Regression For Real Estate Market Analysis

[Kurt Wilkening]

Hi All,

Any Realtors on the line here?

If I were to get the properties that sold in a neighborhood, i.e., sold
price, square footage, bedrooms, full baths, half baths, pool, screened
lanai, water view, garage bays, etc., could a regression analysis tell
me what an extra half bath is worth, or a pool, or any of the other
variables I just listed?

What considerations should I be looking out for?

Thanks,

Casey Wilkening
www.nwtampa.com

[ViAnn Beadle]

I'm not a realtor but given the old saw that location, location, location is
the most important factor in pricing I'd bet that you have to factor that
into any statistical model.

Try your regression and then see how well it actually predicts prices by
looking at residuals by location and other factors. You might run into some
serious multi-collinearity problems--its highly unlikely that you'd find a
house with one bathroom and a swimming pool.

I'd personally try a more classical CART model. I also think that googling
this topic might be helpful.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Kurt Wilkening
Sent: Thursday, August 23, 2007 6:07 PM
To: [hidden email]
Subject: Regression For Real Estate Market Analysis

Hi All,

Any Realtors on the line here?

If I were to get the properties that sold in a neighborhood, i.e., sold
price, square footage, bedrooms, full baths, half baths, pool, screened
lanai, water view, garage bays, etc., could a regression analysis tell
me what an extra half bath is worth, or a pool, or any of the other
variables I just listed?

What considerations should I be looking out for?

Thanks,

Casey Wilkening
www.nwtampa.com

[Ornelas, Fermin]

What you may want is to research the literature dealing with hedonic
estimation. You will find examples dealing with this type of situation.
I am thinking that price per square footage could be a function of your
predictors. Some of those predictors could be entered as binary
variables to represent the extra amenities found in the properties. They
could actually have an effect on the intercept and represent the
additional price increase. Or they could also be entered as interaction
effects with the continuous variables.

F. Ornelas

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Kurt Wilkening
Sent: Thursday, August 23, 2007 5:07 PM
To: [hidden email]
Subject: Regression For Real Estate Market Analysis

Hi All,

Any Realtors on the line here?

If I were to get the properties that sold in a neighborhood, i.e., sold
price, square footage, bedrooms, full baths, half baths, pool, screened
lanai, water view, garage bays, etc., could a regression analysis tell
me what an extra half bath is worth, or a pool, or any of the other
variables I just listed?

What considerations should I be looking out for?

Thanks,

Casey Wilkening
www.nwtampa.com

NOTICE: This e-mail (and any attachments) may contain PRIVILEGED OR
CONFIDENTIAL information and is intended only for the use of the
specific individual(s) to whom it is addressed.  It may contain
information that is privileged and confidential under state and federal
law.  This information may be used or disclosed only in accordance with
law, and you may be subject to penalties under law for improper use or
further disclosure of the information in this e-mail and its
attachments. If you have received this e-mail in error, please
immediately notify the person named above by reply e-mail, and then
delete the original e-mail.  Thank you.

[Marks, Jim]

Yes-- this is standard practice for municipal property assessment--
google "Municipal Property Assessment Corporation" for a CDN government
site.

You will need a large dataset of valid sales across a number of
neighbourhoods (the # of sales should be in the thousands to ensure
enough cases for the many property characteristics).

You will need
        area of lot
        area of improvements
        quality of construction
        age of improvements
        neighbourhood
        indicator variables for property features (type of garage, type
of lot, type of heating, fireplace...)
        indicator variables for the location-- corner lot, school, green
space, frontage

From experience-- lot size and  quality-adjusted area will get you 80%
of the variability.

The assessment of an individual property is tested by the sales
comparison method-- review of 4 comparable properties that were recently
sold

The assessment to sale ratio (ASR) is a primary measure of the quality
of the regression, and SPSS has added a number of ratio testing
statistics to support this.

--jim


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Kurt Wilkening
Sent: Thursday, August 23, 2007 7:07 PM
To: [hidden email]
Subject: Regression For Real Estate Market Analysis

Hi All,

Any Realtors on the line here?

If I were to get the properties that sold in a neighborhood, i.e., sold
price, square footage, bedrooms, full baths, half baths, pool, screened
lanai, water view, garage bays, etc., could a regression analysis tell
me what an extra half bath is worth, or a pool, or any of the other
variables I just listed?

What considerations should I be looking out for?

Thanks,

Casey Wilkening
www.nwtampa.com