Housing Starts Fall

Housing starts decreased in March, falling 62,000 from February to an annualized pace of 510,000 and holding near January's record low of 488,000. The tally was 30,000 fewer than expected. Starts had jumped 84,000 in February, almost entirely because of an increase in starts of multi-family units. Starts are low enough to continue to whittle down the supply of unsold homes, and they will likely stabilize in the current zone, reducing a major drag on GDP.

Single-family starts, which tend to account for the vast majority of starts, were flat in March after a small increase of 2,000 in February. Favorable weather conditions probably boosted starts in February, essentially borrowing starts from March. Supporting this idea are the figures on February starts in the Northeast and the Midwest, which posted gains of 70.3% and 49.2%, respectively.

Building permits decreased 10.8% in March following a 6.2% increase in February. At 513,000, the current level of permits suggests starts will hold fairly steady in their current zone in the time ahead, contrasting with last year's sharp declines.

Housing completions, which in January fell to a record low pace of 773,000, increased to an 824,000 pace in March from 796,000 in February. The recent decline in completions is extremely important, because until January, builders were still completing homes at a pace too strong relative to household formation, preventing inventory levels from falling more rapidly than they recently have. Now that fewer homes are hitting the market for sale, the growing U.S. population will have fewer homes to choose from. This will accelerate the recent decline in home inventories. This will undoubtedly be a game-changer for inventories and prices.

Doubters on the inventory idea will surely point to the difficulties that prospective homeowners face in obtaining credit to purchase homes. In doing so, they will ignore the most important concept, which is to compare the net change in the housing stock with population growth and household formation.

The concept is simple: a basic element of human survival is shelter, and the need for shelter increases along with the population. Housing starts have now fallen to levels well below what is needed to support population growth. Whether people can afford to purchase a home or obtain the credit necessary to do so is not as important as the fact that they need shelter and will rent space if necessary. The bottom line is that empty homes will become occupied one way or another so long as builders under-build relative to population growth.

Keep in mind the key metrics: the U.S. population increases by about 3 million per year. Household formation will average 1.2 million per year, although it tends to slow during recessions for obvious reasons. Eventually, however, people will do the inevitable and get their own space (note: the amount of births in 2007 was the most since 1957 in the U.S.). The current level of starts yields very few new shelters, say 350,000 or so, as many of the homes started are re-starts, homes rebuilt following damage from teardowns and storms and such. Considering that figure in relation to population growth and household formation makes one thing clear: Inventories are headed down.