Methodology & data notes

Where the data comes from

CMAScope is built on 14 public research datasets, each loaded from its publisher and named on every chart it touches. The full catalog with coverage windows lives on The data; in brief:

Realtor.com® Economic Research — the listing backbone: monthly prices, inventory, days on market and price cuts (July 2016 onward; national, state, metro, county, ZIP), market hotness (Aug 2017+, ~300 largest metros), and the weekly national pulse.
Zillow Research — ZHVI home values (including bedroom-count and home-type segments) and ZORI observed rents.
Redfin Data Center — closed-sale medians, homes sold, months of supply, sale-to-list and sold-above-list shares for ~2,800 counties.
FHFA — the all-transactions House Price Index for long-run appreciation.
Freddie Mac PMMS — the 30-year mortgage rate series back to 1971.
U.S. Census Bureau — Building Permits Survey (new construction) and American Community Survey 5-year estimates (income, population, education, housing stock) for neighborhood context.
IRS SOI migration — county-to-county moves and the income they carry.
BLS LAUS — monthly county unemployment rates.
HUD — Small Area Fair Market Rents by ZIP.
FEMA National Risk Index — flood, wildfire, hurricane and tornado risk layers.

The hotness score

Realtor.com's hotness score is an equal-weight blend of two 0–100 percentile scores: a supply score based on how quickly homes go pending (median days on market — faster is hotter) and a demand score based on listing page views per property. A hotness rank of 1 is the hottest market at that geography level.

What we compute ourselves

Where source files ship month-over-month and year-over-year columns, we recompute every delta from the underlying history series at load time and reconcile against the published values (typically 100% within ±0.5pp). Derived metrics — price-cut share, $/sqft vs national, price percentile among peers, gross rent yield (annual ZORI rent ÷ ZHVI value), and the market momentum z-score (price direction up, inventory and days-on-market down = hot) — are calculated from the reconciled series. Cross-dataset joins are by county FIPS and ZIP code; each metric anchors to its own latest published month, since publishers run on different cadences.

Quality flags

Realtor.com flags month/geography combinations affected by coverage changes (quality_flag = 1). Rankings and screeners exclude flagged rows by default; trend charts keep them so series stay continuous. Roughly half of ZIP-level rows carry a flag in some months — ZIP data is best used for direction, not precision.

Known limitations

Realtor.com metrics are listing data — asking prices, not closed sales. Closed-sale truth comes from the Redfin series where county coverage exists.
Hotness covers only the ~300 largest metros and their geographies.
The weekly file is national-only and publishes YoY deltas, not levels.
Zillow segment coverage varies by market; small segments can be missing.
IRS migration is annual and lags about two years; ACS is a 5-year rolling estimate.
Small markets can swing hard month to month; we filter rankings by market size where noted.

Attribution

Every chart, export, embed and API response names its source — Realtor.com® Economic Research, Zillow Research, Redfin, FHFA, Freddie Mac, U.S. Census Bureau, IRS, BLS, HUD or FEMA — per each publisher's attribution guidelines. Full source list: data sources.