The new body of knowledge for valuation incorporates much of the old appraisal body of knowledge.  What is the difference?  The fundamental dividing line is actually simple.

Traditional appraisal practice was developed almost 100 years ago — in a dramatically data-different world than today.  There were two primary sources of data then.  One source was courthouse records, which documented transactions.  There was little information on physical features.  In many states, even the sale price was prevented from being disclosed.  The second source was other people, primarily real estate agents, lawyers, and participants in the transaction.  Almost every “comparable sale” required a phone call or a personal visit.

Our appraisal organizations, especially the Society of Real Estate Appraisers and the American Institute of Real Estate Appraisers, were formed to improve professional service, and also served as a source of information exchange—appraisers sharing information.  When I became a new appraiser, the MLS was still quarterly printed ‘sold’ books, with weekly or monthly listing books.  There were no compiled sources of income or land sales.  The best source was tax records confirmed by parties to the sale.  I attended three chapter meetings each month, plus a lender broker meeting.

The main point is that the appraisal body of knowledge was formed around collecting four or five comps to support your opinion.  Market knowledge was important.  Who you knew was even more important.

Then things changed.  Data became available, not only MLS electronically, but (in California) the Market Data Cooperative.  Soon after income, land, and even special use property data became available electronically.  As things changed, so did the challenge of analysis.  It changed from scraping together 4 or 5 comps, to discarding all but 5 or 6 comps.  None of the current legacy body of knowledge touches on the loss of accuracy and precision from the data discarding.  Data discarding is opposite to the Fisher Information function, which mathematically proves that more information is always better than less, but at a declining rate – until encountering unneeded variability.  GIGO some call it.

The modern methods, and the current “recognized” body of knowledge regarding data analysis – account for this optimal data size.  The point is it’s not always three or six.  It depends on the amount and quality of data available.  It may be two, it may twenty, it may be 20,000.

The quasi-governmental Appraisal Foundation requires the appraiser to “correctly employ those recognized methods and techniques”.  There is no reference to where these methods and techniques are to be found.

The Appraisal Institute has declared itself to be the recognized source of recognized methods and techniques in the form of its “body of knowledge” which has been exactly defined by its board of directors (February 2018).  It specifically includes Appraisal Institute books, courses, and “Body of Knowledge”.  It specifically excludes its own peer-reviewed The Appraisal Journal, and any reference to other professional organizations and non-AI literature.

There is no doubt that the AI materials are well suited to legacy valuation problems.  They are also well-suited to valuation problems where data is still paper-based, or restricted by legal or social power issues.

There is no doubt that AI material has been substantially accepted in law and judicial practice.  Legal implications can be heavy, if these self-declared authorities are accepted as the whole truth.

In the next part of this topic, we will consider other sources of accepted methods and techniques.  Some of these are quite recent, such as the robust data science body of knowledge, or quite extensive and deeply vetted, such as basic economics and econometrics.  Finally, we may also look at the broader source of valuation knowledge, including non-appraisal valuation methods and techniques.