Improving Starts, Under Construction, Completion Data


Tuesday we received the new home construction data. The problem is that once the report is released it is forgotten. We hit rock bottom for new home construction during 2009. Anything is better than that. The problem is that we need to look at the starts data, the under construction data, the completion data and the revisions to the recent data.


The starts data is at the highest level since 2008 - still lower than 1992-2007. The data looks good. There is improvement over the March levels for the past decade . The problem is that the levels we have had for the past decade have been lower than what we had seen since 1975, with very few exceptions. If you look at the rolling year level of starts you can see that we have had fewer starts during the past twelve months that we saw during 1993. We are even doing worse than March of  2008 while we in a full-blown housing recession. The rate of growth is slower than 1983 or 1993. Will it take 14 years to return to pre-recession levels of starts? will it take closer to 20 years?


We have returned to near peak levels of homes under construction. We need homes under construction in order to move to completions and completions to have sales. The under construction data has been surging past 1 million units for an extended period of time. The inventory level is up over 100,00 more units than 2016 and 200,000 more units than 2015.  We are also 23% lower than we were during the peak construction year of 2006.


The completions data is following a similar patter to the start data. The data for March 2017 is virtually identical to those seen during 1992.The monthly data is better than we saw during March 1992. The rolling year data is comparable to what we saw during March 1994.   The rolling year data almost mirrors the movement seen in the monthly data.


Three months of data is the minimum amount of data to measure a trend. Next month we will have four months of data.That being said the current trend for starts is better than 2016 with 180,600 starts versus 170,400 starts during the first three months of 2016. This is a pace that is still slower than 1982, 1993, and 2007. If we are slower than 1993 as we were improving and slower than 2007 as we were in decline then you know that we are at a slower pace of starts than 1993 through 2007.


The completions trend data is better than the starts data. We have 246,200 units completed through March. This is comparable to the rate we saw during 1983. It is a better pace than 1992 or 2016 and is still worse than where we were during 2007.


This is going to be a long recovery process. People lost homes during the recession. People went through bankruptcies and foreclosures which tainted their credit records and possibly their appetite for buying homes. We need more full-time jobs for people to qualify for mortgages. We have a sluggish existing home market with exceptionally low inventory levels during the past few months. People who buy new homes often have owned existing homes. The existing home sales data will be released April 21st. The New Home sales data will be released April 25th.  We have seen upward pressure on both new and existing homes.  Are we living in 2017, 1992, or 1983?


It's the economy.

Jack Dunn - Reclaiming Common Sense