Permit Data for Lenders: The HELOC Blind Spot Hiding in Public Records

By Raf Howery | Kukun

Every building permit is a homeowner raising their hand and saying, “I’m about to spend real money on this house.” Most lenders never see it. Not because the data isn’t public (it almost always is) but because of a gap nobody built a fix for: the distance between what a permit says and what a HELOC underwriting model can actually use.

We ran the numbers on a recent 550,000-permit, multi-state file. 83% of those permits got a construction label of some kind. Only 57% mapped to a category any home-equity lending model would recognize as HELOC-eligible. The 26-point gap in between isn’t bad data; it’s electrical work, plumbing, HVAC, siding, drywall. Real permits, real dollars, filed with a city, sitting in public records, and functionally invisible to the systems that are supposed to be watching for home-equity opportunity.

If you work in home equity lending, mortgage servicing, or property risk, that gap is the whole story. Here’s what’s actually in it, and what to do about it.

What Building Permits Actually Capture

A permit record on its own is thin: an address, a status, a dollar figure, a free-text description written by whoever filed it. To make that usable at scale, it needs a taxonomy (a consistent way of saying “this permit is a kitchen remodel” or “this one is a roof replacement”) applied across millions of records from thousands of different jurisdictions, each with its own format and its own conventions.

That’s harder than it sounds. Two label systems are often needed side by side: an atomic taxonomy (kitchen, bathroom, roof, HVAC, and so on) and a second layer that maps those atomic labels to whatever schedule a specific underwriting or credit model actually uses. The two rarely line up one-to-one, which is exactly where the gap opens up.

The HELOC Eligibility Gap

Here’s the uncomfortable part: not every dollar a homeowner spends on their house is a dollar a lending model can see. Kitchen and bathroom remodels, additions, ADUs, decks, roofs, solar, pools. These map cleanly to a HELOC-relevant schedule. But electrical, plumbing, HVAC, landscaping, demolition, fencing, flooring, drywall, and siding (categories that represent a huge share of real permit volume) typically don’t map to anything at all.

That’s not a data quality failure. It’s a category mismatch between how homeowners actually spend (a full renovation touches five trades at once) and how lending schedules were originally drawn up (one line item at a time). Any institution relying on off-the-shelf permit labels without accounting for this gap is systematically undercounting home-equity intent, quietly, every single month.

Your County Knows Your Dirt Better Than Your House

A related asymmetry shows up in property assessment data that feeds a lot of alternative data for home equity lending: land value fields fill in around 90% of the time. Building value fields fill in less than 17% of the time. Assessors are consistently confident about what a parcel of land is worth and consistently unsure what’s actually built on it.

The practical implication: a lot of “property value” data pulled from assessor records is doing more land appraisal than structure appraisal, even though it gets used as a stand-in for total property value. For any model weighting assessed value as a HELOC input, that’s worth knowing before, not after, it shapes a decision.

Permits Don’t Reset at Closing

Most financial data resets with the person: a new owner means a new file, a fresh credit history, a blank slate. Permit history doesn’t work that way. It’s tied to the address, not the applicant, so a renovation from three owners ago still shows up on that property’s record today.

That makes building permits one of the only inputs in the home-equity stack that survives every closing. A property’s full seven-year renovation story (not just what the current owner has done, but everything the house has been through) is available the moment you pull the address, no reset required. For anyone building a leading indicator of home-equity behavior rather than a lagging one, that persistence is the whole advantage: permits get filed months before a listing, an appraisal, or a credit inquiry would ever surface the same signal.

What Lenders Should Do With This

Three practical takeaways for anyone evaluating permit data for lenders or building it into a HELOC underwriting pipeline:

First, ask any data provider what percentage of their permits actually map to your eligibility schedule, not what percentage get a label, but what percentage map to your categories. The gap between those two numbers is real money left unmodeled.

Second, treat assessor-sourced building value with the same skepticism you’d apply to any thin field; know the fill rate before you weight it.

Third, weight permit-level signals higher in models designed to predict intent rather than just verify it after the fact. The whole point of a leading indicator is using it before the homeowner tells you themselves.

→ Explore permit data for lenders at mykukun.com/developers

Full Knowledge Graph and standalone Permits API available · mykukun.com/blog/business

Billy Guteng

Billy is a content writer for Kukun. He was born in France, and has lived in multiple cities around the world ever since. He loves art, design and architecture as well as construction and real estate.

Published September 30, 2026