Latest / Fintech Conversations with Fexingo: Banking, Payments, and Financial Software Companies / How Fintechs Use Real-Time Housing Data for Dynamic Rent-to-Own Underwriting
Transcript
- Lucas: So there's this company called Divvy Homes — they've been around since about 2017 — and they do rent to own, but with a fintech twist. They use real-time rent payment data to dynamically adjust the terms of the contract. Luna: Right, so the idea is: instead of a traditional mortgage, you rent a home and a portion of your rent goes toward a down payment? Lucas: Exactly. But the key innovation is that the underwriting isn't static. It updates every month based on how you actually pay. Divvy pulls data from the property management system — every time you pay rent, it hits their risk engine. Luna: So if you pay consistently for six months, what happens? Do you get a better price? Lucas: You get a lower purchase price multiplier. They set the home's future purchase price at the start — say, 1% of the current value each year. But if you have a perfect payment record after 12 months, they might reduce that multiplier by ten basis points. Not huge on its own, but over a 30-year path to ownership, that's meaningful. Luna: And they're using real-time data from the rent payments themselves? Or are they pulling in other sources? Lucas: They layer it with property tax records and county deed data to verify ownership. But the core is the rent payment stream. They partner with credit bureaus like TransUnion to get alternative data — on-time rent history from platforms like Cozy or Avail. That data flows in monthly, so the model updates. Luna: It's like a dynamic credit score, but tied specifically to housing. Makes sense — rent is the biggest expense for most people, and it's traditionally invisible to lenders. Lucas: That's the whole thesis. The CFPB actually issued guidance in 2024 saying that using alternative data like rent payments could expand credit access, as long as the models are fair and explainable. Divvy's model is a case study in that. Luna: But there are risks. What if someone pays late one month? Does their multiplier go up? Lucas: It can. The contract usually has a grace period, but if the payment doesn't come in within 15 days, the risk engine flags it. The purchase price multiplier might increase by five or ten basis points. And if they miss two consecutive months, they could lose the option to buy altogether. Luna: So it's essentially a real-time repricing of the option. That's pretty aggressive for the consumer. Lucas: It is. But the proponents argue it's more transparent than a traditional lender who just denies you after a late payment. You see the terms change in your app. And if you get back on track, the multiplier can decrease again. Luna: That's actually a fair point. It's dynamic, but bidirectional. Lucas: Right. And that's the fintech promise — using data to make credit more responsive. But it also raises questions about algorithmic fairness. If the model penalizes someone who has an irregular income but always catches up, is that fair? Luna: Yeah, the CFPB would want to see that the model doesn't disproportionately impact protected groups. Divvy's model uses payment history, not demographics, but the correlation could still exist. Lucas: Exactly. And that's the tension. On one hand, more data should mean better risk assessment. On the other, if the data is noisy or biased, you could end up with a system that's less fair than a traditional mortgage. Luna: But what about the scale? Divvy is still relatively small — they've done maybe a few thousand homes. How replicable is this model? Lucas: The data infrastructure is getting cheaper. Companies like Plaid and Yodlee already connect to thousands of property management systems. So any fintech can potentially tap into rent data. I think we'll see more players — maybe even traditional banks — offering dynamic rent to own products. Luna: That would be a big shift. Banks have been slow to adopt alternative data for mortgage underwriting. Lucas: For sure. But if this proves out — if default rates are lower and homeownership rates go up — regulators might push for it. There's a lot of interest in using housing data to close the wealth gap. Luna: Honestly, if this episode was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Keeps the show ad-free and lets us dig into niche topics like this. Lucas: Yeah, seriously. We love diving into the mechanics, and listener support is what makes that possible. So if you got something out of today's deep dive, that link is buy me a coffee dot com slash fexingo. Luna: Alright, back to the data. Lucas, you mentioned Plaid — how do they integrate with property management systems specifically? Lucas: Plaid has a product called Plaid Exchange that connects to thousands of financial accounts, including property management software like AppFolio and Yardi. So a fintech can request a user's rent payment history directly from the source, with the user's permission. That data comes in as a feed of transactions — monthly rent, dates, amounts. Luna: So the data is structured. That makes it easier to build a model than scraping PDF statements. Lucas: Exactly. And it's real-time. So when a payment clears, the model can update within hours. That's the technical backbone of dynamic underwriting. Luna: What about property tax data? How does that feed in? Lucas: Public records — county assessor offices publish tax rolls. Companies like CoreLogic aggregate that and sell it. Divvy uses it to verify that the property's tax assessment matches the expected value. If taxes spike, the model can adjust the rent to own terms to account for higher carrying costs. Luna: So the data isn't just about the tenant — it's also about the property itself. Lucas: Right. The risk engine looks at both sides. If the property's tax bill goes up 20%, that's a real cost that affects the viability of the contract. The tenant might see their monthly rent increase, or the purchase price might adjust. Luna: That's almost like a dynamic lease with an embedded option. It's more complex than a standard mortgage. Lucas: It is. And that complexity is why fintechs need to be transparent. If the terms change every month, the consumer needs to understand why. Divvy sends a monthly statement that breaks down the data inputs and the resulting multiplier. Luna: Transparency is key. But I wonder — how do they handle disputes? If a tenant says the data is wrong, who resolves it? Lucas: That's a growing area. Divvy has a dispute team, and they rely on the property management system's records as the source of truth. If the tenant claims they paid on time but the system shows late, it's hard to prove. The CFPB has said that fintechs must have a reasonable dispute process, but there's no standard yet. Luna: So the data quality is critical. If the property management system has a glitch, the tenant could end up with a worse deal. Lucas: Exactly. And that's a systemic risk. As more lenders use alternative data, the accuracy of that data becomes a consumer protection issue. The FTC has already gone after some data brokers for providing inaccurate information. Luna: Let's talk about the future. Do you see this model expanding beyond single-family homes? Maybe into multifamily or commercial? Lucas: Potentially. There are startups working on rent to own for commercial leases, but it's more complex because commercial leases have variable terms like CAM charges. For residential, I think we'll see more integration with smart home data — like utility payments, even security system usage — to build a more complete picture of the tenant's reliability. Luna: So your utility payment history could affect your ability to buy a home? That's a big expansion of data usage. Lucas: It is. And it's controversial. But the argument is that it helps people who have thin credit files build a positive history. The key is whether the data is predictive of mortgage default. Early studies suggest that rent payment history is actually more predictive than some credit score components. Luna: That's a powerful finding. If that holds, regulators might embrace it. Lucas: Yeah. I think the next five years will see a lot of experimentation. The infrastructure is there, the data is flowing, and the consumer demand for alternative paths to homeownership is high. It's one of the most interesting corners of fintech right now. Luna: Definitely. Thanks for walking through the mechanics, Lucas. This one gave me a lot to think about. Lucas: Happy to. It's a space worth watching.