On the surface, today's Housing Data [3]was good. Yes, there was a miss in the housing permits number, which declined from a downward revised 939K to 902K, on expectations of a strong 942K print, but let's ignore that: after all bad news is good news (although as the chart below shows even this number was highly skewed due to seasonal adjustments and the NSA number hasn't really budged in the past year). But look at the housing starts: what a whopper: at 1036K, this was the highest print since June 2008 - great news, right? Not really, because the one key indicator here, single-family units, actually posted a sizable drop from 650K in February to 619K in March. The offset: construction starts of multi-family, aka rental units, which in March was a whopping 392K, a 83K seasonally adjusted surge from February, which brings the total multifamily starts to the highest since January 2006 at 423K. Of course, in January 2006, single-family units hit a record 1823K, or about three times as much as the March 2013 number. Thank you Fed and QE for making yet another capital allocation mockery as America is increasingly shifting into a nation of renters. At this pace expect multi-family starts to surpass single unit starts in 4-6 months for the first time ever.
Housing Permits seasonal vs non-seasonally adjusted number:
[4]
And Housing Starts: note the single vs multi-family divergence:
[5]
Source: Census Dept [3] [3]