How Coburbia estimates
Every number on this site is a calculation from inputs you can see. This page lists the inputs, the sources, the weights, and the rules — including the ones that stop us from publishing.
Sources
- HUD Small Area Fair Market Rents by zip and bedroom count, the anchor for long-term rent. FMR is a 40th-percentile figure, so we scale it by ×1.08 toward an asking rent.
- Census ACS 5-year median gross rent, median home value, rental vacancy, and median real-estate taxes per zip (ZCTA). The tax rate on every report is taxes ÷ value for the zip.
- Zillow Research ZORI (rent index) and ZHVI (value index) by zip, used as blend inputs; we never redisplay the series.
- Inside Airbnb market-aggregate nightly rates and modeled occupancy for the cities it publishes, never listing-level data.
- Municipal open data: DC and Boston parcels and recorded sales for comparables; DC short-term-rental license counts by zip.
- Operating actuals contributed through the statement parser under explicit consent, aggregated per zip and mode.
Long-term rent
Weighted blend of the sources present for the zip — FMR 0.5, ACS 0.2, ZORI 0.3, Redfin 0.2 — normalized over what exists. All-unit medians are scaled to the bedroom count (2BR = 1.0; 0BR 0.72, 1BR 0.84, 3BR 1.22, 4BR 1.42).
Hedonic adjustments: size elasticity 0.35 vs the zip median (capped ±15%), 4% per bath vs typical (capped ±8%), property type (single-family +3%, condo −3%, small multifamily −2%), and age bands (2015+ +6%, 2000+ +3%, pre-1970 −3%). Total adjustment is capped at ±35%.
Mid-term and short-term
Mid-term = long-term rent × furnished premium (1.5 default) at 85% utilization. Short-term = market nightly rate × bedroom factor × occupancy × 365, plus collected fees. When a zip has no nightly-rate source we derive one as long-term rent ÷ 365 × 2.2 at 55% occupancy, label it as derived, and widen the band to ±40%.
Tiers and the publication floor
Verified Own actuals for that mode in that zip (or its pooled submarket): at least 5 distinct units and 5 unit-months within 18 months. The prior is blended toward the actual median with weight n ÷ (n + 12), n = unit-months. Band ±10%.
Calibrated A public prior exists for the zip and actuals exist for the same mode in the same metro; the metro correction factor (clamped 0.6–1.6) is applied. Band ±15%.
Modeled Everything else. Band ±25%.
Nothing derived from fewer than 5 distinct units is ever published, at any level. Contributed data is never shown per unit.
Returns and criteria
- NOI = gross revenue − operating expenses (before debt service)
- Cap rate = NOI ÷ purchase price
- Cash-on-cash = (NOI − debt service) ÷ (down payment + closing + rehab + furnishing)
- DSCR = NOI ÷ annual debt service
- Default financing: 25% down, 6.50% for 30 years, 3% closing — all editable
- Default buy box: cap ≥ 6%, cash-on-cash ≥ 5%, DSCR ≥ 1.20×, price ≤ $750,000 — yours to change
- All four met → Meets your criteria; two or three → Partially meets; one or none → Below your criteria
Coburbia reports which mode produced the highest modeled return under your assumptions. It does not say whether to buy. The regulatory section is a checklist of questions for the jurisdiction, never a determination of what is permitted.
Coburbia produces financial models from user-supplied and estimated inputs. It does not provide investment, legal, or tax advice. Verify all figures and local regulations independently before transacting.