by Calculated Risk on 6/20/2011 03:45:00 PM

CR Note: Existing home sales for May will be released tomorrow and the Pending Home Sales Index on Wednesday June 29th. Hopefully the NAR will provide an update tomorrow on the timing of the benchmark revisions.

From economist Tom Lawler: Closed Home Sales Down, Pending Sales Up in May

Based on data from local realtor associations/boards/MLS, it certainly appears as if existing home sales declined on a seasonally adjusted basis in May relative to April, with sales based on the National Association of Realtors’ estimation methodology likely to come in at around a seasonally adjusted annual rate of around 4.75 - 4.80 million, down from 5.05 million in April.

The incoming data also suggest, however, that the NAR’s Pending Home Sales Index – which took a surprisingly sharply tumble in April (down 11.6% on a seasonally adjusted basis from March) – rebounded smartly in May.

Trying to “build up” a pending home sales index estimate from local data is challenging, as not all associations/boards/MLS report “new” pending sales – i.e., contracts signed in a month – to the public. Indeed, not all associations/boards/MLS even TRACK new pending sales, including many in the NAR’s existing home sales sample – and thus, of course, do not report new pending sales to the NAR. According to the NAR, the sample size for its pending home sales index is about half as large as the sample size used to estimate closed existing home sales. The NAR’s PHSI data also only goes back to 2001, making seasonal factors somewhat “imprecise.”

In looking at associations/boards/MLS/etc that DO report on new pending sales/contracts signed, however, it appears as if there was a substantial rebound in pending sales on a seasonally adjusted basis.

Below is a table showing the YOY % change in the NOT seasonally adjusted NAR PHSI compared to the YOY % change in new pending sales/contracts written reported by various associations/etc¹ that together comprised over 52,000 pending sales last month. It is NOT a sample representative of the country as a whole, though it does have data from all broad regions.

	YOY % Change, New Pending Sales, 2011 v 2010		NAR PHSI (NSA)		NAR PHSI (SA)
	NAR PHSI	Select MLS	2010	2011	2010	2011
Jan	-4.4%	1.1%	74.3	71.0	90.3	88.9
Feb	-10.5%	-3.9%	88.3	79.0	98.9	89.5
Mar	-12.9%	-9.0%	119.7	104.3	106.2	92.6
Apr	-26.8%	-24.3%	133.4	97.6	111.5	81.9
May		27.9%	89.0		78.3

¹ Areas covered include the mid-Atlantic region from MRIS, Long Island, Lehigh Valley, Massachusetts, Indianapolis, Omaha, Des Moines, North Texas, Orlando, Northeast Florida, Albuquerque, western Washington, Sacramento, South Carolina, and Hampton Roads.

The table suggests that pending sales in the sample I’m looking at did not decline as much from a year ago as did pending sales in the NAR’s sample, though the YOY declines certainly follow a similar pattern.

Last year, pending sales plunged in most of the country from April to May following the expiration of the federal home buyer tax credit. From last May’s depressed level, however, the vast bulk of associations/etc. have reported sizable gains in pending sales – sufficient to suggest a strong rebound in the pending home sales index from April.

Of course, this May had one more business day than last May, suggesting that this year’s seasonal factor will be higher than last year’s -- meaning that the YOY % change in the seasonally adjusted number will be below that of the not seasonally adjusted numbers.

Still, after allowing for changing seasonal factors the incoming data suggest that the May pending home sales index on a seasonally adjusted basis is likely to show a double-digit gain from April’s level.