With everyone throwing the kitchen sink into creating the illusion that this time housing has bottomed, seriously, this morning's report on housing starts and permits was set to be quite awkward: on one hand, realistic data accounting for weakness due to Sandy would have broken the housing momentum - many were expecting a far weaker than expected print precisely due to the Hurricane. On the other, the Census Bureau could have gone hog wild and completely ignored the same reality that apparently is impacting all other data points, and said housing starts soared to their highest number in 4 years, or a seasonally adjusted 894,000 in October, up 3.6% from a downward revised 863,000 in September, and well above expectations of a Sandy-driven decline of 3.7% to 840K. The CB opted for the latter, while adding a solid pinch of seasonal adjustment to the data, which not annualized and not seasonally adjusted rose from 77.8K to 77.9K sales. In this number was the drop in Northeast housing starts from 4K to 3.5K, the lowest since February. The mystery boost came in the West, where annualized starts rose from 198K to 232K, even as they dropped in the South and Northeast. Finally, and more irrelevant, housing permits dropped from 890K to 866K seasonally adjusted, even as the NSA number rose from 71.4K to 75.0K.
Finally with the trendline still far below historical, this too continues to be merely Fed-stimulus driven noise.
Housing starts and completions +6 month lag:
As a reminder: what it takes to sell a house these days: