You would think that the biggest single industry in America would have really good reporting of sales activity and other market measurements.
Mercifully we do not labor with fake news, but we do suffer under a mountain of alleged news which is not news, and big headlines above non-news.
This week’s click-magnet headlines: “July Sales of New Homes Collapse 9.4%!” followed in couple of days later by “Big Declines in July Sales of Existing Homes!”
First the math, for which I apologize. The definitive reports of home sales are the monthly result multiplied by twelve months to get an annual figure, and then re-computed as the percentage change in the annual figure. If we multiply anything by twelve, we’re going to artificially magnify any actual month-to-month result.
Example: July’s existing-home annualized rate of sale was 5,440,000 versus 5,551,000 in June. Both figures also “seasonally adjusted” by wet-thumb-in-breeze. Any real statistician would say that the variance — 1.3% — is statistically insignificant, and inside the variation just by survey error.
A common sense view: five-and-a-half-million houses is a hell of a lot of houses, and sales are fine. Despite another quirk in data suppressing the July result, it was 2.1% over July one year ago.
Sales of new homes… same song, different verse. New homes sell at an annual rate only a little more than 10% of existing homes, always a tiny market by comparison, which even more magnifies minor changes month-to-month. The July annualized figure was 571,000, down from 630,000 annualized in June. Not many grown-ups remember their multiplication tables to 12, but my calculator reveals that the annualized decline of 59,000 units on a monthly basis is only 4,900. A decline of 4,900 sales from June to July is an accounting error. Remember that new home sales are calculated by contracts written, not closed, heavily affected by fallout (or not).
Year-over-year sales of new homes did faint 8.9% when comparing July 2016 to July 2017, but beware of too-narrow comparisons. July 2016 was an on-fire month, a spike. If we compare the first seven months of each year, 2017 is up 9.2% compared to 2016.
Stripped of all statistical entanglements, the housing market is right where we thought it was: doing the best it can without inventory to sell. In any comparable economic recovery going back to the 1970s, we should be building more than one million new homes each year. Eight years past the Great Recession, new construction is still rising but is barely 60% of need.
Blame lies all over the place. Labor shortage: those who left building trades after the burst of the Great Bubble never came back. Mortgage credit: many say too tight, but I doubt it — what good would it do to finance homes for people only to see the foreclosure rate rise? Prices: new home construction and sales may be limited by cost, too high for too many entry families.
Stick with that last one, price. The key component: land — land is extremely scarce, especially in places with high demand. More than half of the counties in the US are losing population, housing demand there is net-negative. The out-migrating population is moving to urban centers, housing demand super-positive and land very, very scarce. That’s the phenomenon to focus on, missing entirely from aggregate national reports of home sales. (Three charts follow)