What is the Housing Starts?
Estimates monthly housing starts: new privately-owned housing units where excavation for the footings or foundation begins. Field representatives follow a sample of permits from about 900 permit offices and canvass sampled areas with no permit system, and results are adjusted to Building Permits Survey totals. All units in a multifamily building count as started at excavation, and the headline figure is a seasonally adjusted annual rate.
A start is excavation, not a permit or a sale, and estimates include units in structures totally rebuilt on an existing foundation. Starts come from a sample survey with 90 percent confidence intervals, and the Census Bureau notes it may take six months to establish an underlying trend for total starts.
Production facts:
- Publisher: U.S. Census Bureau
- Frequency: Monthly
- Typical release time: 8:30 a.m. Eastern on the 12th workday after the reference month
- Primary metric: Housing starts SAAR
- Market sensitivity: High
- Data type: Hard data
- Economic indicator type: Coincident
- Revision risk: Medium
Key components to watch
The headline (Housing starts SAAR) is the most widely quoted but rarely the most informative number. The components that follow reveal whether the headline result is broad-based or concentrated, improving or deteriorating, and consistent with prior trends.
- 1 unit (single-family)
- 2 to 4 units
- 5 units or more
- Four Census regions
- Authorized, not started
- Under construction
For any release with multiple components, the investor's task is not to memorize every number but to identify which components are most relevant to the current investment thesis and track those with the most care.
How to read the headline: actual versus expected
Markets react to the gap between the published Housing starts SAAR and the consensus forecast, not to the level alone. A stronger-than-expected or weaker-than-expected result is defined against that forecast, so check what was expected before reading the number.
The Housing Starts carries medium revision risk: Two months of data are revised with each month's preliminary estimates, and not seasonally adjusted monthly estimates are not revised again after the second revision. Read any revision to the prior period together with the new figure.
Market impact by asset class
The Housing Starts primarily affects the following markets: Homebuilders, REITs, Banks, Treasuries, Mortgage markets, Stocks.
Stronger than expected result: In an inflation-elevated, tightening environment, a stronger result typically pressures interest-rate-sensitive assets and can strengthen the dollar. In a growth-slowing environment, a stronger result can be equity-supportive if it reduces recession fears.
Weaker than expected result: In a growth-concern environment, a weaker result may reinforce concerns about economic deceleration. If the Fed is watching this specific indicator for policy calibration, a persistent pattern of below-consensus results may shift the rate path.
In-line result: In-line results typically produce muted market reactions unless there are significant component deviations from expectations. The market had already priced the consensus, so confirmation requires little repricing.
Revision risk and methodology notes
The Housing Starts carries medium revision risk. Two months of data are revised with each month's preliminary estimates, and not seasonally adjusted monthly estimates are not revised again after the second revision. With April data, seasonally adjusted annual rates for the previous 63 months are revised for updated seasonal factors. On average, preliminary seasonally adjusted starts are revised 3.8 percent or less.
Primary source: U.S. Census Bureau: Housing Starts.
Publisher documents read for this page:
Common investor mistakes
- Reacting to the headline without checking the components. The headline can diverge significantly from the underlying trend revealed by the components. A headline beat driven by a volatile component may not indicate the same strength as a broad-based improvement.
- Ignoring revisions to prior periods. An initial strong number revised down in the following release tells a different story than a number that stands firm.
- Not knowing what is consensus-priced. Reacting to a number without knowing what was already expected conflates information with noise. The market reaction is driven by the deviation from expectations, not by the level of the number.
- Applying the same interpretation across different regimes. The same headline result has different implications depending on the Fed's current stance, the growth trend, and what is already priced into the yield curve and equity multiples.
Related releases
The Housing Starts should be read alongside:
- New Residential Construction
- Building Permits
- Housing Completions
- New Home Sales
- Existing-Home Sales
Frequently asked questions
What is the Housing Starts?
Estimates monthly housing starts: new privately-owned housing units where excavation for the footings or foundation begins. Field representatives follow a sample of permits from about 900 permit offices and canvass sampled areas with no permit system, and results are adjusted to Building Permits Survey totals. All units in a multifamily building count as started at excavation, and the headline figure is a seasonally adjusted annual rate. It is published monthly by U.S. Census Bureau.
When is the Housing Starts released?
The Housing Starts is released by U.S. Census Bureau on a monthly schedule: 8:30 a.m. Eastern on the 12th workday after the reference month. Exact dates are on the publisher's release calendar.
How does the Housing Starts affect markets?
The Housing Starts primarily affects Homebuilders, REITs, Banks, Treasuries. Its typical market sensitivity is rated High. A result that comes in stronger than consensus expectations generally moves affected markets directionally, while a weaker result can have the opposite effect. The direction depends on the current macro regime and what is already priced into markets.
What is the revision risk for the Housing Starts?
The Housing Starts carries medium revision risk. Two months of data are revised with each month's preliminary estimates, and not seasonally adjusted monthly estimates are not revised again after the second revision. With April data, seasonally adjusted annual rates for the previous 63 months are revised for updated seasonal factors. On average, preliminary seasonally adjusted starts are revised 3.8 percent or less.