Friday, October 21, 2016

Signs of Recession

Bill McBride notices a cool down in the apartment market.  All four measures of market tightness from NMHC fell sharply in the 3rd quarter.  He seems to see this in real terms, as a trend shift in supply and demand, where supply has finally caught up to demand.  Multi-unit building has been relatively strong compared to the previous highly constrained couple of decades.  But, of course, housing starts in general are very low.  There is no way that the supply of homes has caught up to demand.

Here is the chart from Calculated Risk.

Rent inflation remains high, and appears to still be climbing, although the CPI rent measure can lag somewhat.  Mortgages outstanding are just barely growing and homeownership is still dropping, so there is no sign that the single family market is capturing any extra housing demand.

Maybe rent inflation will begin to wane.  But, if it does, I think this will be a sign of generally nominal contraction.

Notice that this level of looseness in the apartment market has previously been associated with the beginning of recessions.  I don't think this has as much to do with real housing supply as it has to do with money.  We are at the beginning of a nominal downshift, and this is one sign of it.  There is no lack of demand for housing.  There is a nascent decline in nominal activity.

Here, it may be worth revisiting BEA statistics on housing expenditures.  Nominal housing expenditures have run pretty close to 18% of PCE for 35 years.  It spiked in 2008-2010, because housing expenditures are sticky and nominal incomes dropped faster than households could adjust.  Consider this long-term flat trend.  Also, note that there was absolutely no rise in this measure from 2002-2006, during the supposed housing bubble, when Americans were recklessly overbuilding homes, as the story goes.

The chart title is wrong.  This is a measure of housing PCE as
a proportion of total PCE, not the other way around.
Basically, what the data tells us is that, in an economy burdened with Closed Access housing markets, where housing is perpetually undersupplied, residential investment is a freebie.  It simply transforms inflationary housing expenditures into real housing expenditures.  Oversimplifying a bit, it means that households live in 2,400 square foot homes instead of 2,000 square foot homes, but their expenditures remain the same.  There is nothing unsustainable about that.

The next graph compares the BEA measure for real housing and utilities expenditures to the measure for real total PCE.  Even in 2002-2006, some of that flat nominal spending on housing was inflationary as real housing continued to decline.  The housing bubble was a moral panic about something that really wasn't even happening.  That is not uncommon for moral panics.

Notice that the jumps in real housing expenditures all happen during recessions, and are a result of falling total PCE, not rising real housing expenditures.  The softening of the apartment market appears to insiders like it is a product of their supply and demand factors, and at very local levels, those factors certainly dominate.  But, in the aggregate, softening demand for housing is probably a sign of softening nominal demand in general.  Growth in real PCE has never really even moved back up out of recessionary levels.  Past contractions have tended to drop at least 3%.  Will real PCE growth drop to sub-1% in the next 18 months?

5 comments:

  1. Great post...and spooky. The Fed wants to tighten?

    Half the globe in deflation, industries everywhere with excess capacity. US economy weakening.

    Time for a rate hike?

    ReplyDelete
  2. Great post...and spooky. The Fed wants to tighten?

    Half the globe in deflation, industries everywhere with excess capacity. US economy weakening.

    Time for a rate hike?

    ReplyDelete
  3. Anecdotally (from personal experience), construction loans are getting harder to get and vrtually impossible without full recourse. At least that is true in the warehouse development sector.

    ReplyDelete
    Replies
    1. Interesting. What's the timeframe you are referring to? Has this been a recent change in the last few months?

      Delete
    2. I'm not in the market for those types of loans often enough to really comment on timeframe, but I'd estimate that the change has occurred in the last 6 months. In the past, we've found a thin market up to 55% loan to cost for non-recourse construction loans if we paid a premium (and provided recourse for completion but non-recourse once we got an occupancy cert). But right now, zero takers for 50% loan to cost or even lower. Not even with partial post completion recourse. Full recourse only. And this is for a project that is otherwise phenomenal.

      Delete