Home > XE Currency Blog > The apartment boom, revisited


XE Currency Blog

Topics4602 Posts4647
By New_Deal_democrat August 21, 2017 11:10 am
  • XE Contributor
New_Deal_democrat's picture
New_Deal_democrat Posts: 468
The apartment boom, revisited

One of the things I have been surprised about in the last several years is the relative lack of apartment construction. With the large Millennial generation in their 20s and 30s, the big increases in rent, and the collapse in supply during the housing bust, I figured there would be a boom.  And yet multi-unit permits and starts have stalled, while single family permits and starts continued to rise (until earlier this year), although more apartments and condos have been permitted since any time in the last 30 years:  

So my interest was piqued when I read an article forecasting that there would be record apartment compleitions this year:

How could that be if more new units haven't been started?

Well, apparently there is an answer.

The Census Bureau measures housing in four chronological phases: permits, starts, construction, and completion. While not every jurisdiction requires permits, in general housing should progress like a pig passing through a python.

That is certainly what it looks like when we compare multi-unit permits and completions:

The total number of completions varies very closely to the total number of permits, with roughly a one year lag.

But look what happens when we throw units under construction into the mix:

There has been a disproportionate rise in the number of apartments and condos under construction. The number now is very similar to the early 1970s when the Baby Boomer generation was in its peak apartment-dwelling years. It is very much an anomaly compared with the last 50 years.

So how could there be such an outsized jump in construction vs. either permits, starts, or completions?  The Census Bureau's numbers don't explicitly tell us, but we can make a pretty good inference.

Since (unlike every other phase) construction takes place over a number of months, if it were taking longer to construct each multi-unit project than has usually been the case previously, then we would see two things in the data: (1) a big relative jump in units under construction than either starts or completions, and (2) a relative delay in completions.

And that is exactly what we are seeing:

Anecdotally, what I have seen is a lot of bigger complexes built, on the order of what is shown in the photo below:

The large three, four, and even five story complexes are very different from the two story garden apartment complexes of the 60s, 70s, and 80s.

In any event, if the above scenario is correct, we ought to see multi-unit completions peak or plateau in the next 6 months or so. Once completions catch up, the next thing we ought to see is a further increase in the vacancy rate (since unfinished apartments and condos obviously don't count!), and then maybe finally we will see a significant abatement in the steep rise in rents:

Sent from my iPad

Paste link in email or IM