How many homes are people buying, and how much did they pay?
The go-to source for that information has long been the National Association of Realtors. But real estate data firm CoreLogic says the trade group's tally of the number of homes sold is too high by a substantial amount:
Historically, the CoreLogic existing sales data have covered about 85% to 90% of all NAR’s existing home sales data. However, in 2006 NAR’s sales data became elevated relative to the CoreLogic, MBA [Mortgage Bankers Association], HMDA [Home Mortgage Disclosure Act] and Census sales related data, and that trend has continued and become more pronounced through 2010. There are several reasons for the divergence, including benchmarking drift, more sales going through MLS systems due to consolidation and a lower share of for sale by owners (FSBO) home sales. Net, NAR’s existing home sales data are overstated by about 15% to 20%.
A CoreLogic economist told The Wall Street Journal that he sees it as the difficulty of adjusting data accurately in times of big changes -- the National Association of Realtors' statistics are built on a sample -- rather than a "gaming-the-numbers issue."
But some were quick to see ill intent from a group whose members benefit if people think it's a good time to buy. "Where are the subpoenas and Congressional hearings?" asked trader Karl Denninger on the Seeking Alpha blog, suggesting that people overpaid for homes as a result of relying on faulty sale data.
The National Association of Realtors is re-examining its numbers, The Wall Street Journal said. (Reuters reported afterward that the NAR was calling any overcounting "relatively minor.")
So what about the local figures?
They're reported by the company that runs the area's multiple-listing service, Metropolitan Regional Information Systems (specifically, its stats arm, RealEstate Business Intelligence). If there's any intentional adjusting going -- starting with a sample and extrapolating upward or trying to account for unlisted sales -- it would be news to me. So that seems to be a more solid starting point, at least. But the recent back-and-forth that came when MRIS's stats arm redesigned its statistics database wasn't exactly confidence-inspiring.
One useful comparison is the multiple-list sales vs. all recorded arms-length home sales. The multiple list should account for most but not all the arms-length sales, because not all homes are listed before they're sold. The state Department of Assessments and Taxation, as it happens, just shipped me data on all arms-length sales recorded in 2010.
Baltimore metro area home sales last year as originally reported by MRIS: just under 21,500.
Baltimore metro area home sales last year as recorded with the state: just over 23,900.
Does that mean everything's hunky dory? I'd need prior years to see if the difference is typical, and since I'm on vacation this week I don't have access to most of my files. But at least the multiple-list figure is less than the state's figure.
Statewide, that's not the case.
MRIS covers most of Maryland, with the rest handled by the Coastal Association of Realtors. Together, they show just under 50,900 homes sold last year, according to the tally by the Maryland Association of Realtors.
The state, meanwhile, has records of 47,817 arms-length home sales -- about 3,000 less than were supposedly sold on the multiple list.
Hmm.
Thoughts?