Today we got the Census HVS data for 1Q18. In addition to providing the update on Household Formation Trends we thought it a good opportunity to redux the contextualization around a number of bigger picture dynamics.
- The Requisite Caveat: The HVS survey is timely and widely cited, but it’s volatile and doesn’t always comport cleanly with the more comprehensive annual Census/CPS housing surveys or a common sense reading of reality. We take the data with a grain of salt and, while we view the magnitude of change as a distorted reflection of the underlying reality, we view directional changes in the data as a largely accurate depiction of the underlying trend. In other words, while it typically gets the magnitude wrong, it generally captures the directional change correctly.
- Household Formation | Owner Momo ↑ : Total Household Formation increased by 1.18mn year-over-year in 1Q, rebounding against a relatively soft 4Q17 print. Most notably, however, is the continued absolute and relative improvement in owner households. As we’ve profiled previously, owner household formation grew at a premium to rental households for the first time this cycle in 1Q17 - a trend that has now persisted for 5 quarters with the spread between Owner and Renter Household growth holding above 1.0mn for a 4th consecutive quarter in 1Q18.
- Millennial Urban Rental Residency ↓: The decline in renter household formation accords with trends observed across major metro’s where the peak in Urban Millennial and Rental residency appears to officially be in. The crest in urban rental residency trends across Boston, Chicago and Los Angeles are pictured below.
- The Next 5 Years: The largest bolus of Millennials are currently 26-27 years old, the average rental duration is approximately 6 years and the average ave of the 1st time home buyer is 32/33 years old. In other words, the 4.8 mn 26/27 year olds are set to replace the ~4.3 mn 32-33 year olds (i.e. the current 1st time buyers) over the next 5 years as they progressively transition from rental demand to single-family purchase demand. In other, other words, 2020-2022 should see a strong bull market as that demographic demand wave pushes through the home purchase market.
- The Housing Cycle: Housing cycles tend to be long and autocorrelated, playing out fully from peak-to-trough and trough-to-peak. And Housing Turnover and peak New Construction activity are generally mid-cycle phenomenon …. except in the present instance. Housing was, of course, a proximate cause of the GFC and, as we remind investors recurrently, the housing recovery lagged the turn in the broader economic cycle by a full 3 years. Thus the housing cycle is, at least somewhat, de-synchronized from its typical, temporal tethering to the larger macro cycle. While the economic expansion is currently late cycle, the housing cycle itself remains only mid-cycle with activity still relatively depressed on a historical basis.
- Housing’s Asymmetry Hierarchy: Charts 5-8 below show the long cycle context for Housing Starts, New Home Sales and Existing Home Sales along with the current risk-reward balance to average peak and trough levels of activity. The asymmetry remains most positive for New Housing Starts and Sales and least positive for existing sales. There will be increasing, late-cycle chop to trade across housing related exposures and you do not want to be long interest-rate sensitive cyclicals alongside a rate shock but from a Trend and Tail perspective, the setup continues to carry one of the more favorable asymmetries in macro.