Bad debt comps
for multifamily.

What’s inside

Investor-grade analytics, sourced and citable. Neutral data — you form the conclusion.

Harris County, TX · Nov 2025 – Jun 2026

Three reads on bad debt: the asset, the market, the operator.

1 · Asset level

2 · Market

Estimated bad debt

0.9% – 7.9%

Estimated bad debt as percent of GPR, across 6 verified multifamily operators spanning Class A, B, and C portfolios.

Bad debt confirmed

$37.2M

Sum of court-awarded judgment dollars on plaintiff wins — the unpaid rent operators have already obtained judgments for.

Filings tracked

39,874

Eviction filings ingested across all Harris County JoP courts in the window.

Operators verified

14

Parent operators whose property or brand attribution has been cross-checked against published portfolios.

3 · Operator

Ranked by court-confirmed bad debt — names unlock for members.

Bad debt % = estimated rent loss ÷ Gross Potential Rent (units × rent × 6-month window). Rent loss is estimated per case from its court disposition — displacements carry court-awarded past-due rent plus a turnover allowance; cured cases carry a partial loss. Unit counts from HCAD assessor parcels where matched; affordable (LIHTC/subsidized) properties excluded. See Methodology below for confidence bands. Source: Harris County Justice of the Peace Courts public records. Operators de-identified by class and scale — identities unlock for subscribers.

Coverage

Houston live. Sunbelt rolling on, quarter by quarter.

• Live

  Houston

  Harris County, TX

• Next

  Tampa

  Hillsborough County, FL

• In licensing

  Phoenix

  Maricopa County, AZ

• Planned

  Atlanta · Charlotte · Jacksonville · Indianapolis · Cincinnati

  Sunbelt expansion

New metros require a verified public records pipeline plus operator-attribution verification before the data goes into the product. We add roughly one metro per quarter once the playbook is established.

Methodology

How we compute bad debt as percent of GPR.

Built for

The investor evaluating the property — not the landlord evaluating the tenant.

• —Multifamily syndicators
• —Asset managers
• —Brokers covering multifamily transactions
• —Lenders & credit analysts
• —Fund analysts and LP diligence teams

What’s in each tier

Four tiers. Aggregated outputs only.

Exports and API responses return aggregated analytics only — operator summaries, submarket trend rollups, property-level breakdowns. Never raw case records, never tenant-identifiable information.

FAQ

What underwriters and analysts ask first.

How do you calculate bad debt as a percent of GPR?

Per case, weighted by court disposition. Displacements (judgment for the landlord, default, agreed) carry the court-awarded past-due rent — real recorded dollars — plus a turnover allowance for the vacancy. Cured cases (the tenant paid and stayed) carry only a small partial loss. Summed across an operator’s cases, that’s the estimated rent loss; divided by 6-month Gross Potential Rent (units × median rent × 6), that’s the bad-debt percentage. The past-due dollars are court data; only the turnover and partial-loss factors are assumptions, and they apply to subsets.

Why disposition-weighted instead of counting filings?

Because not every eviction filing is a write-off. Some operators file the day rent is late as a routine collection tactic — the tenant pays within days and the case is dropped. Charging every filing the same flat loss would massively overstate those aggressive early-filers. Weighting by how each case actually resolved separates lawsuits-filed from dollars-lost, which is the number underwriters care about.

Why ranges instead of single numbers?

The rent-loss numerator is largely court data. The remaining uncertainty is the unit-count denominator, which we’ve sourced per complex (Yardi Matrix, operator sites, HAR) but is not yet authoritative. Each operator carries a confidence band (±20% high, ±35% medium, ±50% low) that produces the range. Our next workstream ingests HCAD assessor records — per-parcel unit counts — which collapses the ranges toward point estimates.

Where does the data come from?

Public county court records. Houston pulls from the Harris County Justice of the Peace Courts public extract service; additional markets (Tampa, Phoenix, etc.) layer in via each county’s public records pipeline. Every metric on this page and inside the product carries a source citation back to the originating county.

Is this a tenant screening service?

No. BadDebt is investor-side analytics. We do not provide tenant background checks, consumer reports, or any FCRA-regulated decisioning. Subscribers see aggregated property-, operator-, and submarket-level views — never tenant-identifiable information.

How are operators attributed to properties?

Court filings name shell LLCs (e.g. RISE CAMELBACK LLC), not the parent. We extract the brand prefix, then verify it against the operator’s own publicly-disclosed portfolio (AUM page, investor materials, SEC filings) before associating filings with a named operator. Unverified brand clusters stay anonymous in the product.

Why only aggregated data — why no raw case exports?

Two reasons. First, court data licensing terms in many counties prohibit re-distribution of party-identifiable records. Second, aggregated outputs are what underwriters actually need to price a deal — operator concentration, judgment outcomes, MoM trends. Raw records are noise to that workflow.

How fresh is the data?

Houston refreshes weekly. Other markets vary by source — monthly bulk drops for Hillsborough and Maricopa, near-real-time for the Phoenix City supplement. Each metric carries the as-of window.

When does the product launch?

We’re in pre-launch with the Houston dataset live and one verified-operator ranking already shipped. The first paid tier opens to waitlist members. Pricing on this page is the at-launch pricing.