Let’s get new critical!

Today there is a confluence of three public policy issues around collateral appraisal.  And what has been argued as the failure of American “appraisal reform” policy.

They are:

• The dissonance between Technology, existing practice and published standards
• The aging and depletion of the existing supply of appraisers
• The failed effort regarding racial (and other) obstacles
• The failed policy to improve appraisal reliability

All this, in the culture of the “five frictions” holding back needed progress.  The “five frictions” is analogous to a bucket of crabs. Crabs at the bottom of the pile pull down any who attempt to escape. The frictions prevent improvements in the real results of safety, understandability, and the reliability of appraisals.  The five are: practice, required education, user expectation, state regulation, and published (Appraisal Foundation) standards.

The old appraiser ways we perpetuate are simply:  “pick comparables, then adjust.”  This was good some 50 years ago when the loan valuation model became the appraisal “opinion” of value.

Since appraisers had to be transparent, and comply with specific (licensing) procedures, it was inevitable that algorithmic computer methods would come to compete with mere human opinions.  AVMs (automated) opaque algorithms were simpler, faster, and above all – cheaper.  The lender/agent simply presumed some of what appraisers normally did in terms of assumptions, inspections, data selection, and explaining.  Easy.

Today, new appraisers are conditioned to the old ways and are not allowed to do “non-transparent” things.  Not allowed to leave out requirements like property inspection, comparable/neighborhood awareness, or finding unexpected things.  To work around these irritations, the law and client ‘rules’ (mostly FannieMae and Freddie Mac) provided exceptions to real appraisals.  Like “waivers” and third-party “inspections” by the non-licensed.

New appraisers and minority trainees continue to be taught by – or by substitutes for – the master-apprentice model.  Substitutes for.

This, in a world completely changed.  This world now is data.  Complete data, or substantially complete data, or a reality of sparse but relevant data.

The pressure to ‘opine’ unrealistic values, or even fraudulent numbers, continues.  With the system continuing to provide commissions and bonuses instead of protection to consumers and taxpayers.  Current evidence of substantial kickback “concessions” and incentives on sale prices, distract from real attention to the required “definition of market value.”

Worse yet, the recent FHFA Staff Working Paper 24-07 (William M. Doerner & Scott Susin) pointed out that some 90% of appraisals failed in the reporting of market conditions.

Hmmm.  I thought that appraisal is market analysis . . .

May be not . . .

Recent technology guidelines (AO-41) force new tech to fit old standards, old regs, and old expectations. We continue to anticipate, to teach, and to require a new generation (regardless of status), to do things the old, “established” way.  That are “recognized” methods, “expected” by users, and what their “peers” continue to do.

Let’s stick to a credible (worthy of belief) opinion, not strive for scientific reliability.

Do not let your crab friend out of the bucket!  No!  Teach the same ol’ ‘recognized’ way.

Click here for the next Stats, Graphs, and Data Science 1 class