Data Methodology
Sales and inventory data is from the Sullivan
County Multiple Listing Service. The Sullivan MLS does not include
all real estate sales made in the county, or all properties
in inventory available for sale. "For Sale By Owner"
homes, private sales, or sales of properties only listed in
a neighboring Multiple Listing Service (Greater Hudson Valley
or Ulster) but located in Sullivan County are not included.
Properties not listed in the MLS, for example offered directly
by a developer or "open listing" sales by brokers,
are also not included. Even with this limitations of the data
set, the vast majority of sales in Sullivan County are MLS-reported,
and trend data extrapolated from this data correlates highly
with the movements in the real estate market.
One very important consideration is that I only
include single family homes in the data. Other news sources
often report statistics based on a larger data set of "all
residential sales" which can include mobile homes, seasonal
homes and multi-family properties. Sullivan County has a larger
proportion of mobile homes and seasonal homes (unwinterized,
and only usable in the warmer months) than neighboring counties,
and the inclusion of these property types unfairly skews Sullivan
County property prices downward compared to other counties.
I try very hard to normalize data to make an apples-to-apples
comparison.
I generally pull data on the 2nd or 3rd of the
month for the previous month and quarter. At that point, not
all sales for the previous month may have been reported to the
MLS. Typically, an additional 3% to 4% of sales will be reported
after that, but seldom has a significant impact on the average
or median prices. I do adjust the data mid month.
Rather than reporting single month sales data,
I use a rolling 3 month average. This is purely a statistical
move, to ensure that there is a sufficient sample size. In a
single month, there may only be 50 or 60 sales, and one or two
particularly large sales can wildly skew the average with such
a small data set. By using a 3 month rolling average, I'm calculating
averages based on 150 to 200 sales, which minimizes the impact
of any large outlying sales and gives, in my opinion, a far
more accurate picture of market trends.
Return to Current Market
Trends page.
David Knudsen
845-887-5855
email: davidk@beechwoods.net