We hear a lot about evidence. For valuation, there may be two relevant definitions. One would be the legal concept, the other being the “support” required to be credible, “worthy of belief.”
In the law, evidence is factual information. It can be testimony, documents, physical, demonstrative, or circumstantial. Relevance can be probabilistic.
Colloquially, evidence is anything that helps prove something is or is not true.
In appraisal, per USPAP, “relevant evidence and logic” is support for results, which are worthy of belief.
We are motivated by the recent appeal by the AI leadership, to HUD (and other regulatory agencies) to apply “evidence-based appraisal regulation.” The underlying motive seems to be the number of recent accusations of racial and other bias.
It is certainly a good move to ask government regulators and legislators to apply good evidence.
So what is evidence?
My school only teaches modern “Evidence Based Valuation” (EBV)©, (instead of the legacy “pick comps, and adjust” procedure). This topic is core for me, and for the CAA (Community of Asset Analysts), the group which uses and promotes the data science methodology. EBV is a giant step towards applying best evidence and best analysis, given today’s data and computer power.
Briefly, there are four “legal” types of evidence: real, demonstrative, documentary, and testimonial.
For analytic/academic evidence, we have six types:
- Empirical: measurable, based on data;
- Expert: judgment from recognized expertise;
- Theory-based: models of explanatory relationships and algorithms;
- Logical/conceptual: deductive/inductive inference, mathematical, or logic;
- Documentary: based on existing prior works of others;
- Anecdotal/perceptual: based on personal selection of individual cases.
Traditional appraisal depends on expert “comparable” selection, usually 3 to 6 sales, providing anecdotal, expert evidence. The data selection criteria is based on “perception/judgment,” but follows established theory. The credible stance is that selected ‘comparables’ are similar to the subject property.
EBV is based on the science of data, applying econometric principles, using “complete-data.” The applied data set is the full competitive market segment (CMS), not some “selected” comps. Expertise applies market-derived similarity parameters to the data. The use of complete information optimizes precision and accuracy, avoiding any selection bias from the old judgment-based selection process.
“Evidence-Based Valuation Enforcement” by regulators and policy-makers, is only possible with empirical evidence. Subjective appraisal methods cannot be objectively enforced!
The appraisal profession, the Appraisal Institute, and others – will have to begin to teach modern methods – if they expect respect.
The intuition is the same, the economic theory is the same.
The analytics are better structured. The results better serve our clients and the public trust.
My 20 year hope remains alive. We must give evidence if we expect evidence in return!