Risk scoring platform offers unique loans to SMEs

Using alternative data to underwrite loans to small and medium enterprises (SMEs) is nothing new, whether it’s leveraging social media or using other non-traditional means.

Today, artificial intelligence (AI) is expanding the universe of alternative data sources – from bill payment history to community reputation – opening up new avenues for lending and giving financial institutions a new way to more inclusively assess the risk of lending money to SMEs and micro-enterprises.

While major lenders are still enamored with FICO scores and typical evidence of viability, the idea of ​​accessing credit risk through new types of trading signals and datasets is a new kind of lending to SME.

Like Uplinq Financial Technologies CEO and Founder Ron Benegbi recently told Karen Webster: “Today the majority of these small businesses are [owned by] minorities or immigrants, like me, and they formed because of the devastation that COVID has had. »

“They just set up Shopify stores, and they don’t have three years of financials,” Benegbi continued. They certainly don’t have the credit to establish themselves as a business. It’s a very, very difficult situation. »

This situation forces companies to see this risk from a new angle.

Benegbi explained that credit scores and financials “tell you a consequence, but they don’t tell you anything about the business itself and where the business is, the world it lives in, the community that lives there. surrounds, its suppliers, its partners, its customers.

Established lending channels also fail to account for pandemics, trade wars, and inflation when sizing unproven SMEs for lending. However, there are other “environmental cues” that work well when you combine them with an artificial intelligence (AI) specifically designed to read them.

Saying that the two met by chance, Benegbi notes that the AI-based decision platform that is now part of the Uplinq solution, originally developed by Pat Reilly, CEO of Verde International, is at the heart of a kind very different from commercial risk rating methodology.

Who benefits? “Our end customer is any organization that lends money,” he said, “however, we are looking to partner with a bureau, or… other FinTechs in the open API economy that serve ultimately some type of small business lender. Who benefits from our product is the small business lender themselves and, of course, the small business owner.”

See also: APIs unlock financing for underserved SMEs

Read Signals

Examining alternative data sources for business loan decisioning opens a fascinating window into how new insights and powerful AI can afford to elicit otherwise missing insights.

As an example, Benegbi explained that analyzing an SME’s electricity consumption over time and whether the utility bill was paid in a timely manner is a statistically revealing measure.

Looking only at electricity usage and payments, he said Uplinq finds “the likelihood that [the] the small business that defaults on a loan is cut in half “- from 8% to 4%” – but “it is very likely that this small business will be refused” for lack of a good FICO score or a sales history .

“This is one example, but we look at many different factors such as macroeconomics, stock markets, commodity markets, supply chain, labor markets, demographic information, traffic and community” , continued Benegbi.

Community reputation-based business lending decisions are another unique unconventional metric that Uplinq considers during the lending process.

“Community is a really big signal in a market like the Middle East, and it’s a really big signal in Latin America,” he said. “What the models are looking at is how active this small business is in their local community, because there’s an incredible cultural shame in not failing and scamming people in certain markets. So there are things like that.

Simply put, a small business that pays utility bills on time and is respected in the local community is less likely to default on a business loan.

Clarifying that Uplinq also takes into account FICO scores and any financial information available, Benegbi added that “thanks to the acquisition of AI/IP, we are integrated into the six largest central banking platforms in the world”, thanks to at which 95% of all loans are accepted worldwide, which provides additional insight into risk.

See also: B2B investors shine the spotlight on trade cards and alternative loans

Trust and Validation

As the company emerges from hiding, getting lenders to trust Uplinq’s data and SME lending risk assessments is a hurdle it faces, although Benegbi is unaffected.

“At the end of the day, [it’s] science and validation,” he told Webster. “It’s highly, highly scientific. We can go to a bank, to a lender and say, ‘If you don’t believe us, here’s the science, we’ll do all the back testing, we’ll prove it to you,’ whereas some analytics companies out there today are guessing. They don’t have that. Our data is regulatory compliant.

As an evolution and enrichment of risk scoring for commercial lending, Webster noted that large traditional banks and specialized FinTech lenders would both benefit from the approach.

“I’ve seen a lot of interest from parties that I didn’t think would initially be at the top of the list,” Benegbi said, citing credit bureaus and “perhaps one of the most world’s largest credit card issuers”.

“Organizations that we can partner with and ultimately bring a new product, a new data solution to their customers to transform the lives of millions upon millions of people around the world,” continued Benegbi. “That’s the impact we’re looking to have.”

See also: Data advances in SME financing beyond legacy business credit rating



On:More than half of US consumers believe biometric authentication methods are faster, more convenient and more reliable than passwords or PINs. So why do less than 10% use them? PYMNTS, in collaboration with Mitek, surveyed more than 2,200 consumers to better define this perception in relation to the usage gap and identify ways companies can increase usage.

Comments are closed.