American Downsizing: Smaller New Homes

By datatrekresearch in Blog American Downsizing: Smaller New Homes

As with yesterday’s note, we have 2 data topics to discuss today:

#1: We’re stuck on Fed watch until Chair Powell’s speech at Jackson Hole on Friday morning, so here is the latest from the Fed Funds Futures market:

Odds remain locked on a 25 bp cut in September, at a 98% probability.
Markets are still fairly certain the Fed will cut again in October, with 69% odds today. This is lower, however, than yesterday’s 80% probability.
Odds remain even on another cut in December, at basically 50-50. Yesterday they did skew modestly in favor of a fourth reduction in rates for 2019.

Why this matters: today’s release of the July FOMC meeting minutes should have been enough to make markets question their assumption of a steady diet of rate cuts through the rest of 2019. There was little support for a 50 bp cut in July, and most FOMC participants saw the 25 bp cut as a “recalibration of the stance of policy, or mid-cycle adjustment.” Not, in other words, the start of a steady path lower…

Our conclusion: we remain concerned markets are too sure of an easy Federal Reserve in the last months of 2019, and that is a recipe for further volatility.

#2: Following on our discussion yesterday regarding US household size, today we want to discuss the size of actual houses. As a reminder:

The number of people/household in the US has declined every year since 1960, going from 3.3 back then to 2.5 today. Much of this is due to single-person households, which grew from just 13% of the population in 1960 to 28% currently. Today this cohort outnumbers households with 4-or-more people, now 22% of the US population.
The reasons for this shift are both societal and demographic. Greater female workforce participation has allowed women to be economically self-sufficient and live alone if they choose. An aging population also adds to 1-person households as men predecease women by an average of 5 years.

The funny thing is that while US household size has been shrinking, the size of the houses they live in has been growing until very recently:

The average size of a new American home in 1973 was 1,660 square feet.
By 2010, the average new home was 2,392 square feet, or 44% larger than in 1960 when there were actually more people to house per dwelling.
Put another way, while the average number of people/household was shrinking by 16% from 1973 – 2010, the size of the new home being built was increasing by almost triple that.
More recent data from the National Association of Homebuilders shows that new home sizes are finally declining. By their measurement (not comparable to the Census data above), “peak house size” was in 2015 at an average of 2,700 square feet. Now, the average is 2,523, or 7% smaller.

How we interpret this data from an investment perspective:

#1: The combination of larger houses and small household size has been an important long run tailwind for the US economy. Larger homes need more “stuff”, and fewer inhabitants mean less sharing of the same. And this does not include the 2.3 billion square feet of self-storage space in the US, which stores everything those +2,000 square feet houses cannot hold.

#2: At the same time, a combination of lower household density, aging demographics, and increasing land costs means new house sizes will continue to decline. Yes, +4,000 sq ft mansions will still get built, but as the typical American household continues to shrink so will the marginal supply of new, larger dwellings.

#3: All this makes us cautious on the shares of homebuilding companies, regardless of where interest rates go. While there are many fundamental inputs that go into this group’s structural profitability, larger homes typically generate higher margins. “Peak house size” is not good news.