Data through: Q2 2026 · updated Sep 4 · Updates: Quarterly · Next: Oct 29
QThe three bars show the last 3 quarters for this indicator, oldest left to newest right. Bar height reflects each reading relative to the other two — the tallest is the highest of the three, the shortest the lowest.
↓ 0.3 pp YoY
Historical context: Currently in the 11th percentile historically — near historically low levels.
What this means right now: Residential investment's share of GDP sits in the wide middle of its own history, where the cell stays neutral. It is colored only below the 5th or above the 95th percentile of that history.
Residential Investment Share of GDP · Quarterly · 1947–2026
NBER recessions
Scroll to read · zoom and pan in Expand
Monthly values
Residential Investment Share of GDP — the last 8 quarters
Every quarter's reading, with the band the heatmap gives it. The chart above shows the same series in full.
Quarter
Value
Band
Q2 2026
3.6%
Within normal range
Q1 2026
3.7%
Within normal range
Q4 2025
3.7%
Within normal range
Q3 2025
3.8%
Within normal range
Q2 2025
3.9%
Within normal range
Q1 2025
4.0%
Within normal range
Quarter-end readings of the same series the chart shows, from res_investment_gdp. Values are the current record, not point-in-time: a month that has since been revised shows its revised value.
What is the Res. Investment / GDP?
This series is residential fixed investment — new housing construction, improvements, and brokers' commissions — as a share of nominal GDP, from the BEA's national accounts, quarterly since 1947. It is the housing sector's macro footprint: how much of the economy's output is being spent building and improving homes.
On the heatmap it is colored only at its extremes, and most of the time the cell is neutral. It turns red when the share falls below the 5th percentile of its own history up to that reading. It turns green when the share rises above the 95th. The middle of its range says little either way: a high share marked both the 1950s building boom and the mid-2000s bubble, and a low share described the 2010s recovery for a decade.
The Housing Cycle's census measured its timing anyway: its four-quarter change correlates most with single-family starts three months earlier, coincident timing. That settled an old question of whether it belongs with the lagging members. It does not vote; it is context. The Housing Cycle's backdrop strip reads it today at 3.6% of GDP, lower than all but about 12.1% of quarters since 1947.
How We Color-Code the Res. Investment / GDP
Our heatmap colors this indicator against historically significant thresholds:
Below 2.9%
Below the 5th percentile of its own history up to the reading — an extreme low, as in 1982, 1991 and 2008–12
2.9% – 6.1%
The wide middle, where the cell stays neutral — about six quarters in seven since 1957
Above 6.1%
Above the 95th percentile — an extreme high, reached only in 2004–06, the housing bubble's peak
Historical Extremes — What Happened Next?
When this indicator reaches extreme levels, history shows consistent patterns:
1950 Q3Post-war building boom
7.3%
The all-time high as the suburban housing stock was built; the share drifted down for two decades
2010 Q3Post-crisis trough
2.4%
The all-time low, 2.4% in Q3 2010 — housing's footprint at a third of its 1950 share. It did not get back to 4% until late 2017.
2005 Q3Housing bubble peak
6.7%
The highest share since 1950 came in Q3 2005, at 6.7%. It fell to 2.4% within five years.
Frequently Asked Questions
How are the color thresholds set?
Only the extremes are colored. The cell turns red below the 5th percentile of the share's own history up to that reading. It turns green above the 95th. Today those edges are 2.9% and 6.1%, and everything between them is neutral. A past quarter's color reads against the history up to that quarter, as it would have read then. A color exists once the series has ten years of history, here from 1957. The colors describe where a reading sits in the record; they are not a forecast.
Why only the extremes?
Because the middle of its range has no clear direction. A high share marked the 1950s building boom and the 2005 bubble alike, and a low share described the long 2010s recovery. Colored by fifths of its history, about two in five quarters would have read red, the whole 2010s among them. At the extremes the share does track the sector. Every red quarter since 1980 came while the Housing Cycle read a downturn, a contraction or a bottoming, and every green quarter came while it read an expansion.
How often has each color appeared?
Rarely, by design. Through the second quarter of 2026, red appeared in 34 of the 278 quarters with a color, about one in eight. Green appeared in 8, all from 2004 to early 2006. The other 236 quarters were neutral.
What counts as residential investment?
In the national accounts: construction of new single-family and multifamily housing, improvements to existing homes, manufactured homes, dormitories, and brokers' commissions on home sales. It excludes the value of land and the imputed rent of owner-occupied housing, which are elsewhere in GDP.
Is this the same as construction spending?
They overlap but are not identical. Residential construction spending (also on the heatmap, year over year) is the Census Bureau's monthly value of work put in place. Residential investment is the BEA's quarterly national-accounts measure, which adds brokers' commissions and uses different source data. Their year-over-year movements agree closely: the two rose or fell together in 106 of the 130 quarters since 1994, through Q2 2026. Quarter to quarter the share is too small to read that way.
Does the Housing Cycle use it?
As a driver in the backdrop strip, never as a vote. Its timing was measured anyway, coincident with single-family starts three months after, which settled a design question about where it belonged. The strip shows the same color as the heatmap cell: neutral across the wide middle, red or green only at the extremes.