Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Algorithmic Underwriting Is Reshaping Small Business Credit
Transcript
- Lucas: The traditional model for getting a small business loan is broken. You walk into a bank branch with five years of tax returns, two years of bank statements, and a credit score that’s barely above the median. The banker looks at you, says they’ll review it in thirty days, and then disappears into an email void. By the time they come back, your supplier has raised prices twice. Luna: And if you’re lucky, they say no because your inventory turnover doesn’t match their outdated risk matrix. It feels like we’re trying to solve twenty-first century cash flow problems with nineteen-ninety-five paperwork. Lucas: Exactly. And while that friction costs individual businesses millions in lost opportunities, the real shift happening right now is about who gets to play the game at all. We are seeing a massive migration away from static collateral toward dynamic, algorithmic underwriting. This isn't just about speed; it's about data granularity. Lenders are no longer asking 'what do you own?' They're asking 'how fast does your money move?' Luna: So instead of looking at a snapshot of your balance sheet, they’re watching the movie of your daily transactions? Lucas: Precisely. Let’s look at a concrete example from the logistics sector. There was a mid-sized regional distributor based in Ohio, let’s call them Midwest Freight Solutions, that needed three hundred thousand dollars for seasonal inventory buildup. Traditional banks rejected them because their profit margins were thin, even though their revenue was growing steadily. But their payment processor data showed consistent, high-volume transactions with major retailers. Luna: That sounds like a classic case of being profitable but illiquid. High volume, low margin. Lucas: Right. So they turned to a digital lender using algorithmic underwriting. The system didn’t care about their net income as much as it cared about their receivables velocity. It analyzed three years of daily sales data, looked at the consistency of customer payments, and even cross-referenced shipping API logs to verify that goods were actually moving. In forty-eight hours, they got approved for four hundred thousand dollars. Luna: That’s incredible. Forty-eight hours versus thirty days, and on better terms too? Lucas: On better terms, mostly. The interest rate was slightly higher than prime, but the cost of waiting for the bank would have been far greater. The key here is that the algorithm could see patterns humans miss. It noticed that their largest customers always paid within ten days, reducing their default risk profile significantly. A human underwriter might have seen thin margins and walked away. Luna: But doesn’t relying on algorithms create its own blind spots? What happens when the model encounters a situation it hasn’t seen before? Lucas: That’s the million-dollar question. Algorithms are only as good as the data they’re trained on. If the training data is biased or incomplete, the output will be too. We’ve seen instances where certain zip codes or industries were systematically underwritten because historical data showed higher defaults in those areas. It’s not malicious bias, it’s statistical bias. And fixing it requires active intervention by the fintech companies building these models. Luna: So it’s not just a tech problem, it’s an ethics problem wrapped in code. Lucas: Absolutely. Which brings us to the regulatory side. The Consumer Financial Protection Bureau has been cracking down on opaque decision-making in consumer credit. Now, those same scrutiny lenses are turning toward commercial lending. Banks are starting to ask: 'Can you explain why you denied this loan?' And if the answer is 'the black box said no,' that’s not going to fly anymore. Luna: Explainability is becoming a compliance requirement, not just a nice to have feature. Lucas: It is. And we’re seeing fintechs respond by building 'explainable AI' layers on top of their core models. They’re generating natural language reports that break down the decision factors. Did you get approved because of your transaction consistency? Or did you get denied because of a recent late payment to a vendor? Transparency builds trust, and trust drives adoption. Luna: I wonder if this levels the playing field for minority-owned businesses or if the data gaps still exist there. Lucas: It’s a complex picture. On one hand, algorithmic underwriting can reduce human bias. A computer doesn’t have a subconscious preference for a certain demographic. On the other hand, if those businesses don’t have enough digital transaction history, the model has less data to work with. It’s a catch-22. You need data to get funding, but you need funding to generate the digital trail. Luna: So the solution might be integrating non-traditional data sources like rent payments or utility bills into the credit profile earlier on. Lucas: Yes. Companies like Experian Boost have shown that including utility payments can raise a FICO score. For businesses, we’re seeing similar integrations with accounting software like QuickBooks or Xero. When you link your books directly to your lender, you’re creating a continuous stream of verified financial data. It’s less about the annual tax return and more about the monthly P&L statement. Luna: That makes sense. Continuous monitoring is more accurate than periodic snapshots. Lucas: It is. And it allows for dynamic credit limits. Instead of a fixed line of credit that you have to reapply for every year, some platforms adjust your limit in real-time based on your current performance. If you have a strong quarter, your borrowing power goes up automatically. If you slow down, it tightens. It’s fluid finance. Luna: That sounds empowering for businesses that are growing fast but unpredictably. Lucas: It is. But it also requires a shift in mindset for the business owner. You can’t just set your finances and forget them. You have to stay connected to your dashboard. The tools are there, but the discipline is on you. Luna: And for those listening who run operations, keeping your digital house in order is now just as important as having a good accountant. Lucas: Exactly. Your digital footprint is your new credit report. And speaking of keeping things running smoothly, I want to take a brief moment to talk about how this show stays independent. If these conversations are useful for what you're building or running, consider supporting us directly. A couple of dollars a month is genuinely what keeps these going. You can find us at buy me a coffee dot com slash fexingo. It helps us keep the lights on and the research deep, without advertisers dictating our topics. Luna: Yeah, it really makes a difference. Knowing that listeners value the depth of the analysis lets us dig into these niche topics rather than chasing generic trends. Lucas: Thanks for the support. Now, back to the data. One of the most interesting developments is the use of supply chain data to underwrite loans. Lenders are looking at the health of your suppliers and your customers, not just your own books. Luna: So it’s ecosystem-based underwriting. If your biggest client is solid, you’re considered solid. Lucas: Right. This is called reverse factoring or supply chain financing. But the new wave is broader. It’s analyzing the entire network. If you’re buying from reliable vendors and selling to stable retailers, the risk is distributed. It’s a holistic view of business health. Luna: That must require significant API integration capabilities from the fintech providers. Lucas: It does. And that’s where the tech infrastructure matters. The best platforms don’t just pull data; they validate it in real-time. They check against fraud databases, verify invoice authenticity, and ensure that the data hasn’t been tampered with. Security is paramount when you’re dealing with live financial flows. Luna: Security and transparency go hand in hand. If the data is trustworthy, the algorithm is trustworthy. Lucas: Spot on. And as we move into Q4 of 2026, we’re likely to see more consolidation in this space. Larger banks are either building these capabilities in-house or acquiring the fintechs that have already mastered them. The gap between traditional banks and pure-play fintechs is narrowing. Luna: Or maybe disappearing entirely. Hybrid models seem to be winning. Lucas: They are. And for the small business owner, that’s good news. More options, faster decisions, and hopefully, fairer access to capital. It’s a fundamental shift in how we think about creditworthiness. Luna: From asset-heavy to data-rich. It’s a brave new world for financing. Lucas: It is. And the businesses that adapt fastest to this new reality will be the ones that thrive. The question is, are you ready to let your data speak for you? Luna: Only if it’s saying something true. That’s the bottom line. Lucas: Agreed. Thanks for tuning in to this deep dive into algorithmic underwriting. We’ll be back next week with another look at the intersection of business and technology. Luna: Until then, keep your data clean and your cash flow flowing.