by Calculated Risk on 6/18/2014 03:48:00 PM

From housing economist Tom Lawler:

Based on local realtor/MLS reports across the country, I estimate that US existing home sales as measured by the National Association of Realtors ran at a seasonally adjusted annual rate of 4.81 million in May, up 3.4% from April’s preliminary estimate (which I believe was too low – see below), but down 6.6% from last May’s pace. Folks who track local realtor reports and look at YOY declines in unadjusted data might be surprised by a May estimate showing a decent-sized monthly pickup in seasonally adjusted sales, as the YOY decline in unadjusted sales in May appears to be similar to (or even a tad larger) than that in April. However, (1) seasonally adjusted sales last May were 3.2% higher than last April; and (2) there was one fewer business day this May compared to last May, and business-day count affects the NAR’s seasonal adjustment factor.

Trying to use publicly-available realtor/MLS reports to project the NAR’s inventory estimate is very challenging in the spring. As I’ve written about before, the NAR inventory number from March to April always shows a substantially larger increase than realtor/MLS reports (or listings data) would suggest, while the April to May change is always lower than realtor/MLS reports (or listings data) would suggest. I’m not sure why, but I’m guessing it is related to differences in the “pull dates” of the NAR reports and the publicly-released reports. Realtor/MLS reports for May, however, clearly indicate that national listings showed a higher YOY increase in May than in April, and my “best guess” in that the NAR’s existing home inventory for May will be about 2.32 million, up 7.9% from last May.

Finally, I estimate that the NAR will estimate that the median existing SF home sales price in May was up by about 3% from last May.

Flipping back to April, local realtor/MLS reports strongly suggest that the NAR’s estimate for existing home sales in April – 4.65 million (SAAR) – was too low, with all of the understatement coming from the South region. I have local realtor/MLS reports from across the South region for April covering a total of over 108,000 sales, and these reports suggest that home sales in the South in April were unchanged from the previous April on an unadjusted basis. The NAR, in contrast, estimates that existing home sales in the South in April were down 4.9% YOY (NSA). I’ve been using local realtor/MLS reports to project NAR data (national and by region) for many years, and I’ve never seen such a huge “gap.” To be sure, some of the publicly-available realtor/MLS reports may be wrong. By the same token, some of the official “NAR” reports may be wrong. Also, I know that a few MLS in the South changed MLS vendors and delayed their release of home sales reports in April, which could have impacted NAR estimates (though probably not by that much). While I have no clue as to whether the NAR will revise its April estimate, I feel pretty confident that the NAR’s April estimate was too low, probably by 100,000 (SAAR) or so, and I would urge the NAR to “double-check” their April numbers.

CR Note: The NAR is scheduled to release the May existing home sales report on Monday, June 23rd.