US New Home Sales October/November 2009

4 posts / 0 new
Last post
Michael Höhne's picture
Michael Höhne
Status: Silver Member (Offline)
Joined: Nov 16 2008
Posts: 119
US New Home Sales October/November 2009

It's really funny to see the old trick applied over and over again. On 11/25/09 it was reported that new home sales in the US climbed to 430,000 in October, up 25,000 or 6.2% from September. Economists had expected a decline of 1%.

That was great news a month ago, wasn't it? I mean it was a 6.2% increase instead of a 1% decline.

Today, 12/23/09, it is reported that new home sales in November declined to 355,000 or 11.3% and the October data was revised to 400,000.

That means:

  1. The revised October data (400,000) is 7.5% lower than the preliminary one (430,000). 7.5% is a lot.
  2. With the revised figure of 400,000, October had a decline of 5,000 or 1.2%, instead of the reported 6.2% increase. And this 1.2% decline would have met the expectations of 1%.
  3. The reported 11.3% decline in November of course is based on the revised data of 400.000. To be fair, it should be compared to the preliminary number of 430,000 though. This is a decline of 75,000 or 17.4%.

Sometimes it's just too obvious.




Subprime JD's picture
Subprime JD
Status: Platinum Member (Offline)
Joined: Feb 17 2009
Posts: 562
Re: US New Home Sales October/November 2009

Great work Michael. Its a shame that the play nasty games with these economic numbers. I wont be surprised when next month they revise downwards the no of existing home sales for Nov.

cmartenson's picture
Status: Diamond Member (Offline)
Joined: Jun 7 2007
Posts: 5975
Re: US New Home Sales October/November 2009

Yes, good work Michael.

You've illustrated the old "Beat by a Penny!" trick where numbers are fudged up, and then revised down well after the fact.  Like Charlie Brown trying to kick a football, the stock "market" never fails to fall for the trick, first by zooming up on the great news and then completely missing the importance of the downward revision the next month.

Once one sees the tricks for what they are, it becomes difficult to believe that there's anything useful contained within stock prices which are, at best, a semi-legitimate reaction to illegitimate information.

The point, of course, is that a nation that fudges and lies to itself at every turn cannot possibly fashion a credible plan to allocate resources towards their best uses. 

It's no different than passing gargantuan, unworkable bills with a clearly fraudulently low cost estimate that everyone and their dog knows is going to get blown out to the upside later on.  The current Health Care bill comes to mind.

I credit these behaviors to a nation too long without any meaningful restraints on its spending and borrowing habits.  And some day when the nation's credit card gets cut in half I expect the petulant displays of aggrieved disappointment to be off the charts. 

redpill's picture
Status: Member (Offline)
Joined: May 6 2010
Posts: 23
Re: US New Home Sales October/November 2009

The US New Home Sales survey, conducted by the Census, has a very small sample size, is prone to error, only covers single-family housing, and does not take into account any sales cancellations.  Their seasonality calculations also tend to skew things upward when you're coming off some of the worst declines in history.  In short, it's horribly flawed to begin with.  Their residential construction permit survey is not much better.  So whether they manipulate the data or not, the data series themselves are of limited use when it comes to relying upon them for month-to-month trending.

I happen to know of a new data source coming down the line that is far superior to the Census data.  Should be out within a few months.  Stay tuned.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Login or Register to post comments