Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Fintechs Use Real-Time Bank Account Data for Instant Mortgage Pre-Approval
Transcript
- Lucas: So Luna, you know how getting a mortgage pre-approval has traditionally meant gathering pay stubs, W-2s, bank statements, tax returns — basically a paper chase that takes days, sometimes weeks? Luna: Oh, I know it firsthand. When I bought my place two years ago, I spent a weekend scanning documents. It was not fun. Lucas: Right. Well, a growing number of fintech lenders are now sidestepping that whole process by pulling real-time bank account data directly through open banking APIs. The result is an instant mortgage pre-approval — sometimes in under ten minutes. Luna: Under ten minutes? That is a massive leap from the old multi-day norm. Give me a concrete example. Lucas: Let's start with Better Mortgage. They've been a pioneer in digital mortgage lending for years, but in early 2025 they rolled out a product called 'Instant Pre-Approval' that uses Plaid to connect to a borrower's bank accounts. Once you give consent, it pulls 12 to 24 months of transaction history, analyzes income deposits, recurring expenses, and cash flow patterns — and issues a pre-approval letter within minutes. Luna: So they're not just checking your credit score and calling it a day. They're actually looking at your financial behavior. Lucas: Exactly. The old model was heavily reliant on FICO scores and self-reported income, which can be inaccurate or outdated. Real-time data lets them verify that your income is actually coming in consistently, that you're not over-leveraged on other debts, and that your spending habits are sustainable. It's a more dynamic picture. Luna: And this is all possible because of open banking regulations, right? The data sharing framework? Lucas: Partly. In the U.S., the Consumer Financial Protection Bureau finalized its open banking rule in late 2024, which requires banks to make consumer data available through standardized APIs — with the consumer's permission. That gave fintechs a legal and technical foundation to build on. Plaid and Yodlee have been aggregating that data for years, but the regulatory clarity accelerated adoption. Luna: So the consumer grants access, the fintech pulls the data, and the underwriting engine processes it in real time. What does that engine actually look at? Lucas: Good question. Most of these systems use a combination of rules-based logic and machine learning models. They look at things like: average monthly deposit amount, volatility of deposits, frequency of overdrafts, debt to income ratio calculated from actual transaction data, and even categorizations like how much you spend on dining or travel. Some models also look at 'income stability' — how consistent your paychecks are if you're a gig worker or freelancer. Luna: That's especially important for the growing number of self-employed borrowers. Traditional lenders often struggle to verify their income because tax returns don't always reflect current earnings. Lucas: Right. A freelancer might have a great year but a low tax bill after deductions, and a traditional underwriter sees that and says 'you don't qualify.' With real-time data, the fintech can see that your bank account has been receiving consistent payments of $8,000 a month for the past six months. That's a much truer picture. Luna: Are there any downsides or risks here? I'm thinking about data privacy and potential bias. Lucas: Absolutely. The biggest concern is that consumers might not fully understand what they're consenting to. When you link your bank account through Plaid, you're giving the lender access to a lot of granular data — every transaction, every subscription, every coffee purchase. There's also the risk of algorithmic bias if the models are trained on historical data that reflects discriminatory lending patterns. Luna: And what about security? If a fintech's database gets breached, that's a treasure trove of financial data. Lucas: Yeah, that's a real threat. Better Mortgage and others use tokenization and encryption, but the aggregation layer — Plaid, for example — becomes a high-value target. The CFPB's rule does require data minimization, meaning lenders should only request the data they actually need, but enforcement is still evolving. Luna: So the regulatory framework is there, but it's still being tested. How are traditional banks responding? They can't be thrilled about fintechs eating their lunch on mortgages. Lucas: They're not. But interestingly, some are partnering. For example, JPMorgan Chase has been piloting a similar instant pre-approval feature using its own internal data — since they already have all your checking and savings account history — plus some external aggregation. They're using it as a retention tool for existing customers. Other banks are building their own platforms or acquiring fintechs. Luna: So the endgame might be that every mortgage lender, whether bank or fintech, uses real-time data as the standard, and the old paper-based process becomes a relic. Lucas: I think that's where we're heading. And it's not just pre-approval. Some lenders are now using ongoing transaction monitoring to adjust rates or offer refinancing at the right moment. If the system sees your credit score improve or your debt load drop, it could trigger a lower rate offer. Luna: That's proactive lending. It shifts the dynamic from 'borrower applies when they need money' to 'lender offers when the borrower's financial health improves.' Lucas: Exactly. And it's more efficient for the lender too — lower origination costs, faster processing, fewer manual underwriting errors. The trade-off is that they're bearing more technology risk and regulatory scrutiny. But the early numbers suggest the conversion rates are higher and the default rates are comparable, if not better, because the data is more accurate. Luna: You mentioned Better Mortgage. Are there other fintechs doing this well? Lucas: Rocket Mortgage has been using its own version called 'Rocket Auto-Approval' which also leverages connected bank accounts. And there's a smaller player called Tomo — they focus on first-time homebuyers and use real-time data to pre-approve within minutes, often with lower down payment requirements. They're betting that the data will show a borrower is less risky than their credit score suggests. Luna: That's a bold bet, but it makes sense for younger borrowers who might have thin credit files but solid income and good savings habits. Lucas: Right. A recent college grad with a good job but only two years of credit history — their FICO score might be mediocre, but their bank account shows steady deposits and responsible spending. Real-time data can capture that nuance. Luna: Okay, so we've talked about the 'what' and the 'how.' What's the biggest obstacle to widespread adoption? Lucas: I'd say consumer trust and education. Many people are still wary of linking their bank accounts to third parties. The data breaches at companies like Equifax and others have made people cautious. Lenders need to clearly explain what data they're accessing, how it's used, and how it's protected. The ones that do that well will win. Luna: And we should note that this kind of conversation — about how technology is reshaping a fundamental financial product — is exactly the reason we do this show. Lucas: Absolutely. And one thing I appreciate is that we can have this conversation without any ads or sponsors pushing a particular product. It's just us digging into what's actually happening. Luna: Right. We deliberately keep the show ad-free because we think the content stands on its own. If you find value in episodes like this and want to support that choice, you can find us at buy me a coffee dot com slash fexingo. Lucas: That's it. Just a simple way to help us keep doing what we're doing. No pressure, no perks — just knowing that listeners find it useful enough to chip in. Luna: And now back to the mortgage data revolution. So Lucas, where do you see this heading in the next two or three years? Lucas: I think we'll see a bifurcation. The big banks will integrate real-time data into their existing mortgage platforms, offering it as a feature for existing customers. The fintechs will push further into the actual closing process — using the same data to streamline document collection, appraisal ordering, and even title search. The ultimate goal is a fully digital mortgage that can be closed in days, not weeks. Luna: And what about the secondary market? Fannie Mae and Freddie Mac — they set the standards for what loans they'll buy. Are they adapting? Lucas: They are. Fannie Mae has been piloting its 'Day 1 Certainty' program which uses automated asset and income verification through third-party data providers. And Freddie Mac has something similar. They're starting to accept verified bank account data instead of paper documents for certain loan types. That's a big deal because if the GSEs accept it, lenders can scale it confidently. Luna: So the infrastructure is coming together — regulation, data aggregation, lender technology, secondary market acceptance. It feels like a tipping point. Lucas: It does. And the next frontier might be using that same real-time data for continuous underwriting — not just at origination but throughout the life of the loan. Imagine a mortgage that adjusts its rate based on your financial health, or one that automatically offers a forbearance if the system detects a sudden income disruption. Luna: That's both powerful and a little scary. But if done transparently, it could make homeownership more accessible and less stressful. Lucas: Exactly. And that's the promise of fintech — not just speed, but smarter, more personalized financial products.