Friday, March 22, 2013

Lawler: The “Strange Case” of the NAR’s Regional Existing Condo/Co-op Sales

by Bill McBride on 3/22/2013 02:57:00 PM

From economist Tom Lawler:

The National Association of Realtors estimated that existing home sales ran at a seasonally adjusted annual rate of 4.98 million in February, up 0.8% from January’s upwardly-revised (to 4.94 million from 4.92 million, though unadjusted sales were not revised) pace. While the NAR’s estimate was slightly below consensus, it was above my estimate based on regional tracking – mainly, it appears, because of a “most strange” surge in condo/co-op sales in the South not reflected in local realtor reports.

The NAR estimated that existing SF home sales ran at a SAAR of 4.36 million in February, down 0.3% from January’s pace – an estimate that seems broadly consistent with regional realtor/MLS reports. In sharp contrast, the NAR estimated that existing condo/co-op sales ran at a SAAR of 620,000, up 8.8% from January’s pace – with condo/co-op sales on the South purportedly up 20.8% on a seasonally adjusted basis on the month, and up 29.4% on an unadjusted basis from last February’s pace. This gain seems especially suspect given that Florida Realtors (formerly the Florida Association of Realtors) reported that existing condo and townhome sales by realtors in Florida (the “condo capital” of the South) in February were up only 7.0% from last February’s pace! Quite frankly, the NAR’s estimates for existing condo/co-op sales don’t “smell” right.

Below is a table showing historical estimates of regional existing condo/co-op sales from the NAR, expressed as a seasonally adjusted annual rate. Note all the periods where seasonally adjusted sales in a region were exactly the same for at least three consecutive months.

As the table indicates, seasonally-adjusted sales show periods of remarkable stability in all four regions – partly, of course, because the NAR rounds seasonally adjusted sales to the nearest 10,000 (and unadjusted sales to the nearest thousand), which seems like “excessive” rounding. E.g., if “unrounded” condo/co-op sales on a seasonally adjusted annual rate basis in the Midwest went from 45,100 to 54,500 -- a 20.8% jump – sales rounded to the nearest 10,000 would show sales as being flat!). Conversely, if unrounded SAAR sales went from 54,500 to 55,100 – 1.1% increase – SAAR sales rounded to the nearest 10,000 would jump from 50,000 to 60,000, a 20% gain! According to the NAR, existing condo/co-op sales in the Midwest ran at a SAAR of 50,000 in ten of the 12 months of 2011.

When the NAR released its “rebenchmarking” of existing home last year, there were “astonishingly” big revisions in regional sales of existing condos and co-ops – suggesting either issues with the rebenchmarking, some serious issues with its previous methodology or both.

NAR estimates for existing condo/co-op sales only go back to 1999, and it first published monthly estimates in 2005.

2010Jan 100,00060,000200,000130,000
 Feb 110,00070,000210,000140,000
 Mar 110,00050,000200,000140,000
 Apr 130,00060,000210,000140,000
 May 100,00060,000230,000150,000
 Jun 100,00060,000210,000130,000
 Jul 80,00040,000170,000100,000
 Aug 90,00040,000190,000110,000
 Sept 80,00040,000190,000110,000
 Oct 90,00050,000180,000110,000
 Nov 90,00050,000210,000110,000
 Apr 80,00050,000220,000130,000
 Jun 100,00050,000190,000120,000
 Jul 70,00050,000210,000110,000
 Aug 120,00050,000200,000130,000
 Sept 100,00050,000210,000130,000
 Oct 80,00050,000210,000130,000
 Nov 70,00060,000210,000120,000
 Apr 110,00060,000210,000130,000
 Jun 90,00060,000220,000110,000
 Jul 100,00060,000220,000130,000
 Aug 110,00070,000240,000130,000
 Sept 110,00070,000250,000130,000
 Oct 110,00070,000250,000130,000
 Nov 110,00080,000280,000130,000
 Feb p110,00080,000290,000140,000

Last 10 Posts