Everyone talks about the “market,” but no one tells us exactly what it is.

The Dictionary of Real Estate, 5th ed. Appraisal Institute, defines “Real Estate Market” as “Buyers and sellers of particular real estate and the transactions that occur among them.”  Close, but no kewpie doll prize.  I have had at least three graduate level classes in the topic, from econometrics, business, and statistics departments.

In the appraisal literature, especially The Appraisal of Real Estate, 14th ed., it uses the word as a noun, a verb, and several adjective meanings.  It is used over 2500 times in this important book.  We are admonished many times how important it is.  But no kewpie doll.  No bell.

This is something we must see and understand.  But we do not know how to specify it!  If we are to analyze a market, we must be able to measure and define it.  We can’t measure a market until we know what IT is!

In my research over the years, working to apply statistical and data science methods, this hit me in the nose.  Before I could do believable statistics, truthful graphs, and meaningful words, I would have to figure out how to identify a market.

Appraisal material did not help.  Business marketing classes helped some.  Econometrics classes helped a lot.  What I learned was that to analyze something, you have to first define it and measure it.

I had started teaching classes to appraisers on my own, because none of the professional organizations seemed to have any interest in what was to evolve into ‘data science’.

What I discovered was that you could not define a market in words.  You had to use numbers or categories, or at least put things in order.  It turned out that this was about “data types.”  Some were measure variables (like building size), some were categorical (like parking spaces), and others were ordinal (like view).

Worse yet, variables measured different economic things.  For real estate it includes transaction terms, (like financing and motivation), market conditions (time-series), spatial (location), and property features.

Geography, use, time, property type, and features have little or no direct association.  For example, for a given type of property, different buyers will have varying desires.  In statistics and econometrics classes, I learned to put a name to this problem.  It is called ‘dimensionality.’  Ugh!

But with ugh news comes yay news.  When you have statistical independence or geometric perpendicularity – the two variables are independent.  They can be treated separately from other variables.  For example, you can and should look at seller financing separately from property location.

This creates the possibility of simpler analytics.

And simpler analytics is what we learn in Evidence Based Valuation (EBV)© in Valuemetrics.info classes.  We use simple descriptive statistics, like mean, median, range, and percentiles.  We learn to use visuals, to help the human expert brain make sense out of the market.