternative credit scoring models to determine whether or not to extend credit to borrowers. However, there are challenges to the widespread adoption of alternative credit data: 

• ‍Data quality and reliability – The accuracy of alternative credit data can vary. Some data sources may have errors or lead to misinterpretation
• Compliance risks – The Fair Credit Reporting Act (FCRA) and Equal Credit Opportunity Act (ECOA) impose specific requirements on credit reporting agencies and entities that use credit reports or other sources for decision-making. When using alternative data, organizations must ensure they adhere to FCRA and ECOA guidelines
• Privacy and security concerns –Using data like website activity or social media history can present privacy concerns. Financial institutions must also make a plan to ensure privacy and security
• Ethical concerns – The use of some alternative credit data sources, like social media profiles or personal information, can raise ethical concerns related to data collection or introducing biases into the process

Alternative data can also be expensive and may not produce the accuracy or measurable ROI for the expense. 

While alternative data can be beneficial to level up some credit decisioning, financial service organizations also need more automation, more sophisticated processes, more forward-looking predictions, and greater speed-to-decisioning. And to this end, they need machine learning. 

With the power of machine learning, it’s possible to gain profound insights into the thin file population with less alternative data than one might expect. Machine learning can be leveraged strategically and transparently to ensure not only regulatory compliance but also to establish a track record of accuracy and fairness in lending practices. 

It’s crucial to recognize that without a proven and robust machine learning tool to guarantee accuracy, compliance, and transparency, the risks associated with utilizing alternative data can be significant. Therefore, harnessing the potential of machine learning is essential to mitigate these risks and realize the full potential of alternative data for the benefit of thin file consumers. 

Thin file consumers deserve equal access to financial services and opportunities for financial growth. AI technologies have the potential to level the playing field by providing a more accurate assessment of creditworthiness and tailoring financial products to their unique needs — in a compliant and responsible way. As the financial industry continues to embrace AI, we can look forward to a future where thin file consumers are no longer excluded but empowered to take control of their financial lives. 

___________________________________________________

Denise Wymore is an inductee to America’s Credit Union Museum and a cheerleader for passion and commitment. Currently, she is the Marketing Manager for Small Credit Union Initiatives at Zest AI and is proud to be a credit union lifer who started her career as a teller.