← Back to Research

Alternative Data

Federal Housing & Construction

Building permit momentum mapped across 3,135 US counties — and into 16 tradable homebuilder equities.

Permit acceleration precedes home price appreciation by one to four quarters, and is observable before official price indices update.

3,135
Counties covered
+0.099
Permit Information Coefficient (1-Quarter)
t = 5.91
65%
Homebuilder Out-of-Sample Hit Rate
Geographic panel
3,135 counties, 911 Metropolitan Statistical Areas (Fiscal Years 2010–2025)
Data sources
7 fused: Census, USAspending, Low-Income Housing Tax Credit (LIHTC), SAM.gov, FEMA
Equity coverage
16 homebuilder and building materials names
Validation
Geographic vs. Zillow Home Value Index · Equity out-of-sample 2020Q1–2025Q3

Primary Signal

Flagship
permit_sf_yoy

Year-over-year change in single-family building permits. Counties with rising permit counts reveal strengthening demand before it appears in home prices — typically one to four quarters ahead. The signal has been validated against Zillow ZHVI across all 3,135 US counties.

+0.099
1-Quarter Information Coefficient vs Home Price Index
5.91
1-Quarter t-statistic
+0.129
4-Quarter Information Coefficient

Quarterly Home Price Index appreciation by permit_sf_yoy quintile

Q1 = lowest signal quintile  ·  Q5 = highest

Top quintile counties appreciate ~3.5% per quarter versus ~0.8% for the bottom quintile — a 2.7pp spread that persists across market regimes. The 4-quarter Information Coefficient of +0.129 is stronger than the 1-quarter Information Coefficient, confirming the signal strengthens as prices catch up to the permit supply.

Why It Works

Permits are filed before construction begins. They reveal builder confidence and pipeline supply well before home prices adjust to the new supply outlook. Permit data is published quarterly — before the Zillow ZHVI and Case-Shiller indices that most market participants rely on for real-time housing signals.

hb_sector_adaptive_indexHomebuilder Equity Signal

The same permit flow data, mapped into each homebuilder's geographic footprint and weighted by sector-specific transmission factors. The lead OOS equity signal across 16 tradable housing-sensitive names.

+0.204
Out-of-Sample Information Coefficient (16 names)
4.25
Out-of-Sample t-statistic
65%
Hit rate (23 out-of-sample quarters)

Interested in Federal Housing & Construction?

Get in touch to discuss geographic coverage, the homebuilder equity signal, or how it fits your housing research.

Get in Touch →