New Home Sales versus Housing Starts and Permits
Today, we will get the first New Home Sales (at 10:00am) after the stupendous plus 13% number last month. Recall our analysis of that back on November 30 found two interesting wrinkles:
• a plus 13% is meaningless when the margin of error is 17%;
• Historically (past 15 years) double digit gains are typically followed by flat to negative numbers.
Do note that New Home Sales are a very different animal than Housing Starts and Permits. The strong numbers we saw in Housing Starts and Permits are more reflective of how Builders do business than anything else. As I noted Wednesday:
From clients in the Real Estate industry, I have learned that Builders build -- and keep building. They do not try to guess when the cycle will end.
Why not? Look at their risk versus reward: If they guess wrong and stop building too soon, they miss out on some of the boom. If they build past the cycle's end, they may get stuck with excess inventory, forcing price cuts. But that usually means they are selling at a smaller profit; they paid much less for the land (bought years prior); it cost them relatively little to erect a house -- materials plus labor.
Even in the most severe slow down, the end game is merely the cost of doing business, and hardly dents the accumulated profits of the prior 3-5 years. So rather than just guess when the music will stop, they simply keep on building -- until sales slow dramatically.
Bottom line: New home starts and permit apps are not a leading indicator of the housing cycle.
I expect New Home Sales today to be flat to negative -- consensus is -8%, or 1,300,000 annualized.
Because of the limitations in the data gathering -- its a very small sample, with a large statistical error -- we could just throw out the November data.
Instead, lets think of it in terms of reporting the two months together, and then averaging them. A plus 13% and a minus 8% gives us about a ~5% growth rate in New Home Sales, which seems about right . . .
TrackBack URL for this entry:
Listed below are links to weblogs that reference New Home Sales versus Housing Starts and Permits:
» Builders Build: New Home Sales Are Not A Leading Indicator, Got It? from Matrix
The stats for new home construction showed more brisk activity than was expected [WSJ]. According to the Commerce Department: Housing starts increased 5.3% to a seasonally adjusted annual rate of 2.123 million units. Permits for future building ... [Read More]
Tracked on Dec 23, 2005 10:02:05 AM
And now the numbers are out, and even weaker than predicted, down 11.3 percent to a 1.245 million annual rate.
According to Bloomberg
"...homes for sale at the end of the month increased to a record 503,000 in November [and supply] at the current sales pace rose to 4.9 months, the most since December 1996 ... Sales fell in three of four regions."
As I noted in comments to an earlier post, Census Bureau housing statistics here, especially the Estimates of the Housing Inventory by Age of Householder in the Homeownership Data Historical Tables section, show that the percentage of surveyed households reporting that they owned their home rose steadily through the '90s after 1993 at about half a percentage point a year though construction ran only at rates in the 1.1 to 1.2 million range.
As the tables break the data down by age range, they clearly show that while ownership rose in all age ranges, it rose most among under-30s, whose rates are now at record highs. With a little spreadsheeting they also reveal that the number of under-50 households owning is 3 million higher than it would have been had ownership remained at 1993 rates; for all age ranges, the number is 4.3 million higher.
Note that the 4.3 million figure impies that over those 11 years an average construction pace of around 1.3 million units a year was 0.4 million above what would have held ownership rates constant, indicating that the constant-ownership rate is around 0.9 million new units a year (note that this takes second-home and investment buying into account).
If the constant-ownership construction rate is indeed 0.9 million a year, that makes the November inventory of 503,000 more than a 6-month supply at the constant-ownership rate, while builders are set to churn out homes at more than twice that rate. Holy Toledo.
We face an excess of home supply that would depress prices for years even without any forced selling by overextended homeowners and speculators who have bet everything on continually rising prices. Add that additional supply to the market and we'll be looking at a glut of stupendous proportions. (And if a flu pandemic takes out a few percent of the population ...)
Posted by: jm | Dec 23, 2005 11:54:04 AM
The comments to this entry are closed.