Warning:  Hardcore Real Estate Talk

(Like Lenn Harley would say)

I just found out that I have been using bad data on most of my sales analysis for years due to a flaw/weakness in my local MLS that may exist in many MLS's throughout the country.  This flaw is so profound that it could actually be impacting our national data on home sales!

It all started while gathering up the data for my August Sales Analysis when I decided to take a look at every property that sold, not just the statistical summary.  As I was going down the list I found that there were a good number of homes that shouldn't have been included in the closed August sales because they actually closed in July.

To give you an example of how bad those prior numbers were, in today's report the August sales for Lake Norman using the flawed search would be 79 rather than 62.  There are 22 sales in the August 2009 search that actually closed in July and 5 sales that are included in September when they actually closed in August for a net error of 17 ora 27% more than the actual. 

If I had not corrected the numbers I would have had 85 sales in July and 79 sales in August which would be interpreted as a pretty minor decline in sales rather than the actual -31% decline!

I called my MLS and was told that the standard MLS search for properties that have sold by date actually include the properties based upon when the listing agent entered the sold data into the MLS, NOT when the property actually sold! Now we all know how long it takes some agents to report their listing as sold.  Given the volume sales on the last day of the month, most of these actually fall into the next month's sales data!

I was told by my MLS that the only way to get the accurate sales numbers is to create a "Custom Search".  And sure, enough, using the custom search I got only properties that actually sold in August.

This in itself is very disturbing but here is what really worries me:

According to the person I spoke to, all of the statistics provided by our Charlotte Regional MLS use data based upon when it was entered by the agent, not the actual sales date!

If this is true for our large MLS, is this true throughout the country?  Is it true in YOUR MLS?  Is NAR using accurate data?  YIKES!