Gimme Shelter – US Housing Inflation

By in
Gimme Shelter – US Housing Inflation

Lost in much of the chatter about US inflation rates is one simple fact: the cost of housing is the single most important driver of changes in consumer prices. A few numbers here:

  • Shelter (the catchall name for the category in the Bureau of Labor Statistics data) is 32.7% of the US Consumer Price Index.
  • Shelter is 41.4% of core CPI, which excludes Food (13.3% of the CPI) and Energy (7.7%).
  • The largest piece of the Shelter is “Owners’ Equivalent Rent”, at 68% of the category. OER represents 23.6% of the entire US CPI (more than Food and Energy combined) and 29.7% of core CPI.

Owners’ Equivalent Rent is a funny measure, because the BLS collects it by surveying consumers about what they think their house/apartment would rent for and then tracking those responses over time. We can’t help but think there is a TV reality show lurking somewhere in this analysis. “So you think you can rent your house for $3,000… Let’s find out!”

On a more practical note, this makes OER very difficult to forecast. The BLS doesn’t give much color on the data that goes into this calculation (link at the end of this section for what they do share), so reliably predicting/replicating this measurement is difficult.

Fortunately, there is a data-driven approach to assessing and forecasting OER: Zillow data. Everyone’s favorite site for seeing what their friends and coworkers really paid for their house has a more useful function. It works like this:

  • Every month Zillow collects the asking rental prices for properties across the country, both in aggregate and defined by key characteristics (house vs. apartment, number of bedrooms, etc.)
  • Zillow then calculates a median asking price per price per square foot, again both in aggregate and by major category.
  • By looking at national trends in asking rental prices across all types of housing, you can essentially see where OER inflation will be in future months.

The correlation between the Zillow data and OER works best on a six-month lead basis. The correlation in percent changes since 2011 (first Zillow data available) is 0.65, or an r-squared of 42%. The half year lag makes intuitive sense: changes in asking prices take some time to filter through to actual rentals and then inform the BLS’ survey takers’ perspectives on local rents.

So what does the current Zillow data tell us about future CPI readings? Two points:

#1. US headline and core inflation should both accelerate in the second half of 2018. Rental asking prices were soft in the first 4 months of last year, rising by less than 2% year-over-year. This improved from June onwards, with Zillow reporting rental comps of 4-8% since June 2017. The average for Q1 2018 was a +5.7% increase in rental asking prices, or 4x the average increase since 2011.

#2. The increases should not, however, be dramatic enough to put pressure on the Federal Reserve to raise rates more quickly than currently anticipated.While rental asking prices are firmer now than much of the last 6 years, the year-on-year comps are no worse than prior peaks in 2015 and 2016. Yes, other CPI inputs like Energy may accelerate, but remember that OER is still +3x more important to consumer inflation math than gasoline or heating oil.

Bottom line: the Zillow data shows that Shelter inflation will very likely rise over the balance of 2018 and drag overall CPI inflation higher.

BLS Description of Owners’ Equivalent Rent: