According to a report by fintech company Fundera, about 48% of small businesses meet their financing needs, with 20% getting loans and 28% “lacking enough capital without a loan.” That means that 52% of all businesses get no financing or only get a portion of what they need.
Meanwhile, the U.S. Small Business Association (SBA) indicates that there are currently more than 34 million small businesses in the country. We’ll let you do the math (of how many of these small businesses are unlikely to get adequate financing).
Three veteran Live Oak Bank employees took notice of the scale of this problem and the lack of loans being distributed to small businesses. They realized that one issue is that most banks lack the modern technology to access underleveraged data that can determine how large a loan they can give.
Enter their solution: Lumos Data, a Wilmington-based data analytics venture designed to help financial institutions discover actionable insights through informed predictive models and thereby make accurate loan decisions.
Co-Founders Brett Caines (CEO), Youri Nelson (CTO) and Stephen Hayes were still at Live Oak Bank when they realized that the banking industry was in dire need of a better analytical solution. Nelson said that the primary mission of financial institutions is not to develop complex data analytics to inform lending decisions. Rather, they serve communities by accepting deposits and using those funds to provide loans to businesses and individuals.
This approach leads these institutions to make inadvertently biased decisions based on factors that don’t tell the whole story of each individual small business—such as the owners’ personal credit score or the lending history of how an industry did 15 years ago.
Nelson conveyed that rather than employing automated analytical solutions, institutions were using a more traditional underwriting workflow—relying on credit officers, high-level metrics, and past experiences of the underwriters.
“There’s an opportunity to use data science to be able to make better lending decisions, in terms of analyzing the risk that banks are taking with small businesses.”
With Lumos’s predictive models, institutions can not only take into account credit scores, but also three decades of an industry’s performance data, changing economic environments at both state and national levels, expected potential credit losses, how many employees that business has, and more. Also known as its PRIME+ scoring model, this solution from Lumos provides a fairer and more accurate risk assessment for small business loans under $500K.
Lumos in action
Lumos’s platform integrates as an API, where the financial institution will provide the platform with nondescript data from the company being assessed, and that data will feed into their predictive models. Then, Lumos’s platform will output a score, along with an expected loss and default probability.
Nelson said that Lumos’s fast, accurate and comprehensive data analysis is what makes them an appealing product offering to banks, even though the startup is only three years old. He said that comparisons have shown that their overall score performance is greater than the Statistical Product and Service Solutions (SPSS) scoring banks often use.
“We take into account a lot more factors than what traditional scores typically look at,” Nelson said. “In our analysis of historical data, we’ve seen that when banks were making smaller loans to small businesses, the decision was highly correlated to what that owner’s personal credit score was… If credit score is the primary thing banks look at, they’re missing out on fantastic business owners that have sacrificed their personal credit in order to make their business successful.”
On top of their predictive integrations, Lumos also offers actionable insights and tailored data advisory for financial institutions to make strategic decisions with greater clarity.
Lumos’s platform is currently being utilized by more than 90 unique banks across the country. They have even partnered with local fintech companies such as Wilmington-based nCino to streamline small business loan approvals even more quickly.
QUICK BITS
Startup: Lumos Data
Co-Founders: Brent Caines (CEO), Youri Nelson (CTO) and Stephen Hayes
Founded: 2021
Team size: 16+
Location: Wilmington, NC
Website: https://lumosdata.com/
Funding: closed a 5m round, not currently raising; running off subscription model
Nelson relayed that the core mission of Lumos is to shine a light on underserved, complex data. (Hence the name Lumos, if you’re a Harry Potter enthusiast!)
He said that he sees Lumos growing to become a de facto small business credit bureau, in the sense that for any data analytics needs that banks or small businesses may have have, Lumos will be there to turn that complex relationship into something that empowers entrepreneurs to do what they love.
“There’s plenty of small business owners that can’t access capital, and they’re having to turn to predator lenders and put themselves in jeopardy,” Nelson said. “I ask your audiences to understand why it’s important to support initiatives like this, because at the end of the day, it empowers the communities that we all live in, and it empowers the banks to help more entrepreneurs.”
