Real Estate and the Normal Distribution

REAL ESTATE AND THE NORMAL DISTRIBUTION

Many things in nature, economics, and other areas have quantitative data which can be normally distributed.  If one were to graph this data, it would tend to look like a bell, whereby the majority of the data points would fall in the middle and fewer and fewer points would fall to the left or to the right of this central tendency the further the points were from this center.  Generally, this phenomonon is true in real estate.  If the sales price from a particular region for like homes sold in a relatively close period of time were plotted, then they would probably form a bell curve.  The central tendency would most likely correspond to the average of the sample, and the points which were to the left or right of the center would still be within a statistically predictable distance of the average.

No two houses are exactly alike, if only because they are in two different locations even when they are side by side, although there are usually more factors which make them different, such as liveable square footage, lot size, condition, amenities, and so on.  Of course, when an appraiser tries to determine the value of a given property, the appraiser will limit the sample to the best comparable properties and then make cost adjustments for the differences so that they most accurately represent the subject property.  When these points are  plotted without making cost adjustments, then they will likely fall within a normal distribution.

Some banks use this normal distribution, sometimes in lieu of and sometimes in addition to an appraisal, although they do not use it with complete mathematical accuracy.  They use an automated value model.  Suppose that a particular neighborhood had an equal combination of townhouses and individual houses and that the townhouse average price was $200,000 and the individual house average value was $400,000.  If these were lumped together, it would make the townhouses look more valuable and the individual houses look less valuable.  If all of the data points were plotted what should occur is not a single normal distribution, but rather two distinct normal distributions.  One would notice this immediately by seeing two high points in the curve, or a bimodal curve. (The mode is where the majority of the points fall.)  This would be a clear sign that two different types of data were being incorrectly grouped together.

Another example of where this may occur is in "up and coming" neighborhoods.  There are some established neighborhoods where the sales prices will fall right into a normal distribution.  However, in some types of "up and coming " neighborhooods, there will be a set of houses which are in need of significant work and some houses which have already been renovated and rehabbed.  Although the sales may take place around the same time and the houses are similar types of houses, there will be a clear difference in price between the two and it will show up on the graph.  Typically when this phenonomon takes place, it is a clear sign that it is an "up and coming" neighborhood.

For more information about your house or your neighborhood, contact Ron Trzcinski of Zenith Realty at 410-935-5844.