Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Fintechs Use Real-Time Expense Data for Corporate Card Limits
Transcript
- Lucas: If these conversations are useful for what you're building or running, that's exactly why we make them. But there's a quiet reality behind this show. Luna: Go on. Lucas: A handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making this many of these. No ads, no sponsors — just people who find value and toss in a few bucks. Luna: Yeah, it's a small group, but it keeps us independent. Anyway, back to the topic — corporate card limits. I've been curious how Brex actually sets those limits day to day. Lucas: That's the heart of this episode. Traditional corporate cards — think Amex or Chase — they set a static limit based on your credit score and some historical data. Brex does something fundamentally different. Luna: They use real-time data, right? Not just your credit report. Lucas: Exactly. Brex connects directly to your company's bank accounts, accounting software, and even your payroll provider. So they see your cash balance, your burn rate, your recurring subscriptions, your revenue inflows — all in near real-time. Then they adjust your spending limit dynamically, sometimes daily. Luna: That feels invasive but also incredibly useful. If I'm a startup with a lumpy cash flow, I'd rather have a limit that expands after a funding round than be stuck at a low ceiling. Lucas: Right. And Brex's model is built for that. They launched in 2017 targeting venture-backed startups. Their underwriting is less about personal credit and more about the company's cash position and runway. The limit can go up or down in 24 hours. Luna: So what's the tech stack behind that? How do they pull data fast enough to update limits daily? Lucas: They use Plaid for bank account linking — that gives them read-only access to transaction history and balances. Then they integrate with accounting platforms like QuickBooks or Xero, and payroll systems like Gusto or Rippling. All that data feeds into a machine learning model that predicts your future cash position. Luna: And the model decides the limit based on that prediction? Lucas: Essentially. They set a limit that's a fraction of your available cash, but also factors in your spending patterns. If you consistently spend $50,000 a month on software subscriptions and ad spend, they might give you a limit of $100,000 — higher than your average monthly spend, but not so high that you could drain your account in a week. Luna: What about the risk of false positives? Say a startup has a big inflow from a customer payment, then the limit spikes, but the customer reverses the charge. Lucas: That's a real risk. Brex's model has to smooth out noise. They look at rolling averages, not just a single day's balance. They also monitor for anomalies — like a sudden spike in spending that doesn't match historical patterns. Luna: So it's not just a simple formula of 'you have $1 million, your limit is $200,000'. It's more nuanced. Lucas: Much more. They also consider your industry. A SaaS company with recurring revenue gets a different treatment than a hardware startup that has lumpy inventory purchases. The model is trained on thousands of companies, so it learns patterns. Luna: Have we seen any data on how often limits actually change? Is it daily for everyone? Lucas: Brex says limits update at least every 24 hours for most customers. But in practice, if your financials are stable, the limit might not move much. The real value is when you have a big event — a funding round, a large customer payment, or a significant cost reduction. Luna: So it's more about responsiveness to life events than constant fluctuation. Lucas: Exactly. And that's a huge shift from traditional cards, where you have to call the issuer and request a limit increase, wait for a human to review your financials. With Brex, it's automated and immediate. Luna: But there's a downside, right? If your cash balance drops, your limit could shrink without warning. That could disrupt operations. Lucas: It's a trade-off. Brex does notify you when the limit changes, but it's still a risk. Some finance teams prefer the predictability of a static limit. But for startups that are cash-conscious, the dynamic limit can actually enforce discipline — you can't overspend beyond what you have. Luna: Let's talk about the competitors. Stripe issued corporate cards too, and Ramp, and Divvy. Are they all doing the same thing? Lucas: Similar, but with different twists. Ramp, for example, uses real-time expense data but also offers software to automatically categorize spending and enforce policies. Divvy focuses on budget controls — you can set limits per employee or per department. Brex was the pioneer in dynamic underwriting, but now everyone is moving toward real-time data. Luna: So the trend is clear: static credit limits are dying. The question is, how far can this go? Could we see personal credit cards using real-time data? Lucas: Maybe, but there are privacy concerns. For businesses, it's more acceptable because companies want tight controls. For individuals, sharing bank account access for a credit card is a tough sell. But there are early experiments — like Tally, which uses bank data to manage credit card payments. Luna: That's a different use case, though. Tally helps you pay off cards, not set limits. Lucas: Right. For corporate cards, the real-time data model makes sense because the card issuer and the business have aligned interests — both want to avoid overspending. I think we'll see more fintechs adopt this approach, especially for small businesses that don't have long credit histories. Luna: So what's the next frontier? Maybe integrating with project management tools to set limits based on project budgets? Lucas: That's a great thought. Some companies are already doing that — linking card limits to specific initiatives. Imagine a card that only allows spending on a particular marketing campaign, and the limit is the campaign budget. That level of granularity could be powerful. Luna: We'll have to keep an eye on that. For now, the key takeaway: fintechs like Brex are using real-time financial data to make credit limits smarter and more responsive, which benefits startups that need flexibility. Lucas: And it's a great example of how AI and API connectivity are reshaping a traditional banking product. The corporate card is no longer just a piece of plastic — it's a dynamic financial tool.