Data Disclosure
← Back to Analytics
RE/TERM · Milairo

Data Sources, Methodology
& Limitations

RE/TERM provides market intelligence derived from public government datasets, Zillow Research public data, and Federal Reserve economic series. This page discloses every data source used, how values are calculated, and what this tool cannot tell you.

Last updated: March 15, 2026 · Version 1.1
On This Page
What RE/TERM Is and Is Not Data Sources Calculations & Methodology Known Limitations No MLS or Transaction Data Not Financial, Legal, or Tax Advice Contact Geographic Boundary Data
01 · What RE/TERM Is and Is Not
RE/TERM is a free market intelligence tool. It is not a licensed real estate broker, investment advisor, financial planner, or automated valuation service.
What RE/TERM is
What RE/TERM is not
02 · Data Sources
Zillow Research — Public Data

RE/TERM uses publicly available CSV files published by Zillow Research at files.zillowstatic.com. These files require no API key or data agreement. They are fetched nightly at 3am UTC via a scheduled serverless function and cached in Netlify Blob Storage. Users receive cached data — not a live query — which is refreshed once daily.

SeriesWhat It MeasuresFrequency
ZHVIZillow Home Value Index — smoothed, seasonally-adjusted median for homes at the 35th–65th percentile. Includes all homes in the area, not just those recently listed or sold.
Metro and ZIP level.
Monthly
ZORIZillow Observed Rent Index — repeat-rent methodology tracking the same units over time, all bedroom types.
Metro level.
Monthly
ZHVFZillow's 12-month home value forecast. This is Zillow's proprietary model output — RE/TERM does not independently verify or validate this forecast.
Metro level.
Monthly
ZHVI is not a listing price or sale price. It is a statistical index of all homes in a geographic area — including long-held properties that have not recently transacted. Active listings on Zillow, Redfin, or the MLS will typically show prices 5–40% above ZHVI depending on market conditions. RE/TERM estimates a listing price range from ZHVI using a market-velocity-adjusted formula. See Methodology below.
FRED — Federal Reserve Economic Data

The Terminal tool uses live data from the St. Louis Federal Reserve's FRED API. All series are publicly available. Data is fetched at session start.

Series IDDescriptionSource
MORTGAGE30US30-year fixed mortgage rate (Freddie Mac PMMS)Freddie Mac / FRED
MORTGAGE15US15-year fixed mortgage rateFreddie Mac / FRED
DGS1010-year Treasury constant maturity yieldU.S. Treasury / FRED
CSUSHPINSAS&P/Case-Shiller U.S. National Home Price IndexS&P Dow Jones / FRED
HOUSTHousing starts (new privately-owned units, thousands)U.S. Census / FRED
PERMITBuilding permits authorized (thousands)U.S. Census / FRED
MSACSRMonthly supply of new homes (months)U.S. Census / HUD / FRED
HSN1FNew single-family homes sold (annual rate)U.S. Census / FRED
FIXHAIHousing Affordability Index (NAR methodology)NAR / FRED
MSPUSMedian sales price of houses sold in the United StatesU.S. Census / HUD / FRED
EXHOSLUSM495SExisting home sales (seasonally adjusted annual rate)NAR / FRED
RVUUSRental vacancy rate, U.S.U.S. Census / FRED
CUUR0000SEHACPI: Rent of Primary Residence (used for rent adjustment)BLS / FRED
U.S. Census Bureau — American Community Survey

ZIP-level demographic and housing data is sourced from the ACS 5-Year Estimates, 2023 vintage (covering survey years 2019–2023). This dataset is accessed via the public Census API at api.census.gov. No authentication key is required.

Variables used: median home value (B25077), median household income (B19013), median contract rent (B25058), population (B01003), owner/renter tenure (B25003), gross rent (B25064), rent burden (B25071).

ACS 5-year data has a significant temporal lag. The 2023 vintage reflects surveys conducted from 2019 to 2023, with a statistical midpoint of approximately mid-2021. In markets that have appreciated significantly since 2021, the Census home value figure will be substantially below current conditions. The Terminal applies a Case-Shiller HPI forward-adjustment to correct for this. Analytics shows the Census vintage label on all ZIP data.
HUD Fair Market Rents

Fair Market Rent data is sourced from the U.S. Department of Housing and Urban Development via the HUD User API. RE/TERM uses FY2026 FMRs (effective October 1, 2025). FMRs are published at the metropolitan area or county level — a single FMR applies to all ZIP codes within a county.

HUD FMR represents the 40th percentile of gross rent for unassisted units. It is not the same as median market asking rent. Well-maintained units in desirable sub-markets will typically rent above FMR.

If the HUD API is unavailable, RE/TERM estimates FMR from Census gross rent using bedroom-scaling ratios. Estimated values are clearly labeled.

IRS Statistics of Income — Migration

Metro-level migration patterns are sourced from IRS SOI Tax Stats, which tracks year-over-year address changes on tax returns. The most recently available data is FY2022. This data is embedded directly in the tool and is not fetched live. It may not reflect migration pattern changes since 2022.

03 · Calculations & Methodology
Listing Price Range Estimate

RE/TERM estimates an active listing price range from ZHVI using a days-adjusted premium band. Active listings trade above the ZHVI index because they represent currently available, select inventory rather than the full stock of all homes. The premium varies with market velocity.

Days to PendingPremium BandMarket Condition
< 15 days+25% to +40% above ZHVIExtreme sellers market
15–30 days+18% to +30% above ZHVISellers market
30–50 days+10% to +22% above ZHVIBalanced-to-seller
50–70 days+5% to +15% above ZHVIBalanced-to-buyer
> 70 days−2% to +8% above ZHVIBuyers market

The days-to-pending figure used is a metro-level average. Individual ZIPs within a metro may have faster or slower absorption rates. This estimate is not sourced from MLS data and does not represent actual active listing prices.

HPI Forward-Adjustment (Terminal)

The Terminal adjusts Census home values forward using the S&P/Case-Shiller HPI (CSUSHPINSA) or a metro-level HPI when available from FRED. The adjustment factor is the ratio of the current HPI to the HPI at the approximate Census survey midpoint (mid-2021).

Formula
Adjusted Value = Census Median Value × (HPI_current / HPI_2021_baseline)
Adjusted Rent = Census Gross Rent × (CPI_SEHA_current / CPI_SEHA_2021_baseline)

This adjustment is a statistical approximation. It applies a national or metro-wide index to a specific ZIP — local sub-market conditions may differ materially. The adjustment factor and data source are displayed as a tooltip in the Terminal.

Gross Rent Yield
Formula
Gross Rent Yield = (ZORI × 12) / ZHVI × 100

This is a gross yield only. It does not account for vacancy, CapEx reserves, property management, taxes, insurance, or financing costs. Use the Deal Analyzer for net cash flow analysis.

Housing DTI
Formula
Monthly PITI = Monthly P&I × 1.25 (P&I × 1.25 approximates taxes and insurance escrow)
Housing DTI = Monthly PITI / (Median Household Income / 12) × 100

Assumes: 20% down payment, 30-year term, current FRED MORTGAGE30US rate
MBS Spread (Terminal)
Formula
MBS Spread = MORTGAGE30US − DGS10
Historical norm: 150–200 basis points

This is a derived signal, not a published series. It represents the risk premium above the 10-year risk-free rate that the mortgage market is demanding.

Deal Analyzer — All Metrics

The Deal Analyzer uses standard real estate investment finance formulas. All inputs are provided by the user. RE/TERM does not populate property tax, insurance, or other expense fields from any data source — these must be entered manually from county records and quotes.

Key Formulas
NOI = Gross Income − Operating Expenses (before debt service)
Cash-on-Cash Return = Annual Cash Flow / Total Cash Invested × 100
DSCR = NOI / Annual Debt Service
Cap Rate = NOI / Purchase Price × 100
GRM = Purchase Price / Annual Gross Rent
04 · Known Limitations
LimitationImpact
Census data lag (2–3 years) Home values and income from Census ACS reflect 2019–2023 survey data. In appreciating markets, Census home values may be significantly below current conditions. HPI adjustment applied in Terminal mitigates but does not eliminate this gap.
ZHVI is not a listing price ZHVI includes all homes, not just listed or recently sold properties. Active listings will show higher prices. Listing range estimate is derived, not from MLS data.
Days-to-pending is metro-level Listing premium bands use metro-level days-to-pending. Individual ZIP codes within a metro may have meaningfully faster or slower absorption. Sub-market conditions are not captured.
HUD FMR is county-level A single FMR applies to all ZIP codes within a county. High-demand sub-markets within a county will typically see asking rents above FMR.
ZHVF is Zillow's model The 12-month home value forecast is produced by Zillow's proprietary algorithm. RE/TERM does not modify or independently validate this forecast. Past forecast accuracy varies by market.
Migration data lag IRS SOI migration data reflects FY2022. Migration patterns may have shifted in 2023–2026 due to remote work normalization, rate-driven affordability changes, or other factors.
No property-level data All data is area-level (metro, county, or ZIP). Individual property values, conditions, histories, or assessments are not available.
Embedded city data is manually maintained Days-to-pending, inventory months, and property-type breakdown figures for the 16 tracked metros (Miami, Austin, Atlanta, Phoenix, Charlotte, Dallas, Denver, Nashville, Seattle, New York, Los Angeles, Chicago, Boston, Houston, Las Vegas, Tampa) are updated manually on a monthly basis. Values may lag recent market shifts by up to 30 days.
05 · No MLS or Transaction Data
RE/TERM does not have access to Multiple Listing Service (MLS) data, Zillow's internal listing API, Redfin transaction data, CoreLogic, Attom, or any proprietary property database. This is important to understand before using any listing price estimates from this tool.

All listing price ranges shown on RE/TERM are estimates derived from the ZHVI index using the days-adjusted premium methodology described above. They are not sourced from active listings, recently sold comparables, or any MLS system.

For actual current listing prices and recently sold data, use:

06 · Not Financial, Legal, or Tax Advice

RE/TERM is built and maintained by Victor De la Cruz, CPA, through Milairo. While the tool incorporates CPA-grade thinking around real estate tax concepts (§280A, Schedule E, DSCR, depreciation), nothing on RE/TERM constitutes professional financial advice, investment advice, tax advice, or legal advice.

The Deal Analyzer produces mathematical outputs based on user-provided inputs. These outputs are illustrative. Actual investment performance will differ based on market conditions, financing terms, tenant behavior, capital expenditure timing, local regulations, tax law changes, and many other factors not modeled by this tool.

Before making any real estate investment decision, consult:

If you would like to work with a CPA who specializes in real estate investment analysis — including STR §280A classification, cost segregation, 1031 exchange planning, and entity structure — visit milairo.com to learn about working with Victor De la Cruz, CPA.
Geographic Boundaries
Census TIGER/Line — ZIP Code Boundaries

The ZIP choropleth map on the Analytics page uses ZIP Code Tabulation Area (ZCTA) boundary polygons from the U.S. Census Bureau's 2020 Cartographic Boundary Files. These are publicly available shapefiles at 1:500,000 scale, optimized for web mapping use.

These boundaries are uploaded to Mapbox as a vector tileset and served as pre-tiled map tiles. The boundary data is static and updated only when Census releases new ZCTA vintages (typically following each decennial census). ZIP code boundaries do not precisely match USPS delivery routes — they are statistical approximations.

Mapbox — Map Rendering

All maps on RE/TERM are rendered using Mapbox GL JS, an open-source mapping library. Map tiles are served by Mapbox under their standard tile hosting terms. The base map style uses Mapbox's public light style. The ZIP choropleth layer uses a private vector tileset hosted on the RE/TERM Mapbox account.

Mapbox usage is subject to their Terms of Service. No user location data is collected or transmitted in connection with map rendering.

07 · Contact & Corrections

If you identify an error in any data displayed on RE/TERM, or have a question about the methodology described on this page, contact us at hello@re-term.com.

RE/TERM is a free tool built and maintained by Milairo. It is provided as-is without warranty of any kind. Use of this tool constitutes acknowledgment that you have read and understood the limitations described on this page.