Transforming auto lending beyond FICO: AI’s impact on expanding creditworthy customers

Auto lenders have been using the venerable FICO score to gauge customers’ creditworthiness for decades. But there’s a new kid in town: artificial intelligence (AI). And using it to assess credit risk can boost dealership sales by finding more customers that are eligible for loans – customers that would normally be disqualified by traditional credit-scoring methods. Here Amitay Kalmar, the Chief Executive Officer of Lendbuzz, a leading financial technology firm, explains how AI is changing the credit-assessment landscape – and how that can benefit dealers.

Q: Dealerships have been relying on FICO scores for years, so why should a dealership use another approach to assess the creditworthiness of potential vehicle buyers?
 

Amitay Kalmar: Traditionally, FICO scores have served one particular group of borrowers well: those who have built up multi-year credit histories here in the United States. For car dealers – especially those in a major metropolitan area or a state with a large migrant/immigrant population – that system leads to a lot of potential car buyers being either underscored or having no score at all. If you want to sell cars to those customers in a way that’s good for your business in both the short and long terms, you need to be able to score them accurately and then offer them a better deal – two things that traditional methods don’t enable.

Q: Artificial intelligence (AI) has been a topic of considerable conversation – and not always in a good way. Is using AI to assess credit risk better than the traditional approach?
 

Kalmar: Most industries are going to be impacted by AI in some way or another. Since it’s a very large industry, financial services are likely to be significantly impacted in several aspects. The core underwriting engines are the first ones being transformed and that’s where we are focused. Our use of AI is built around artificial intelligence risk analysis (AIRA), which can analyze significantly more data than humans – or the traditional FICO scoring method – can. Then we use that data to make a better decision on credit risk for specific consumers. Basically, when we use AI, we’re able to look at more data. This makes it possible for us to serve consumers that have been traditionally underserved by lenders because they were outside of the prime FICO ranges and provide them with a better solution.

Q: “Alternative data” and “machine-learning algorithms” are a mouthful compared to “FICO score.” Could you briefly demystify these terms to make this new approach to assessing credit risk easier to understand?
 

Kalmar: For us, “alternative data” means non-traditional information sources that can provide both deep and broad insights into a borrower’s creditworthiness. This includes a lot of what can be found in a person’s bank statements, like utility bill payments, rent payments and overall spending habits, together with employment history, educational background and more. Lenders can then paint a more comprehensive picture of an individual’s financial behavior and repayment ability beyond what’s captured by traditional credit scoring methods like FICO, which relies solely on the data reported to the credit bureaus.

Machine-learning algorithms are part of the underpinnings of an AI platform; they allow it to learn from data and rapidly make predictions or decisions. In the context of credit risk assessment, machine-learning algorithms – together with deep neural networks – allow us to analyze large volumes of historical data, including both traditional and alternative data points, to identify patterns and relationships that correlate with creditworthiness. This is what’s at the heart of our ability to provide a better level of service to the consumer and the dealer.

Q: Every new technology brings with it another risk – investing in the technology before it’s fully mature or investing in it too late. How does a dealer determine the right time for such an investment?
 

Kalmar: While AIRA is constantly becoming more and more sophisticated, it’s not something that’s new or unproven. It’s a maturing technology – one that’s built on a solid, well-documented and secure foundation, but with plenty of room for advancement. From an investment perspective, this technology seamlessly integrates into Dealertrack and RouteOne, the two platforms used by virtually every dealer in the industry. It’s similar to onboarding any other new lender to a dealership, in that it can happen quickly, and it doesn’t require dealers to invest in additional hardware or software. Given AIRA’s benefits, the biggest risk for most dealers will be in not adopting this technology.

Q: OK, so provide some proof of concept. How are dealerships that use AIRA to assess credit risk benefiting from the switch?
 

Kalmar: AIRA’s ability to accurately score thin and no-file borrowers results in many benefits for both dealers and their customers. First of all, it allows borrowers who would otherwise not be offered a loan – or offered one at maximum state rate – fast financing at a better rate – and for better vehicles. This results in a better overall experience for them, which also helps their relationships with dealerships.

In addition, AIRA unlocks more market segments for dealerships to sell and service to, including groups that they may have completely given up on in the past, due to their reliance on traditional credit scoring systems. More markets mean more potential sales – something that’s welcome no matter what kind of economic market dealers are experiencing.

Furthermore, AIRA reduces the amount of time and effort a dealer typically spends on labor-intensive tasks associated with stipulations, including verifying details about the borrower.

About the Panelist

Amitay Kalmar
CEO and Co-Founder, Lendbuzz
Before founding Lendbuzz, a financial technology company based in Boston, Kalmar was a Vice President at Deutsche Bank’s Technology Investment Banking practice, where he led mergers and acquisitions, initial public offerings and debt financing for tech companies. Prior to that, he served as a captain in the IDF, leading research and development teams for digital communications systems.

To learn more about Lendbuzz, visit www.lendbuzz.com.