Goldman Sachs economist that estimated new home demand in the United States (from the combination of new household formation and the replacement of old homes) of approximately 1.5-1.6 million units per year. Given that the U.S. population is around 300 million, this figure does not really stand out as being unreasonable, and it is in-line with other forecasts I have seen.
In the short term, current inventory combined with foreclosures, weak loan demand from the recession, and tighter credit standards all contribute to the fact that new housing starts in the U.S. today are near record low levels, coming in at an annualized rate of around 500,000 per year. At some point, however, it does seem likely to me that housing starts would have to begin to trend upward toward that 1.5 million figure, which is three times the current annual run rate.
Before you dismiss this potential need for new homes as being years and years away, consider the graph below showing annual U.S. housing starts from 1991 through 2009. (Click to enlarge)
You can easily see the effects of the housing bubble (from the early 2000’s through the 2005 peak of more than 2 million units), which resulted in home construction far outstripping demand (by 400,000-500,000 units if you use the 1.5-1.6 million base demand estimate). However, we also see if we ignore the bubble period that housing starts of 1.5-1.6 million per year would simply put us back to the level housing starts were in the mid 1990’s, when the U.S. population was much lower than today.
Despite the foreclosure glut we have in many states nowadays, this chart makes me think that the current housing start rate of 500,000 or so per year really is not sustainable for any prolonged period of time.
