From a small northwestern observatory…

I stumbled on a very interesting graphic on the inter-web the other day.  I can’t provide the citation just yet — it was posted anonymously on a data visualization web site.  Nonetheless, I’ve done a bit of research to semi-validate these numbers, and even if they’re off a bit, it’s a very useful graphic.

First, it tells us that since 2002, the median price of a new home in America has approximately doubled, from $175,000 to about $350,000 (depending on exactly which metrics you use, this is about right).  That’s an inflation rate of about 100%, more or less, in 15 years (end of 2002 to end of 2017).  In a paper I presented at the American Real Estate Society annual meetings about 10 years ago, I noted that post-WW2 data indicated that house prices/values should be expected to grow annually at a rate of about 2 percent points above the inflation rate.  I checked, and the actual inflation rate over that period measured by the CPI totaled 36%, more or less.  That averages about 2.1% per year, compounded.  The doubling of house prices in 15 years equates to an inflation rate of about 4.7%.  So…. 4.7 minus 2.1 = 2.6.  Thus, by my estimation based on historical averages, house prices have been growing about 0.6% per year faster than they should have since 2002.

You might argue that some of that was the last few years of the housing bubble, but that sponge got squeezed out in the post-bubble collapse.  Nope, folks, what we’re seeing is the echo bubble.  You might also argue that 0.6% doesn’t sound like much, but here’s what it amounts to over time.  If house prices had actually grown at the rate suggested by previous post-WW2 data, then prices would only have gone up by about 170% over that time period.  That means that a $175,000 house from 2002 should today be selling for about $295,000.  The difference (350,000 minus 295,000) of about 17% is the measure of the echo bubble — it’s the degree to which houses are currently overpriced.

Ahem…. that’s NOT the point of this story.  That’s just the introduction.  The more important story comes from deconstructing house prices into various tranches.  This graphic I found does a wonderful job of that:

Here’s my point in a nutshell. Note that in 2002, the plurality of homes built were in the “less than $200,000” category.  Today, that’s the smallest category (the one in red).  Conversely, we’re building about twice as many homes in the expensive category (the green bar) as we were in 2002. While all housing starts are down from the peak, compared to the earlier years, we’re now building the bulk of the housing in the two most expensive categories, which is a real shift from 2002.

Why?  The market is constantly screaming about the lack of supply for “affordable housing”.  Why aren’t builders building to that tranche of the market?  The answer is cost.  Two very disruptive forces are plaguing the homebuilding industry today.  First, the labor and infrastructure for building died off during the recession.  We have relatively fewer trained and skilled tradespeople, fewer developed lots (and a shrunken pipeline for development) and more expensive construction lending.  Second, the building materials themselves — lumber and steel — are in short supply, have been affected by price increases, and are now faced with tariffs.  Builders have no choice but to build more expensive homes to be able to cover the cost of construction.

Are we headed for a new bubble?  Back in the dark ages, when I was in graduate school, we were taught that inflation could be caused by either demand-pull (too much money chasing too few goods) or cost-push (increases in commodity costs).  Either way you look at it, the cost of owner-occupied housing is going thru the roof (pun intended).