By Karl Smith
U.S. Home Construction is coming in later than I had expected earlier this year, but seems to be largely tracking the pattern we laid out. There are, however, key differences between the actual path and the path I began promulgating late last year.
First up, here is where we stand. Housing starts are gapping upward:
From the long term view, we can see that this is a significant turn upwards, on the order of the jumps traditionally consistent with the end of a recession.
Let's zoom in to talk about timing:
That early 2012 stall out wasn’t “supposed” to happen, and whatever caused it seems to be responsible for housing's late arrival to the recovery. Indeed, it's deeper than that because there is likely a sort of increasing returns to scale, or amplification effect.
That was the same time that we saw car sales and consequently, manufacturing construction surging forward. If that had happened at the same time as housing, then a significant chunk of the blue collar workforce would have found itself with rapidly improving employment prospects all at once.
This, in turn, would have allowed a surge in demand for business that support them and so on through the multiplier effect.
Crucially, however, if it all happens at once, then suddenly an entire segment of the population has improved creditworthiness – or decreased money demand, if you want to look at it that way. This means that it is less difficult to make loans to them because the screening doesn’t have to be as careful. Everyone in this pool is doing better, so all loans should be less risky.
More lending, in turn, means less risk of cash-flow shortages, which means more intense household capital deepening (replacing the washer dryer, sending the kid to tech school, etc.), or equivalently, an increase in the natural rate of interest.
That’s the kind of thing we need to escape from the Zero Lower Bound.
Why it didn’t happen is a complex kind of question. A good candidate for the precipitating event was the seeming near breakdown of the eurozone that happened earlier this year and a resulting credit tightening. However, it is precisely the job of the Federal Reserve to offset such things. Moreover, it's not clear that some other set of events pushed this into motion.
For one thing, the composition of the housing turn-up is different than expected. Here is year-over-year growth in single family vs. multi-family:
It’s the absolute change in starts that determines contribution to GDP growth. As you can see, multi-family has been growing strongly for a while now, but it has not been growing at an ever faster rate. Given the build-up in rent pressure and the decline in vacancies, this is surprising.
An interpretation here is that this is a direct result of the failure of the Fed to make a credible commitment to be irresponsible. Though the fundamentals of the renting market were tightening quickly in 2011, banks perceived the Fed as unwilling to allow rents – the largest component of core CPI – to grow above trend.
Hence, in the face of high demand and rapidly improving fundamentals, banks were reluctant to lend. This hypothesis is bolstered by the fact that private equity started to move into the rental game. One way to think about private equity is as a way to get around constraints placed on you by the Fed. You drive down liquidity demand among a somewhat insulated set of rich people using personal relationships as a substitute for liquid traction instruments. This allows you to make far less liquid investments or take on far higher levels of leverage than would otherwise be possible.
Since we see that in the rental market, it is a sign that the rental property market is liquidity constrained and hence, is a place where private equity could earn abnormally high returns.
I don’t have time for much more, but I did want to look at absolute quarterly change in permits, as this is the best proxy we have for contribution to GDP growth. Indeed, in a dynamic period like this, it’s likely better than the real-time releases by the BEA. That is, I suspect in the years to come, the BEA will revise its 2012 estimates towards what quarterly permit change now suggests.
As you can see, absolute quarterly permit growth is closing in on levels seen during the 2000s, suggesting a contribution to GDP growth that may be inline with, and eventually even above, what we saw during the early 2000s.
This confuses some folks because housing construction is a small fraction of GDP. It makes huge percentage fluctuations, however. So, if housing construction is 2% of GDP and it doubles over 12 months, then its direct contribution to GDP is 2%. That’s not including any multiplier effects or amplification, and the 2% added on top of what GDP growth would have been otherwise.
That is, of course, assuming the Fed does not respond by cranking down on the rest of the economy, or that the housing growth was not prompted by monetary moves in foreign countries. In those cases, there would be some almost necessarily offsetting reduction in either Non-Residential Investment or Net Exports, respectively.