Weak housing starts? Permits are a better indicator

April housing starts were sharply lower than expected (853K vs. 970K), but building permits were much stronger than expected (1017K vs. 941K). The past behavior of these two series suggests that in cases like this where starts are this low relative to permits, they usually come back into line with permits in the subsequent month. Thus it's reasonable to expect that starts will bounce back next month, and that the boom in residential construction activity is ongoing.

Not surprisingly, building permits and housing starts tend to track each other very closely, and it's reasonable to assume that builders must first acquire a permit before starting construction. Logic would thus suggest that the sharp divergence in these two indicators should be resolved in the direction of permits.

The above charts show the long-term history of the level of starts and permits and their ratio (on the bottom half of each chart). Note that the ratio averages just about 1 over time, which means that indeed the two series are measure two sides of the same thing: residential construction activity. The top chart gives a long-term view, while the bottom chart zooms in on the past several years. Note that April 2013 marked the lowest level of this ratio in the past 13 years. In fact, it was the lowest ratio in the entire history of these two series going back to 1960, and this strongly suggests that April starts were the outlier. Something similar occurred in December 2010, when permits rose strongly but starts were flat. The following month the ratio shot higher as starts surged.

Starts and permits have been rising at about a 30% annual pace for over a year, and there is no reason to think that this boom has suddenly come to an end.