What we expose, where it came from, how fresh it is, how often we refresh, and the minimum tier that unlocks it. Pair with the API reference for fields and the Data sources page for the upstream providers.
Every CBSA we score, with the latest signal, risk grade, and 12-month forecast.
cbsa| Field | Type | Notes |
|---|---|---|
| cbsa | string | 5-digit Census CBSA |
| name | string | — |
| signal | 'expand'|'hold'|'contract' | Nullable if unscored |
| risk_score | 'A'|'B'|'C'|'D'|'F' | — |
| yoy_price | number | Trailing-12-month HPI change (pp) |
| forecast_12m | number | Model 12-month forecast (pp) |
Full quarterly FHFA House Price Index series for every metro.
(cbsa, period_start)| Field | Type | Notes |
|---|---|---|
| cbsa | string | — |
| period_start | date | — |
| value | number | Indexed HPI level |
What the model said at each past run — enables walk-forward analysis.
(cbsa, run_at, horizon_months)| Field | Type | Notes |
|---|---|---|
| run_at | timestamptz | — |
| model_version | string | — |
| horizon_months | integer | — |
| target_period | date | — |
| point | number | — |
| lower_80 | number | null | — |
| upper_80 | number | null | — |
Metadata about each model run — version, timestamp, features.
id| Field | Type | Notes |
|---|---|---|
| id | integer | — |
| model_version | string | — |
| run_at | timestamptz | — |
| features | string[] | null | — |
| notes | string | null | — |
Raw observations for macro series (Fed funds, mortgage rate, CPI shelter, etc.).
(series_id, period_start)| Field | Type | Notes |
|---|---|---|
| series_id | string | e.g. FEDFUNDS, MORTGAGE30US |
| period_start | date | — |
| value | number | — |
Historical signal flips per metro with the trigger reason.
(cbsa, flipped_at)| Field | Type | Notes |
|---|---|---|
| cbsa | string | — |
| flipped_at | timestamptz | — |
| previous_signal | string | null | — |
| signal | string | — |
| trigger | string | e.g. 'forecast_crossed_threshold' |