The 1004MC “market conditions” appraisal addendum was created by Fannie Mae and Freddie Mac to require a form-based market analysis.  This added to the traditional practice to collect a handful of ‘comps’ to opine and foretell market price.  There are problems and unintended consequences.

The ANSI requirement is that residential appraisers measure houses according to the ANSI (American National Standards Institute) method.  This constraint can have good long-term results for consistency, if it is adopted by the full universe of participants.  The requirement would have to include tax assessors, building permit agencies, real estate agents, investors, insurance companies, architects, contractors, and unlicensed appraisers.  This part might take years, if ever.

We can take a quick look at some similarities, some differences, and where things may go.

What’s similar?

It appears similar motives may have preceded both the MC and the ANSI.  Each strives to either simplify or make consistent input or analysis.  Each requires showing fixed calculations to arrive at results.

Both the 1004MC and the ANSI requirements deal with variables that normally have high correlation of sale prices.  Living area, especially in homogeneous neighborhoods strongly predicts price.  Market conditions deal with changes in price levels.  The predictor variable is time.  A ‘market conditions’ adjustment is a misnomer, in that it combines changes in relative market conditions (supply/demand), and includes changes in the value of the dollar (inflation/deflation).  A “time” adjustment includes both macro and micro effects on nominal contract prices.

What’s different?

1004MC asks the wrong question, features the wrong data, and gives the wrong answer.  (Read more about that here from George Dell’s The Appraisal Journal article, Common Statistical Errors and Mistakes: Valuation and Reliability.) It asks about the heterogeneous neighborhood, not the homogeneous competitive market, displays irregular time periods, and gives misleading answers for time adjustments.  (Sometimes even in the wrong direction!)  Good appraisers effectively have ignored or fudged to get around the biased analysis.

On the other hand, the 1004MC did create controversy and attention toward the forgotten reality:  “Appraisal is market analysis.”

The ANSI measurement standard focuses on data collection, not prediction (making adjustments).  On the “collection” side, it strives toward consistency.  On the “adjustment” side, it can create a problem.  Different areas and different participants have different conventions for measurement.  For example, in some areas assessors have a different convention, then repeated by agents in MLS listings.  In such situations, there is a discrepancy or bias.  This bias reflects the consistent differences in measurement.  It will follow through and bias the final value opinion.

To avoid this type of analytical/information bias, appraisers (and pre-programmed automated models) will have to adjust for the bias.  (Perhaps call it a “measurement method bias adjustment?)

End results.

In the very long run, as licensed and unlicensed ‘measurers’ adapt to ANSI — the bias will disappear.  In the meantime, everyone will have to adjust.  In the shorter run, the GSEs (Government Sponsored Enterprises) are able to improve their future automated models based on the appraiser measurements.

A replacement for market conditions analyses requires two separate analytics:

  1. Identify the competitive/similar sales, not some heterogeneous group, such as neighborhood, or compiled published local trend. This is the core of modernized data-science based valuation.
  2. Set a trend (regression) line to match the actual similarity-matched sales, and apply the simple regression coefficient to the relevant market segment.

Our Stats, Graphs, and Data Science1 class, the Market Conditions class and related open-source (free) solutions provide a consistent, reliable, auditable path for market delineation, analysis, and visually understandable, definitive market price indexing.

Unfortunately, the largest investors, guarantors, agencies, and originators of real estate collateral loans do not yet require clear transparent analytics ensuring reproducible results.

Without demand for the superior analytic results, our public trust will continue to be threatened.

It is our hope that coming new regulation (and deregulation) will release the potential for truly modernized, unbiased, and reproducible financial valuation risk analytics.